USF Libraries
USF Digital Collections

The effects of American Diabetes Association (ADA) diabetes self-management education and continuous glucose monitoring ...

MISSING IMAGE

Material Information

Title:
The effects of American Diabetes Association (ADA) diabetes self-management education and continuous glucose monitoring on diabetes health beliefs, behaviors and metabolic control
Physical Description:
Book
Language:
English
Creator:
Meisenhelder-Smith, Jodee
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Health belief model
Glycosylated hemoglobin
Diabetes patient care assessment
Diabetes outcomes
Outpatient program
Dissertations, Academic -- Public Health -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Abstract:
ABSTRACT: The purpose of this study was to determine whether adults with type 2 diabetes participating in American Diabetes Association (ADA) diabetes self-management education (DSME) randomly assigned to an intensive follow-up group (IFG), utilizing continuous glucose monitoring system (CGMS), or a standard follow-up group (SFG) have any significant differences in mean HgbA1c values and health belief scores over time. Baseline HgbA1c values and health beliefs were measured using the revised Expanded Health Belief Model (HBM) questionnaire. The questionnaire measured the 8 HBM domains: perceived susceptibility; severity; treatment benefit; cues to action; motivation; barriers; self-efficacy and structural elements. Twelve weeks after DSME, patients returned for follow-up based on random assignment. The SFG received routine follow-up care: HgbA1c measurements; behavioral goals and education assessments. The IFG received routine follow-up and CGMS.^ Patients wore the CSMS for 72 hours and recorded their daily food, blood glucose values, medications and physical activities. Results were analyzed and reviewed with patients. Both groups returned in 24 weeks for HgbA1c measurements and to complete the HBM questionnaire. A repeated measure ANOVA analysis showed a statistically significant reduction in mean HgbA1c at each time period (F=86.75. p>.0001 ) from week 1 to week 12 (SFG 8.6-7.1; IFG 8.5 --7.1,) and from week 12 to week 24 ( SFG 7.1 to 6.9; IFG 7.1 --^ 7.0). There were no significant differences found between the groups. (F = 0.17 p > 0.87). Following DMSE and follow-up intervention some health belief scores improved but no significant differences were found between groups except for severity scores. (SFG 27.05, IFG 25.00, p=0.03). The power of the study to detect small differences between the groups was affected by the higher than anticipated attrition and the significant lowering of HgbA1c in the education arm of the study. Both groups achieved a high success rate (58% IFG; 55% SFG) to lower the HgbA1c to the ADA goal of less than 7. DSME and follow-up care (both standard follow-up and more intensive follow-up) achieved a significant lowering of HgbA1c (1.6%), which has been shown to reduce diabetes related morbidity and health costs.
Thesis:
Dissertation (Ph.D.)--University of South Florida, 2006.
Bibliography:
Includes bibliographical references.
System Details:
System requirements: World Wide Web browser and PDF reader.
System Details:
Mode of access: World Wide Web.
Statement of Responsibility:
by Jodee Meisenhelder-Smith.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 387 pages.
General Note:
Includes vita.

Record Information

Source Institution:
University of South Florida Library
Holding Location:
University of South Florida
Rights Management:
All applicable rights reserved by the source institution and holding location.
Resource Identifier:
aleph - 001920107
oclc - 187304443
usfldc doi - E14-SFE0001837
usfldc handle - e14.1837
System ID:
SFS0026155:00001


This item is only available as the following downloads:


Full Text

PAGE 1

The Effects of American Diabetes A ssociation (ADA) Diabetes Self-Management Education and Continuous Glucose Moni toring on Diabetes Health Beliefs, Behaviors and Metabolic Control by Jodee Meisenhelder-Smith A dissertation submitted in partial fulfillment of the requirement s for the degree of Doctor of Philosophy Department of Community and Family Health College of Public Health University of South Florida Major Professor: Rober t J. McDermott Ph.D. Carol Bryant, Ph.D. Elizabeth Gulitz, Ph.D. Jeffery Kromrey, Ph.D. Brendan O’Malley, M.D. Date of Approval: November 14, 2006 Keywords: Health Belief Model, Glycosyl ated hemoglobin, Diabe tes patient care assessment, diabetes outcome s, outpatient program Copyright 2006, Jodee Meisenhelder-Smith

PAGE 2

Dedication To my dad, Joseph Meisenhelder who insp ired me to be a diabetes educator, to my mom, Twila Meisenhelder-Hiatt, my husband, Michael Smith, and all my family: Tyler, Dwana, Leslee, Barbara, Herb, Jean, Joe, Jolie, Cliff, Clifford, Clodagh, Bill, Brigette and Philip.

PAGE 3

Acknowledgements Special thanks to the staf f of the Diabetes Care Institute, University Community Hospital, Tampa, Florida, my doctoral committee, Monika Wahi, M.P.H. and MiniMed Medronic.

PAGE 4

Note to the Reader The original version of this document contains color that is provides assistance in understanding some of the data presented. The original dissertation is on file with the US F library in Tampa, Florida.

PAGE 5

i Table of Contents List of Tables ........................................................................................................v List of Fi gures...................................................................................................... vii The Effects of American Diabetes A ssociation (ADA) Diabetes Self-Managementviii Education and Continuous Gluc ose Monitoring on Diabetes Health Beliefs, Behaviors and Metabo lic Contro l.................viii Chapter One: Introducti on...................................................................................1 Diabetes Self-Managem ent Educat ion......................................................6 Statement of the Probl em........................................................................18 Purpose of the St udy...............................................................................25 Significance of the Probl em.....................................................................27 Hypothes is...............................................................................................30 Research Q uestions ................................................................................31 Research Plan.........................................................................................34 Assumpti ons............................................................................................39 Delimitati ons............................................................................................40 Limitati ons................................................................................................41 Definitions of Terms.................................................................................42 Chapter Two: Review of the Lite rature...............................................................47 Overview of Diabet es...............................................................................48

PAGE 6

ii Etiology and Cla ssificati on............................................................48 Epidemio logy.................................................................................50 Diabetes Related Complicat ions...................................................51 The Expanded Health Belief Model.........................................................54 Factors Affecting Me tabolic C ontrol.........................................................61 Diabetes-Specific Health Belief Inst ruments............................................66 Comprehensive Diabetes Self -Care Management Traini ng.....................74 Continuous Glucos e Monito ring...............................................................81 Implications for Fu ture Res earch.............................................................83 Chapter Three: Methods.....................................................................................86 Purpose of the St udy...............................................................................87 Hypothes is...............................................................................................89 Research Q uestions ................................................................................90 Research Plan.........................................................................................93 Population and Sample.................................................................93 Procedur es....................................................................................99 Physiological Inst rumentatio n.....................................................100 Psychosocial Instru mentati on......................................................103 Expanded Health Belief Model Questionnaire Reliability and Validity Re sults.................................................................108 The ADA Diabetes Self-Care Man agement Training Program....128 Intensive Follo w-up Care............................................................144

PAGE 7

iii Documentat ion............................................................................146 Data Collection and Analysi s......................................................147 Summary...............................................................................................150 Chapter Four: Results ......................................................................................151 Introducti on............................................................................................151 Participant Charac teristics.....................................................................152 Drop Out versus Complete rs.......................................................159 Research Question 1 result s..................................................................182 Research Questions 2-11 result s...........................................................188 Chapter Five: Discussion ..................................................................................196 Summary...............................................................................................197 Limitati ons..............................................................................................200 Conclusion s...........................................................................................207 Implications for Diabetes Education, Clinic al Practice, and Glycemic Cont rol.........................................................................208 Implications for P ublic Heal th.................................................................211 Lessons Learned and Recommendations for Future Research.............213 Referenc es.......................................................................................................217 Appendice s.......................................................................................................253 Appendix 1: Assessment and E ducation Records and Baseline Evaluation Form..........................................................................254 Appendix 2: Sample Flyer fo r Advertising Program..............................260

PAGE 8

iv Appendix 3: Inform ed Consent Form....................................................262 Appendix 4: St udy Forms ......................................................................276 Appendix 5: Follow-up Evaluation Form and Dex Glucometer Monitoring System Summary......................................................302 Appendix 6: Diabetes Self-manag ement Education Schedule..............308 Appendix 7: Patient Sati sfaction Questi onnaire....................................310 Appendix 8: Educati onal Materi als........................................................313 Appendix 9: Division of Nursi ng Standards/Policie s/Procedures..........316 Appendix 10: Box-plot and Stem-leaf Di agrams of Continuous Variables at Baseline ..................................................................324 Appendix 11: Chronbach Alpha Result s for Individual Questions.........330 Appendix 12: SAS Output fo r Research Q uestions ..............................339 About the Au thor...............................................................................................388

PAGE 9

v List of Tables Table 1: Exclus ion Crit eria .................................................................................95 Table 2: Minimum Sample Sizes per Group for Te sting the Hypothesis of Equal Group Effects in a Two Group RM Desi gn...........................97 Table 3: Revised Diabetes Health Belief Model Questionnaire Formulae ..........................................................................................110 Table 4: Reliability of Health Belief Scales and Subscale s..............................113 Table 5: Cronbach Alpha Results for Health Be lief Scales and Subscales at Ti mes 1 and 2.............................................................115 Table 6: Diabetes Self-car e Practice A ssessment ...........................................121 Table 7: Matrix of the Educational Sessions and Applications of the Expanded Health Beli ef Model C oncepts...................................136 Table 8: Characteristics of Study Populatio n Before Education Interventio n.......................................................................................148 Table 9: Baseline Characteristics by Tr eatment Group: Categor ical Covariat es........................................................................................154 Table 10: Baseline Age, Health Belief Scor es, and Hemoglobin (HgbA1c) by Group........................................................................ 156 Table 11: Characteristics of Completers vs. Dropouts: Cat egorical Covariat es......................................................................................161

PAGE 10

vi Table 12. Characteristics of Completers vs. Dropouts: Contin uous Covariates at Baseline ...................................................................163 Table 13. Characteristics of Completers vs. Dropouts by Treatment Group: Categorical Covariat es......................................................168 Table 14. Characteristics of Comple ters vs. Dropouts by Treatment Group: Continuous Co variates ..................................................................172 Table 15. Baseline Characteristics by Treatment Group for Those with Complete Follow-up: Cat egorical Cova riates .................................177 Table 16: Baseline Characteristics by Treatment Group for Those with Complete Follow-up: C ontinuous Cova riates ................................191 Table 17: Results from Repeated Measur es ANOVA Analysis of Hemoglobin at Weeks 1, 12, and 24, Control vs. Intensive............185 Table 18: Hemoglobin (Hgb1ac ) at 1-year Follow-up ......................................187

PAGE 11

vii List of Figures Figure 1: St udy Desi gn......................................................................................38 Figure 2: Theoretical Fr amework: Factors Influenc ing Metabolic Control of Diabet es.........................................................................................63 Figure 3: Model of Diabetes Educ ation..............................................................67 Figure 4: The Health Belief Model as Pr edictor of Diabetes Self-care Behavio r.............................................................................................85 Figure 5: Pilot St udy Result s...........................................................................102 Figure 6: Loss to FollowUp.............................................................................160 Figure 7: Mean Hemoglobin at Weeks 1, 12, and 24 by Gr oup.......................186 Figure 8: Baseline Health Be lief Profiles by Group..........................................189 Figure 9: Week 24 Health Be lief Profiles by Group.........................................190 Figure 10: Health Belief Profiles in Differences fr om Baseline to Week 24 by Group.........................................................................191 Figure 11: Health Belief Profile s at Weeks 1 and 24 by Gr oup........................192

PAGE 12

viii The Effects of American Diabetes Association (ADA) Diabetes SelfManagement Education and Conti nuous Glucose Monitoring on Diabetes Health Beliefs, Behavior s and Metabolic Control Jodee Meisenhelder-Smith ABSTRACT The purpose of this study was to determine whether adults with type 2 diabetes participating in American Di abetes Association (ADA) diabetes selfmanagement education (DSME) randomly assigned to an intensive follow-up group (IFG), utilizing continuous gluc ose monitoring system (CGMS), or a standard follow-up group (SFG) have any sign ificant differences in mean HgbA1c values and health belief scores over time Baseline HgbA1c values and health beliefs were measured using the revi sed Expanded Health Belief Model (HBM) questionnaire. The questionnaire meas ured the 8 HBM domains: perceived susceptibility; severity; treat ment benefit; cues to action; motivation; barriers; selfefficacy and structural elements. Twel ve weeks after DSME patients returned for follow-up based on random assignment. The SFG received routine follow-up care: HgbA1c measurements; behavio ral goals and education assessments. The IFG received routine follow-up and CG MS. Patients wore the CSMS for 72 hours and recorded their daily food, blood glucose values, medications and

PAGE 13

ix physical activities. Results were analyz ed and reviewed with patients. Both groups returned in 24 weeks for HgbA1c measurements and to complete the HBM questionnaire. A repeated measure AN OVA analysis showed a statistically significant reduction in mean HgbA1c at each time period (F=86.75. p>.0001 ) from week 1 to week 12 (SFG 8.6-7.1; IFG 8.5 –7.1,) an d from week 12 to week 24 ( SFG 7.1 to 6.9; IFG 7.1 – 7.0). Ther e were no significant differences found between the groups. (F = 0.17 p > 0.87). Following DMSE and follow-up intervention some health belief scores im proved but no significant differences were found between groups except for se verity scores. (SFG 27.05, IFG 25.00, p=0.03). The power of the study to det ect small differences between the groups was affected by the higher than anticipated attrition and the significant lowering of HgbA1c in the education arm of the study. Both groups achieved a high success rate (58% IFG; 55% SFG) to lowe r the HgbA1c to the ADA goal of less than 7. DSME and follow-up care (both standard follow-up and more intensive follow-up) achieved a significant lowe ring of HgbA1c (1.6%), which has been shown to reduce diabetes related morbidity and health costs.

PAGE 14

1 Chapter One: Introduction Diabetes mellitus is a major public health concern associated with substantial morbidity, mort ality, health care utilization, and costs (American Diabetes Association, 2005; Wagner, S andhu, Newton, McCulloch, Ramsey, & Grothaus, 2001; Gilmer, O ’C onnor, Manning, & Rush, 1997; Carter Center of Emory University, 1985; Ray, Thamer, Ta ylor, Fehrenbach, & Ratner, 1996; Ray, Willis, & Thamer, 1993; Rubin, Altman, & Mendelson, 1994; UKPDS Group, 1998; Westerhall, Olson, & Destafano, 1992; Winburger, Cowper, Kirkman, & Vinicor, 1990). The Centers for Disease Control and Prevention (CDC)) estimates over 20 million Americans are afflicted with the disease (CDC, 2006). Diabetes contributes to nearly 200,000 deaths annually, and is the leading cause of kidney failure, blindness, leg and foot amputations. Persons with diabetes are between two and four times more likely to have heart disease, and five times more likely to have a stroke than non-diab etics (Bellenir & Dresser, 1995). The estimated direct and indirect costs of diabetes are $132 billion (ADA, 2005). Research over the past decade has found that many diabetes-related complications once thought to be an inevit able progression of the disease can be prevented with strict glycemic control. The Diabetes Control and Complications Trial (DCCT, 1993) demonstrat ed that by keeping blood glucose levels as close to normal as possible, microvascula r complications such as retinopathy,

PAGE 15

2 neuropathy, and nephropathy can be averted. Patients with insulin dependent diabetes mellitus (IDDM) who already s howed early vascular complications were able to slow the progression of their comp lications with tight glycemic control. Fishbein and Palumbo (1995) also r eported delay of onset and slowing of progression of diabetes-related complicat ions. Several studies, including the DCCT, provided overwhelming evidence that a direct relationship exists between the level of glycemic control and the risk of developing complications such as diabetic retinopathy, neuropathy, nephropathy, and other vascular complications (DCCT, 1993; Klein, Klein, Moss, Davi s, & DeMets, 1984; Nathan, Singer, Godine, Harrington, & Permu ter, 1986; Ohkubo, et al ., 1995; UKPDS 33, 1998). In the United Kingdom Prospective Diabet es Study (UKPDS), a clear relationship between control of hyperglycemia and reduction of diabetes morbidity and mortality with type 2 diabetes strengthened the evidence for aggressive glycemic control in all diabetes populations. Intensive diabetes management has relied heavily on blood glucose monitoring. Persons with diabetes w ho regularly monitor their blood glucose levels and work closely with their heal thcare providers have fewer diabetes complications. In one study, people who increased the frequency of their selfmonitoring of blood glucose or exercise after attending DSME lowered their HgbA1c by a mean of 2.9% (Clement 1995). The Intern ational Diabetes Center, (IDC,2001) practice guidelines recommend type 2 patients not on insulin should

PAGE 16

3 test their blood glucose three times per day when newly diagnosed, adjusting therapy if test results ar e outside the target range. Type 2 patients on insulin should test their blood glucose four times a day. Diabetes care requires a high degree of self -care. Persons with diabetes are often asked to make major lifestyle changes such as practicing glucose selfmonitoring, maintaining reasonable body weight, adjusting meal composition, size, and timing, following a medication regimen and performing other preventive practices. Behavioral changes such as m odifying food intake and exercise were found to be the most difficult adjustment s to make (Robiner & Keel, 1997). Simplifying regimen complexity and prov iding guidance about the specific meal plan or exercise recommended may help ov ercome perceived barriers. Glucose monitoring barriers such as fear of needles and soreness to fingers can be addressed to increase patient’s accept ance and willingness to practice selfmonitoring. New glucose monitoring techniques can be used to help identify glucose excursions and help patients understand the importanc e of monitoring their blood glucose levels. New, less in vasive glucose monitoring procedures can assist in motivating the patient as to the importance of the various diabetes therapies.

PAGE 17

4 A new glucose monitoring device approved by the Food and Drug Administration (FDA) in 1999, the Continuous Glucose Monitoring System (CGMS) developed by MiniMed Inc., provides a novel and minimally invasive technique that continually measures glucose levels usi ng a miniscule (0.5ul) sample of dermal interstitial fluid. The CGMS records ti ssue glucose levels at five-minute intervals for up to three days. Previous studies hav e shown that dermal interstitial glucose measurements by the CGMS closely correlate to capillary glucose meter measurements and venous blood laboratory va lues (Jansson, Fowelin, Smith, & Lonnroth, 1988; Service, O’Brien, Wi se, Ness, & LeBlanc, 1997; Bolinder, Hagstrom-Toft, Ungerstedt, & Arner, 1997). Although capillary self-monitoring of blood glucose (SMBG) levels has been a mainstay in intensive management of diabetes, the true variations in blood glucose levels may not be detected by conventional SMBG recording. The glucose sensor data in conjunction with frequent SMBG measures can be helpful to identify glucose excursions incl uding asymptomatic hypoglycemia. The information obtained from the sensor can facilitate decisions concerning diabetes management and provide evaluation of treatm ent options. The sensor may also be used to help patients see the glycemic effects of food intake and exercise directly, and may motivate or act as cues to facilitate behavior change. A pilot study using the MiniMed continuous gl ucose monitoring system in nine poorly controlled type 1 subjects (mean HgbA 1c 9.9%) showed a reduction of mean

PAGE 18

5 HgbA1c to 8.8 % in a five-week period (Bolinder, Hagstrom-T oft, Ungerstedt, & Arner, 1997). This study attempts to evaluate the potential for continuous glucose monitoring to identify glucose patterns and help patients understand why appropriate changes in their diabetes management is needed. During the followup visit perceived barriers to self-car e can be addressed. Patients can be assisted with any difficulty in following their meal plan, exercise, taking their medications or moni toring blood glucose. Achieving good metabolic control is dependent on effectively educating and motivating the person with diabetes to comply with the medical regimen and adopt certain self-care behaviors. This st udy seeks evidence as to whether the data obtained from the sensor can lead to improved HgbA1c values by facilitating understanding of patient-specific patterns of glucose control and in facilitating diabetes self-management. T he proposed hypothesis also suggests that patients receiving intensive follow-up care are mo re likely to improve their diabetes selfmanagement behaviors, and thus, improve their glycemic control than those receiving standard follow-up care.

PAGE 19

6 Diabetes Self-Management Education The concept of diabetes education is not new. Diabetes education has been viewed by experts as an essential par t of diabetes care. Dr. Elliott P. Joslin, one of the pioneers of diabetes educ ation and creator of the renown Joslin Diabetes Clinic in Boston, Massachusetts stated that “the diabetic who knows the most lives the longest” (Krall, 1985, p. 465). Clement (1995, p. 1204), defined diabetes self-management educati on as a “process of providing the person with diabetes with the knowledge and skills needed to perform self-care, manage crises, and make lifestyle changes required to successfully manage this disease.” Most researchers agree that a mi nimal threshold level of diabetes knowledge is essential before any impr ovements in metabolic control can be expected. Factors such as the patient's beliefs, a ttitudes, medical treatment, medical history, psycho-social suppor t, and environment may also influence whether a person with diabetes is willi ng or able to make the necessary behavioral changes to improve metabolic control (Alogna, 1980; Anderson, Fitzgerald, & Oh, 1993; Ary, Toobert, W ilson, & Glasgow, 1986; Beeney & Dunn, 1990; Cerkoney & Hart, 1980; Glasgow, Hamp son, Strycker, & Ruggiero, 1997; Haire-Joshu, 1988; Herman & Dasbach, 1994; Rubin & Peyrot, 1992; Schafer, McCaul, & Glasgow, 1986; Schlenk & Ha rt, 1984). Meta-analyses of diabetes education by Norris (2003), Brown ( 1993; 1988) and Padgett, Mumford, Hynes,

PAGE 20

7 and Carter (1988) concluded there is an overall beneficial effect of diabetes education. Although these earlier studies unanim ously reported the beneficial effects of diabetes education, they seldom focu sed on tangible clinical outcomes such as improved glycemic control, redu ced incidence of diabetes-related complications, decreased utilization of heal th care, or healthcare costs. One exception, a study conducted by Miller, Goldstein, and Nicolaisen (1978), found that after implementing a comprehensiv e diabetes self-management training program, the rate of acute complications requiring hospitalization was reduced. This dramatic reduction in complications was estimated to save three million dollars in healthcare costs. During the 1970s, groups such as the American Diabetes Association (ADA), the American Association of Diabetes Educators (AADE), and the Diabetes Education Study Group (DESG ) of the European Association began initiatives to establish diabetes educ ation standards. These standards were designed to provide more structure and le ss variability in quality of education received by patients with diabetes. In 1983, the National Diabetes Advisory Board established by Congress develo ped the National Standards for Diabetes Education. Diabetes selfmanagement education (DSME) programs accredited and recognized by the American Diabetes A ssociation can be measured in terms of structure, process, and outcomes. Program content was standardized to

PAGE 21

8 include an overview of diabetes, stre ss and psychosocial adjustment, family involvement and social support, nutrition, exercise and activity, medications, blood glucose monitoring and ketone testing, interpretation and application of test results, relationships among nutrition, exercise, medications, and blood glucose levels, prevention, detection, and treatm ent of acute and chronic complications, foot, dental and skin care, behavioral change strategies, benefits of good glycemic control, community resources, and special considerations for travel, sick days, and pregnancy (ADA, 1996b; 199 8). The National Standards for DSME are designed to define quality diabetes self-management education that can be implemented in diverse settings and will facilitate improvement in health care outcomes. Standards are, therefore, periodically reviewed and revised to reflect advances in scientific knowledge and health care. ADA approved programs must have standar dized curricula that meet the needs of their target populat ion. Instructional objecti ves are assessed to ensure that patients are equipped with the knowledge and skills to manage their diabetes day-to-day. Participants practi ce self-care skills such as insulin administration, glucose monitoring, meal planning, and getting actively involved in planning goals and strategies to overco ming barriers to self-care. Programs meeting the standards set by the National Advisory Bo ard are recognized and accredited by the American Diabetes Asso ciation. ADA approved education programs employ a goal-oriented, empow erment approach that may enhance the

PAGE 22

9 patients' ability to identify and accomplish realistic goals. Problem-solving and coping skills are also utilized to help patients and their families deal with the barriers that interfere with goal achi evement. ADA approved programs employ multi-disciplinary team mem bers who must meet stringent qualifications such as having valid professional licenses and regi stration, and experience in their area of expertise. Certification as diab etes educators is recommended and 16 hours of approved diabetes education prin ciples are required per year. Studies have shown that structur ed diabetes selfmanagement education programs have been successful at improvi ng metabolic control (Garrard, Mullen, Joynes, McNeil, & Etzwiler, 1990; Glasgow, Toobert, & Hampton, 1992; Mazzuca, et al., 1986; Peyrot & Rubin, 1994). Reductions in the number of hospitalizations, ketoacidosis episodes, and amputations have been attributed to comprehensive diabetic education pr ograms (Miller & Goldstein, 1972; Mulhauser, et al., 1983). Structured di abetes education that stresses both knowledge and self-care behaviors assumes a causal path from learning to changed patient performance, and from altered performance to changes in clinical and psychosocial outcomes (Mazzuca, et al., 1986). Peyrot and Rubin (1994) found that pat ients who improved their diabetes self-care behaviors improved their glycemic control. Assessment of the priority populations needs help to maximize t he effectiveness of self-management education. Educational programs that identify demographic variables, ethnic

PAGE 23

10 background, formal education level, diabetes health beliefs, barriers to participation, and other fact ors in diabetes self-managem ent are more effective than ones just applying a didactic approach (Ingersoll, Orr, Herold, & Golden, 1986; Peyrot & Rubin, 1988; Rubin, Peyr ot, & Saudek, 1993; Skyler, Seigler, & Reeves, 1982). Peyrot and Rubin (1994) suggested in their study that it is possible to help patients who are at different levels of self-care proficiency. A step-wise approach to diabetes self-care management helping individuals to improve where they need the most improvement can favorably influence t heir glycemic control. Whereas some patients may advance from not adhering to certain aspects of diabetes care to better adherence, others may acquire advanced self-care skills. For DSME to reduce HgbA1c values, participants must not only learn the diabetes care skills, but also be motivated to maintain them. A further goal of diabetes education is to help patients dev elop coping skills to overcome their attitudinal and emotional barriers to di fferent aspects of diabetes management. Educators discuss potential options fo r addressing problematic situations. The most frequently utilized behavioral model for understanding and predicting diabetes self-car e behavior is the expanded Health Belief Model. Although the model is c oncerned with the person’s obj ective environment, and past history and experience, the main c oncern is with the person’s subjective state or personal orientation. The H ealth Belief Model (HBM) proposes that a

PAGE 24

11 person is likely to take action concer ning a health condition if the perceived benefits of action outweigh the perceived costs or barrier s. Barriers are defined “as those things which arouse negative f eelings or attitudes toward a given action” (Salazar, 1991, p. 129) Even if an action is perceived as beneficial in reducing a threat, if it is also perce ived as inconvenient, expensive, unpleasant, painful or upsetting, the person may still decide not to take action. A variable added recently to the HBM to strengthen its ability to predict and explain behavior is Bandura’s concept of self-efficacy (1977a; 1977b; 1992). Self-efficacy is defined as an individual’s beliefs about their ability to perform the behavior. Bandura postulated that self-e fficacy is one of the most important prerequisites for behavior change. The Expanded Health Belief Model (B urns, 1992), which incorporated selfefficacy, has been useful in explaining wh y some persons with diabetes are able to improve their metabolic control pos t-intervention and others are not (Alogna, 1980; Becker, Maiman, Kirscht, Haefner & Drachman, 1977; Cerkoney & Hart, 1980; Harris, et al., 1982; Ha rris, Linn, Skyler, & Sandi fer, 1987; Pham, Forton, & Thibaudeau, 1996; Wdowik, Kendall, & Harris, 1997). The Expanded Health Belief Model theorizes that for persons to modify behav ior, they must perceive the benefits are greater than the perceived disadvantages. The Expanded Health Belief Model is es pecially useful when expl aining how the patient's attitudes can influence self-care behaviors and ultimately, metabolic outcomes.

PAGE 25

12 Applying the components of the Health Be lief Model, (the perceived susceptibility to diabetes related complications; perce ived severity or threat of these complications; perceived benefits of aver ting these complications; perceived psycho-social and financial cost of tr eatment, perceived self-efficacy) help to predict who will be compliant with diabetes self-care behaviors. In a comprehensive review, Janz and Becker (1984) provided strong empirical support for the HBM and reco mmended that the dimensions of the HBM be incorporated into health education programs. This review identified perceived barriers as being the most pow erful dimension of the HBM. The second most powerful dimension was perceived susceptibility. The study concluded that educators should identify and focus on perceived barriers and perceived susceptibility. The Expanded HB M is especially useful in diabetes management where focus on lifestyle behav ior requires long-term changes. People must not only feel the benefits outweigh the cost to perform and sustain diabetes self-care behaviors, but they also must feel competent to make the desired changes in behavior. Numerous studies related to health behavior have examined self-efficacy. Strecher, DeVillis, Becker, and Rosenstock (1986) reviewed 21 studies employing the self-efficacy construc t and found self-efficacy to be a strong predictor of success at changing health behav iors. In a similar review of selfefficacy studies, Lawrance and McLeroy (1986) concluded that self-efficacy was

PAGE 26

13 a strong predictor of behavior. Programs that address these beliefs and help motivate patients to modify their lifestyles, by developing coping and problem-solving skills to overcome barriers to self-care, have been more succ essful at improving glycemic control (Garrard, et al., 1990; Mazzu ca, et al., 1986; Mulhauser, et al., 1987; Peyrot & Rubin, 1988; Peyrot & Rubin, 1994; Rubin, Peyrot, & Saudek, 1991; Rubin, Peyrot, & Saudek, 1993; Wooldridge, Wa llston, Graber, Brown, & Davidson, 1992). The intensive follow-up is hypothes ized to motivate patients to modify their exercise and food intakes by prov iding a better understanding of their glycemic effects as observed by the continuous glucose sensor. The data obtained from the sensor can be used to t each patients how their lifestyle affects their metabolic control, and what they c an do to control blood glucose levels. Numerous studies demonstr ate that health beliefs are positively correlated with diabetes self-care adherence (Al ogna, 1980; Cerkoney & Hart, 1980; Harris & Linn, 1985; Kirscht, 1974; Kirscht, 1976; Maiman & Becker, 1974) and that both health beliefs and adherence to di abetes self-care are correlated with glycemic control (Harris & Linn, 1985; Schafer, Glasgow, McCaul, & Dreher, 1983). Wooldridge, et al. (1992) found that Hgb A1c improved significantly from preto post-education in a subgroup of patients who also improved health beliefs; however, they were unable to a ssociate these improvements with self-

PAGE 27

14 reported adherence or Hgb A1c directly. The question as to whether improvement of health beliefs and attitudes towards diabetes can be achieved by diabetes selfmanagement education, and whether thes e changes can in turn, improve diabetes self-care behaviors and glyce mic control, need to be answered. The proposed study will attempt to answer this question by employing a health belief assessment tool, the Expanded Heal th Belief Model Questionnaire, to assess health beliefs, attitudes, diabetes self-care adherence, and other factors that influence diabetes control before and after an educational intervention. A standard follow-up group will be compared to those in an intensive follow-up group pre-intervention (week1), and at the 24th week post follow-up. Using the patient assessment and t he revised Expanded Health Belief Model Questionnaire (EHBMQ), the diabet es educators during the education and follow-up intervention will attempt to modify beliefs and attitudes that received a poor or low score (<2). To improve low scores on perceived benefits/importance of care, the diabetes educat or will emphasize that diabetes is a serious disease with shortand long-term complicati ons; will describe microvascular and macrovascular complications and their re lationship to chronic high blood glucose levels; and will describe di abetes hyperosmolar coma, symptoms of uncontrolled diabetes and their relationship to high bl ood glucose levels. For those scoring low on perceived benefits of treatmen t, emphasis on consequences of poor control will be discussed. Benefits of preventing complications, increased well-

PAGE 28

15 being and physical health will be stressed. The psycholog ical aspects of being in control of diabetes rather than bei ng controlled by the disease will be emphasized. Participants with a poor score on barriers will be assisted in identifying and dealing with specific ba rriers to self-care management. For patients scoring low on perceived ability the educator will emphasize that all diabetes self-care skills can be learned. Self-confidence can be enhanced by teaching coping skills, and assisting parti cipants in mastering self-management skills. Patients will be assisted in forming their own behavioral goals, and setting up personal contracts. The literature suggests that the response to comprehensive diabetes education is complex and is mediated by a number of variables. Patient characteristics such as type and mode of therapy, co-morbidi ty, duration of disease, ethnicity, educational backg round, socio-economic background, and age are among the numerous factors that ca n all influence response to education efforts (Anderson, Wolf, Burkhart, Corne ll, & Bacon, 1989; Glasgow, Toobert, & Hampson, 1992; Irvine, Saunders, Blank, & Carter, 1990; Jacobson, Adler, Wolfsdorf, Anderson, & Derby, 1990; Wing, Epstein, Norwalk, Scott, & Gooding, 1987). To control the influenc e of these factors a random assignment to either the control or the intervent ion group and the incorporatio n of a repeated measure design will be used to decrease systematic error and individual variance. Several subgroups exist within the di abetes patient population. These

PAGE 29

16 groups are based on treatment type. Treat ments may include diet control, oral diabetes medications, and insulin. Treat ment type can act as a cofounder. The literature suggests that pati ents receiving insulin may have different health beliefs and Hgb A1c values than those treated without insulin (Greene, et al., 1991; Hartwell, et al., 1986; Jacobson, et al., 1990; Mulhauser & Berger, 1993; Peyrot & Ruben, 1988). The study sample will be limited to persons with type 2 diabetes. After blocking for insulin use and applying random ization, the groups tends to equalize with sample size. The groups will be evaluated for any pre-intervention differences that may conf ound the results. In some studies (Glasgow, et al., 1997; Wooldridge, et al., 1992) the expanded HBM has been used to not only i dentify and explain di abetes self-care behavior, but also to address these be liefs and help motivate persons with diabetes to modify their lifestyles. Educational programs focused on helping participants develop coping and problem-solving skill help to overcome barriers to self-care. These studies have prov ided strong evidence that those programs that can modify health beliefs and im prove self-care performance are successful at improving glycemic control. The extent to which ADA diabet es self-management education and subsequent follow-up care modify se lf-care behaviors and beliefs, thereby improving HgbA1c values, needs to be a ssessed. Comparing these variables

PAGE 30

17 preand post-intervention will help dete rmine if the proposed intervention has an effect on psychosocial factors such as attitudes, beliefs, self-care adherence, understanding of diabetes, self-care skills and social support. Does intensive follow-up care make a difference or is standard follow-up just as effective? Standard follow-up care consists of a 12th week evaluation of individualized behavioral and educational goals and objective s and post-HgbA1c determination. Intensive follow-up differs from standard foll ow-up care by use of the continuous sensor placed during their 12-week follow-up visit. The sensor is worn for 72 hours while maintaining their usual daily activities. Physical activities, medications, blood glucose readings and food intake are recorded on a daily log. The subjects return for removal of the sensor and the glucose data are downloaded to the computer. The info rmation gathered during the sensor period is examined and used to help patients make appropriate lifestyle changes. The glycemic effects due to food intake and exercise can be illustrated to the participant, possibly motivating the person to perform certain diabetes self-care behaviors. Strategies are then develope d to help overcome obstacles to diabetes management. The expanded HBM was selected to provide the theoretical framework for this study because the model has been significantly related to health behaviors and enables the practitioner to design interv entions that focus on mutable factors such as beliefs, attitudes, and self-effi cacy that can be influenced by education.

PAGE 31

18 Statement of the Problem Despite the establishment of diabet es education standards more than a decade ago, most diabetes education offer ed to patients has consisted of piecemeal instructions, at best (Clemen t, 1995). Coonrod, Harris, and Betschart (1994) found that in a representative samp le of the U.S. adult population with diabetes, only 35.1% had ever attended a di abetes education class. Even more alarming was the fact that more than 75% of persons with type 2 diabetes had never received formal diabetes education. Similarly, Snyder (1996) reported that 65% of patients receive no diabetes educ ation. Despite the existence of diabetes standards of care recommendat ions for HgbA1c determinations -minimally once or twice a year fo r type 2 patients with stable control and quarterly for those who required changes in diabetes therapy or in poor control -only 40% had HgbA1c determinations (ADA, 2001). Fueled by the escalating cost of heal th care and the high healthcare utilization by persons with diabetes, c ontroversy and debate continue as to whether diabetes education can effectivel y improve glycemic control and improve self-care behaviors. Researchers, heal thcare providers, and the person with diabetes need to arrive at some mutual agreement on what defines “effectiveness.” Barnard (1938) defined e ffectiveness as “the accomplishment of recognized objectives of cooperative effort .” This definiti on is especially applicable here where multiple perspectives need to be considered.

PAGE 32

19 Diabetes education programs must not only be evaluated on whether they effectively lower Hgb A1C levels, but whether they fo ster changes in attitudes, beliefs, and diabetes self-care behaviors. Standards provide a benchmark for quality for DSME based on a combination of best practice and best scientific evidence. Standards of diabetes care are continually updated evolving as new evidence warrants. New technologies o ffer health professionals new tools to evaluate diabetes management. If pers ons receiving intensive follow-up employing these new technological advanc es have better outcomes, important evidence for a new standard of diabetes management will have been identified. The use of home glucose monitori ng has become the mainstay of intensive management despite its limitati ons in detecting true variations in glucose levels. Studies have shown that patients who are willing to monitor their blood glucose levels regularly and work with health professionals have lower HgbA1c values and fewer diabetes-related complications (Peyrot & Rubin, 1994; DCCT, 1993). Despite growing evidence for tighter control, many patients do not receive adequate training or management of their diabetes. To obtain optimal blood glucose control, patients are asked to make major lifestyle changes such as following a meal plan, monitoring thei r blood glucose, taking their medications, and getting regular exercise. Diabetes self-care is demanding and frequently involves tasks that are unpleasant such as fingersticks, food restrictions or insulin injections. Following a meal plan and exerci sing cause the greates t difficulty.

PAGE 33

20 Lack of proper training is often to blame. In one survey, 90% of the patients surveyed with diabetes never received diet instructions. Many patients are told to lose weight and to exercise, but are giv en no instructions as to what to eat or what amount or mode of exercise is best (Ford & Herman, 1995). In addition, few type 2 patients are in structed to use a glucose meter or how to interpret the values they obtai n. The utility of SMBG depends on several factors such as the accuracy, the frequency of testing and ability to use the data to make daily diabetes care decisions Long-term glucose control may not improve unless frequency of testing is incr eased to five times a day. Increasing the frequency of glucose monitoring is view ed by many authorities as unrealistic in light of the U.S. survey that repo rts that even among type 1 patients, only about 20% tested their own blood glucose. Of those type 1 patients participating in self-testing, only 15% tested blood glucos e at least three times a day. Harris et al. (1993) reported in a diabetes self -care survey of 1895 respondents that 67% of patients did not monitor their own blood glucose levels. The most common reasons given for noncompliance were finger soreness, pain, inconvenience and fear of needles. Al though glucose monitoring has improved by requiring less blood, many pati ents still fail to test the recommended frequency. Barriers such as soreness from multiple punctures, lack of ability to use the data obtained, or lack of insuranc e coverage for training and supplies are commonplace.

PAGE 34

21 Many patients are taught the mechani cs of how to use a glucose meter but receive little training in how to use the glucose data. Harris et al. (1993) found that the frequency of home glucose monitoring was related to whether patients had received diabetes self-managem ent education or not. Those who had DSME were three times mo re likely to monitor thei r blood glucose levels at least daily than those who had no formal ed ucation. Diabetes educators rely on SMBG to teach patients daily decision making. The glucose results can be used to reinforce food intake decisions such as limiting carbohydrate portions to lower post prandial blood glucose levels, and to re inforce the beneficial glucose effects. Strategies to overcome the patients’ barriers to monitoring included use of alternate site glucose testing, use of less invasive glucose monitoring sensors, and proper training of the use of data. Further studies are needed to evaluat e whether glucose data obtained from conventional blood glucose m onitoring is adequate for making daily diabetes care decisions and for determining the necessary frequency of testing. Preliminary studies found that using a less invasive method of continuous glucose monitoring with the Minimed s ensor was able to reflect the true variations that occur during ordinary daily life that were missed by conventional SMBG. The sensor data were useful in making decisions regarding the effects of medication, food intake, and exercise. B ode (1999) found in a five-week pilot study with nine type 1 patients, HgbA1c va lues were lowered significantly using

PAGE 35

22 the sensor to adjust therapies. The sensor also may be useful to show patients the effects of their food choices on t heir blood glucose levels. Sensor data coupled with a log that documents food intake, medications and activities can be used to assess glucose trends and provide feedback to the patient. A myth that compromises diabetes ca re is that glucose monitoring and diabetes education are only for persons wit h diabetes that requires insulin. The findings of the UKPDS study (1998) pr ovided strong sup port of the ADA’s position that rigorous treatment of ty pe 2 diabetes is needed as well to lower HgbA1c levels and prevent complications. For every 1% reduction in HgbA1c, microvascular complications were lower ed 35%. Despite the growing evidence that strict glycemic control is important in type 2 patients, the third National Health and Nutrition Examination Survey (NHANES III) showed the majority of patients with type 2 diabetes had HgbA1c levels over 8% (Harris, Eastman, Cowie, Flegal & Eberhardt, 1999). Nati onally, the mean HgbA1c for the diabetes population was 9.2%. The same survey revealed that 59% of patients with diabetes never practiced SMBG. Only 11% to17% practiced SMBG at least once a day. ADA recommends blood glucose testi ng in type 2 patients at least twice a day. Inadequate diabetes care practices by healthcare providers may be partially to blame. In an HMO setting su rvey by Peters, Legorreta, Ossorio, and Davidson (1996), only 44% of the diabetes patients had HgbA1c levels tested at

PAGE 36

23 least twice a year. The ADA (2001) recommends that healthcare providers conduct thorough medical examinations, physical and laborator y examinations including HgbA1c, and formulate a co mplete management plan with the diabetes team and patient. Investigators still have been unsuccessf ul at overcoming the many obstacles of study design and methods that plagued earlier studies. They have found it difficult to separate the effe cts of diabetes educa tion from medical treatment and other demographic, psychos ocial, and clinical factors that influence glycemic control. The major obstacle to performing a long-term, prospective, randomized clinical trial is t he ethical issue of denying education to a control group. No long-term comparis ons can be made between the metabolic outcome of an education group versus a no-education group. When comparing various types of education interventions co mpared to a control there is a smaller treatment effect requiring impractical sa mple sizes. Long term studies have more difficulties distinguishing the e ffectiveness of diabetes education from diabetes treatment and other factors (B loomgarden, et al., 1987; Clement, 1995; Delamater, et al., 1991; Hanefeld, et al., 1991; Kaplan, Hartwell, Wilson, & Wallace, 1987; Korhonen, et al., 1983; Litzel man, et al., 1993; Mazzuca, et al., 1986; Raz, Soskolne, & Stein, 1988; Re ttig, Shrauger, Recker, Gallagher, & Wiltse, 1986; Vanninen, Uusi tupa, Siitonen, Laitinen, & Lansimies, 1992; Vinicor, et al., 1987).

PAGE 37

24 The International Diabetes Cent er in Minneapolis (2001) used a staged diabetes management approach. During t he medical nutritional stage they reported that up to a 1% reduction in HgbA 1c may be expected. In the oral agent stage, a cumulative 2% reduction of HgbA 1c may be expected. The combination oral agent stage or combination oral agent and insulin stage has a potential 2% to 4% reduction in HgbA1c values. The various insulin stages have a 4% or greater potential reducti on in HgbA1c values. Prior to the 1980s, no universal standards for diabetes education were in place. Diabetes education being evaluated in studies varied tremendously in its intensity and composition as well as the type of sa mple population employed. Early studies could only suggest that ce rtain educational interventions might be helpful in some populations, but coul d not generalize to other populations (Clement, 1995). These early studies failed to describe the education interventions adequately, or the basic education being compared. Both the education intervention and the basic educ ation varied from study-to-study, making comparisons difficul t. In some studies, basic education consisted of printed education materials and verbal inst ructions given to patients from their physician’s office (Delamater, et al., 1991; Litzelman, et al., 1993; Mazzuca, et al., 1986). Others defined their control as those receiving minimal education from a dietitian and physician (Korhonen, et al., 1983; Raz, et al., 1988; Rettig, et al., 1986) or no specific education (Hanefeld, et al., 1991).

PAGE 38

25 Purpose of the Study The purpose of the study is to dete rmine if persons with type 2 diabetes participating in ADA diabet es self-management education (DSME) assigned to an intensive follow-up group achieve any si gnificant differences in capillary HgbA1c values over time compared wit h controls receiving standard follow-up care. Intensive follow-up differs from standard follow-up by ut ilizing the MiniMed continuous glucose monitoring system (C GMS) to analyze glucose patterns and responses to exercise, medications, food intake and daily activities. During the DSME and follow-up sessions, educators will help patients ident ify factors that influence their diabetes-related behaviors including health beliefs, self-efficacy, self-care and coping skills. Educators help patients identify diabetes-related problems and behaviors that are constructive or non-constructive in regards to meeting their diabetes goals. The main objective of the st udy is to determine whether a statistically and clinically signifi cant difference exists in the level of glycemic control between a randomly assi gned intensive follow-up group after participating in diabetes self-managem ent education and t he standard follow-up group. The second aim of the study is to determine if diabetes behavior and beliefs change differently following DS ME for the intensive follow-up group compared to the standard follow-up group. To examine the effectiveness of DSME and follow-up on these variables, a preliminary psychosocial assessment, the Expanded Health Belief Model Questionnaire, wil l be used. Comparing these

PAGE 39

26 variables preand post-intervention will help determine if the education and follow-up interventions result in any changes in diabetes-specific attitudes, beliefs, understanding, and adherence to recommended self-care procedures.

PAGE 40

27 Significance of the Problem Mortality from diabetes has declined since the discovery of insulin. However, the growing number of people with the disease, and its chronic and debilitating nature, pose a majo r public health and clinical concern. Diabetes is associated with higher hospitalization rate s, greater length of stay, and greater hospital costs (ADA, 2005,1998b; Ray, et al., 1996). Nationally, diabetes costs the health care system nearly $132 billio n annually. The 1992 estimates put the cost of health care for people with diabetes at approximately $11,157 per person compared to $2,604 for people without diabetes. Today, on average, each hospitalized patient with diabetes costs $25,000 compared to $12,200 for patients without diabetes. Near ly 15% of the total U.S. healthcare expenditures is spent on the 6% of t he population with diabetes. One out of every seven health care dollars is spent on diabetes. Current estimates for both direct and indirect cost exceed $132 billion (A DA, 2005 1998b; Bellenir & Dresser, 1995; Ray, et al., 1993; R ubin, et al., 1994). The current cost associated with diabet es is but a fraction of future expected costs because of the aging popul ation and growing nu mber of residents who have the disease. The Centers fo r Disease Control and Prevention (CDC, 2001) estimates by 2025 that 300 million people worldwide will have diabetes. Many elderly persons with diabetes are dependent on Medicaid or Medicare assistance. Studies have shown t hat persons dependent on Medicaid and

PAGE 41

28 persons lacking adequate insurance coverage have a twoto five-fold increase in hospitalizations for acute complicat ions (Rubin, et al., 1994). In addition to the devastating financial costs, diabetes imposes a burden of illness, discomfort, disability, and premature death. The CDC (2005) estimates that over 200,000 deaths per year can be attributed to diabetes. In 1990, there were 2.8 million diabetes-related hospital admissions, accounting for 24.5 million hospital days (Bellenir & Dre sser, 1995). Many of these admissions were a direct result of knowledge defic its in self-management skills such as taking medications correctly, glucose moni toring, meal-planning, proper foot care, and following sick day guidelines. Previous studies have shown that 50% to 80% of adults with diabetes, regardless of duration of the disease, had severe knowledge deficits about their disease (M iller, et al., 1978; Watkins, Williams, Martin, Hogan, & Anderson, 1997). Preventable acute complications, such as diabetic ketoacidosis, remain predominant causes of death among persons with type 1 diabetes. According to a study by Herman, Teutsch, and Geiss (1987), 52,000 cases of ketoacidosis can be averted a year through major educational interventions aimed at improving glycemic control. The Diabetes Control and Complications Trial (1996) demonstrated that intensive therapy coupled with comprehensive diabetes education resulted in a 76% reduction in retinopathy, a 36% to 56% reduction in nephropathy, and a 60% reduction in diabetic neuropathy.

PAGE 42

29 Despite these promising results, clinic al practice lags in applying these scientific advancements, mainly due to resistance by third-party payers to reimburse for comprehensive diabetes educ ation. A study by Weiner et al. (1995) on Medicare patients with type 2 diabetes found 84% had not received Hgb A1C measurements, despite st andards that recommend these measurements be taken at least twice a year. Some data suggest that the average Hgb A1c for persons with type 2 diabetes is 8.5% (Greenfield, Roger, Mangotich, et al., 1995). Effo rts are underway in some stat es, such as Florida, to enact legislation that requires insurance co mpanies to cover the cost of diabetes care supplies needed for blood glucose m onitoring and education services. Since July 1, 1998, Medicare provides re imbursement for diabetes supplies and education when provided by a certified provider. In the present healthcare climate it is imperativ e that the outcomes of diabetes treatment, new technologies and diabetes education be evaluated .

PAGE 43

30 Hypothesis This prospective study compared the effectiveness of intensive follow-up group compared with standard care followup of participants of ADA diabetes self-management education program on me tabolic control, diabetes health beliefs and behaviors. A randomized pretes t-posttest control group design tested the following null hypothesis: Ho: There is no significant difference in mean glycosylated hemoglobin (HgbA1c) value over time (week 1, week 12, and week 24) for participants of a diabetes self-management educat ion program assigned to intensive follow-up group (CGMS utilized) compared to a standard follow-up control group.

PAGE 44

31 Research Questions This investigation also addr essed the following questions: 1. Is there any significant diffe rence in mean glycosylated hemoglobin values over time (week 1, week 12, and week 24) between those participants of the ADA comp rehensive self-care management training randomly assigned to t he intensive follow-up group and the standard follow-up control group? 2. Are there any significa nt profile differences in mean Health Belief Model subscale scores (perceived barriers, severity, susceptibility, benefits, self-efficacy, cues to action, motivation and structural elements) between the partici pants receiving the ADA comprehensive self-care m anagement training assigned the intensive follow-up and the standard follow-up control group over time? 3. Is the difference between t he week 1 and the week 24 mean perceived barrier scores of t he intensive follow-up group significantly different compared to the standard follow-up control group? 4. Is the difference between t he week 1 and the week 24 mean perceived benefits score of the intensive follow-up group significantly different compared to the standard follow-up control

PAGE 45

32 group? 5. Is the difference between t he week 1 and the week 24 mean perceived self-care ability scores of the intensive follow-up group significantly different compared to the standard follow-up control group? 6. Is the difference between t he week 1 and the week 24 mean perceived severity scores of t he intensive follow-up group significantly different compared to the standard follow-up control group? 7. Is there a significant differenc e between the week 1 and the week 24 mean perceived susceptibility score s of participants receiving the ADA comprehensive self-car e management training randomly assigned to the intensive followup group compared to the standard follow-up control group? 8. Is there a significant differenc e between the week 1 and the week 24 mean perceived cues to acti on scores of the participants receiving the ADA comprehensive self-care management training randomly assigned to the intensiv e follow-up group compared to the standard followup control group? 9. Is there a significant differenc e between the week 1 and the week 24 mean structural elements score s of participants receiving the

PAGE 46

33 ADA comprehensive self-care management training randomly assigned to the intensive followup group compared to the standard follow-up control group? 10. Is there a significant differenc e between the mean diabetes selfcare practice scores of the participants receiving the ADA comprehensive self-care managem ent training randomly assigned to the intensive follow-up group com pared to the standard follow-up control group? This study sought evidence as to w hether persons with diabetes type 2 who participate in an ADA–approved DSME randomly assigned to the intensive follow-up group have any significant differenc es in their HgbA1c values, health beliefs, and diabetes self-care behaviors over time compared to a standard follow-up group. The intensive follow-up group differs from the standard followup group by using a continuous glucose sensor to determine blood glucose patterns and responses to exercise, medi cations, food intake, and other daily activities.

PAGE 47

34 Research Plan The study sample was recruited from the Tampa Bay area, and consisted of English-speaking adults, 18 year of age or older who have been diagnosed with type 2 diabetes. Exclusion criter ia included: having type 1 diabetes; pregnant women; patients with concomitant conditions or who are in poor general health that might impair their ability to par ticipate in self-care activities; patients who had undergone a comprehensive diabet es self-care management training program within the past year; patients with any form of hemoglobinopathy; and patients who had Hgb A1c less than 7.0. Volunteers who signed the informed c onsent were assessed initially by the diabetes care manager and complet ed the Expanded Health Belief Model Questionnaire (EHBMQ) .. Each volunteer’s Hgb A1C was assayed. Those persons who met the eligibi lity criteria were randomly assigned after blocking by treatment (insulin or not) to the intensiv e follow-up group following DSME, or to a standard follow-up control gr oup. Demographic and baseline information such as anthropometrics, blood chemistries, die t, medical, and social histories were obtained by the diabetes case manager. Study participants received ADAapproved DSME taught by a mu lti-discipline team of ce rtified diabetes educators (including a diabetes nurse educator, r egistered dietitian, and exercise physiologist). The ADA-approved curriculum “ Life with Diabetes ” included: complications and management of the di sease; and skills necessary for good

PAGE 48

35 control such as glucose monitoring, meal planning, foot care, sick day management, goal setting, stress m anagement, and exercise. Diabetes educators individualized education plans ba sed on information compiled from the diabetes needs assessm ent and the EHBMQ. HgbA1c values were assessed at week 1, week 12, and week 24 (see Figure 1). Health Belief scores were a ssessed at week 1 and week 24. Records of attrition and adherence were main tained and analyzed. Attrition is operationally defined as those research parti cipants lost from the intervention or control group because they dropped out of t he study, missed pretesting or posttesting, or were absent during the majo r educational sessions. For this study, participants had to attend a minimum of 6 hours of sessions covering the major content areas in part 1 and 2. Strategies were employed to minimi ze attrition; including reminder postcards of appointments, follow-up phone calls, assi stance with scheduling or rescheduling and establishing good rappor t. Subject characteristics and treatment conditions were compared between dropouts and completers to evaluate whether differential a ttrition threatened the internal validity of the study. To minimize selection bias, scheduling, random assignment, and data entry were conducted by a program assistant not dire ctly involved in t he delivery of the education program. The study sought a sample size of 159 participants to achieve adequate

PAGE 49

36 power (.80) and alpha .05. Estimates are based on a table of sample sizes for a two-group repeated measure design (Vonesh & Chinchilli, 1997), for a maximum correlation of .9, and to detect a minimu m standard deviation of .50. A group size of 60 is needed. Due to the time and effort required of participants in this study, an attrition or dropout rate of 35% for the intensive intervention group (additional 21) and 30% for the standard fo llow-up control group (additional 18) was added to the sample size. The primary response variable, Hgb A1C, was measured at the beginning of the trial at the first vi sit, and repeated at week 12 and 24. Initially, the control (standard follow-up) versus the interv ention (intensive follow-up) group was compared for significant differences in the response variable and other baseline characteristics. The fo llowing study population charac teristics were compared: gender, age, education, duration since diabetes diagnosis, physician type, hemoglobin A1c, diabetes treatment, and diabetes health belief scores. The Diabetes Health Belief Model subsca les for perceived barriers, perceived benefits, perceived self-effi cacy, perceived susceptibi lity, perceived severity, cues to action and structural elements were measured at week 1 and week 24. A profile analysis was employed to test par allelism using SAS procedures (1979). A test of parallelism determines whether the profiles are similar or whether a groupby-variable interaction exists. The primary response variable, the mean Hgb A1C, was compared at

PAGE 50

37 baseline, week 12, and week 24 for t he two groups using ANOVA for repeated measures. A profile analysis (2 bet ween one within ANOVA) was used to compare the groups week 1 and week 24 on the eight major constructs of the HBM (perceived barriers, benefits, severi ty, susceptibility, cues to action, structural elements, motiva tion and self-efficacy scores.

PAGE 51

38 Figure 1. Study Design

PAGE 52

39 Assumptions For the purpose of this study it was assumed that: Due to the direct relationship bet ween blood glucose levels and the risks of diabetes-related complications, techniques that improve glycemic control over time reduce the ri sks of diabetes complications. The benefits of improved glycemic c ontrol found among persons with type 1 and type 2 diabetes would also apply to other types of diabetes as well. Achieving and maintaining good met abolic control can reduce diabetes mortality and morbidity. Comprehensive diabetes self-managem ent programs that are accredited by the ADA must meet similar struct ure, process, and outcome standards. The benefits of improved glycemic control such as healthier, more productive lives with fewer complications offset the additional expense of comprehensive diabetes programs. Persons participating in the di abetes self-management training program are motivated to control their diabetes and value their health.

PAGE 53

40 Delimitations The study is delimit ed to the following: 1. Persons with type 2 diabetes refe rred to the trial by Tampa-area physicians. 2. Persons with type 2 diabetes 18 years of age that give informed consent to participate in the study. 3. Capillary Hgb A1C determinations. 4. English-speaking adul t volunteers who are ph ysically and mentally able to perform self-care activities. 5. Persons with diabetes with Hgb A1c values between 7.0 and 13.0.

PAGE 54

41 Limitations This study is limited to the following: 1. ADA-recognized diabetes self-managem ent education programs that meet national standards might be differ ent from other diabetes selfmanagement education programs. 2. A sample made up of healthy volunt eers referred to the Diabetes Care Institute at University Community Ho spital Clinic by their physicians may be different from other diabetes populations. 3. Because the sample is drawn from the Tampa Bay area, samples from other areas of Florida or the United States may be different. 4. This is a short-term prospective randomized clinical trial and trials conducted over other time frames may yield different results. 5. The revised diabetes Expanded Health Belief Model questionnaire used in this study is a self-report instrument; and other types of measures may yield different results. 6. An empowerment education approach is used in this study group (the compliance approach and others may yield different results). 7. Over a 24-week period subjec ts may learn diabetes management techniques on their own.

PAGE 55

42 Definitions of Terms Adherence — Extent to which a person’s be havior (in terms of medications, diets or life-style changes) coincides wit h medical adherence (Haynes, Taylor, & Sackett, 1979). Blood glucose — The main sugar that the b ody makes from three elements of food: proteins, fats, and carbohydrate s, but mostly from carbohydrates. Glucose is the major source of energy for living cells and is carried to each cell through the bloodstream. However, t he cells cannot use glucose without the help of insulin (Bellenir & Dresser, 1995). Blood Glucose Meter — A machine that tests how much glucose is in the blood. A specially coated stri p containing a fresh sample of blood is inserted in a machine, which then calculates the correct level of glucose in the blood sample and shows the result in a digital di splay (Bellenir & Dresser, 1995). Diabetes Mellitus — A disease that occurs when the body is not able to use sugar as it should. The body needs sugar for energy for daily ac tivities. It gets sugar when it changes food into glucose (a form of sugar). Diabetes occurs when the body tries to use sugar in the blood for energy — but cannot because the pancreas is not able to make enough of a hormone called insulin, or because the body cannot use the insulin it does hav e. The beta cells in areas of the pancreas called the Islets of Langerhans usually make insulin (Bellenir & Dresser, 1995).

PAGE 56

43 Type 1 Diabetes — This type of diabetes is characterized by beta cell destruction, usually leading to absolute in sulin deficiency. Type 1 diabetes may have two forms — idiopathic referring to forms of the disease with no know etiology and immune-mediated resulti ng from a cellular mediated autoimmune destruction of the beta cells of the pancreas (ADA, 1998a). Type 2 Diabetes — This type of diabetes is defined as those individuals who have insulin resistance and usually re lative insulin deficiency (1998a). Gestational Diabetes — Carbohydrate intolerance of variable severity with onset of first recognition during t he present pregnancy (Freinkel, 1985). Diabetes Self-Management Traini ng or Self-Management Education — Previously termed diabetes patient educati on, this term refers to the process whereby individuals learn to ma nage their diabetes (ADA, 1996). Diabetes Educator — Defined as any health professional who has mastered the core knowledge and skills in biological and social sciences, communication, counseling, and educat ion and has a specific amount of experience in the care of per sons with diabetes (AADE, 1992). Certified Diabetes Educator (CDE) — Certification in diabetes education designates the attainment of knowledge beyond basic core knowledge and skills. A CDE has at least 2000 hours of direct diabetes teaching experience and has successfully completed an examination that verifies contemporary diabetes

PAGE 57

44 knowledge across multiple professiona l disciplines (American Dietetic Association, 1995). Glucose Tolerance Test — A test to see if a person has diabetes. The test is given in a lab or doctor's offi ce in the morning before the person has eaten. The person first gives a blood sa mple. Then the person drinks a liquid that has glucose in it. At specified intervals, the person gives a sample of blood to see how the body deals with the glucose in the blood over time (Bellenir & Dresser, 1995). Hemoglobin A 1c (Hgb A 1C ) or Glycated Hemoglobin — A form of hemoglobin that reflects the average blood glucose concentration over the past three months (ADA, 1993). Health Motivation — Measures concern for one’s own health (Becker, Brachman, & Kirscht, 1974). Incidence — How often a disease occurs; the number of new cases of a disease among a certain group of people for a certain period of time (Bellenir & Dresser, 1995). Insulin — A hormone that helps the body use glucose (sugar) for energy. The beta cells of the pancreas (in the areas called the Islet of Langerhans) make the insulin. When the body cannot make enough insulin on its own, a person with diabetes must inject insulin from other sources, i.e., beef, pork, human

PAGE 58

45 insulin (recombinant DNA origin), or hum an insulin (pork-derived, semi-synthetic) (Bellenir & Dresser, 1995). Microvascular Complications — Means damage to small blood vessels generally affecting three organ systems: the eye (retinopathy), the kidneys (nephropathy), and the nerves (neuropathy) (ADA, 1993). Macrovascular Complications — Means damage due to atherosclerosis of large blood vessels resulting in reduced bl ood flow to tissues such as the heart, brain, and lower extremities (ADA, 1993). Cues to Action — Internal (symptoms) and external factors (health messages, lab reports) that lead or motiva te a person to seek medical care or follow positive health practices (Rosenstock, 1994). Perceived Barriers — Measures the perceived psychological and situational obstacles that would have to be overcome in order to follow recommended actions (Janz & Becker, 1985). Perceived Benefits — Perceptions regarding the effectiveness of the treatment in reducing the th reat, or measures the per son’s faith in curative powers of their physician and the reco mmended actions (Rosenstock, 1974). Perceived Severity — Perception of consequences, and how much one perceives the illness will interfere wit h their lives (Rosenstock, 1974).

PAGE 59

46 Perceived Susceptibility — Perceptions of thr eat of the illness, its complications, assessment of future risks associated with the disease (Rosenstock, 1974). Perceived Self-efficacy — Measures a person’s own assessment of their ability to perform certain tasks or engage in specific actions (Bandura, 1977a; 1977b). Structural Elements — Modifying factors such as locus of control, understanding of the dis ease and the recommended actions, amount and quality of social support or peer influence, and prior experience or contact with the disease that may affect a person’ s perceptions (Rosenstock, 1974). Prevalence Rate — Measures the number of people in a population who have the disease at a given poin t in time (Mausner & Bahn, 1974).

PAGE 60

47 Chapter Two: Review of the Literature This chapter is divided into nine sect ions: (a) an overview of diabetes, (b) etiology and classifications, (c) epidemio logy, (d) diabetes related complications, (e) the Expanded Health Belief Model, (f) fa ctors influencing metabolic control, (g) the diabetes-specific health belief in struments, (h) diabetes self-management education, (j) continuous glucose monitoring and (i) future implications.

PAGE 61

48 Overview of Diabetes Diabetes mellitus is a chronic metabo lic disorder in which the body is unable to produce or properly utilize insu lin, a hormone that facilitates the conversion of sugar, starches, and other s ubstrates into energy. The diagnosis of diabetes is made when: 1) there is an unequivocal elevation of random plasma glucose of greater than or equal to 200 mg/dl (11.1 mmol/l), and classic symptoms of diabetes are pr esent such as excessive thirst, urination, hunger and weight loss; 2) fasting plasma glucose (F PG) is greater or equal to 126 mg/dl (7.0 mml/l); and 3) FPG is greater than or equal to 200 mg/dl during an oral two-hour plasma glucose (PG) toleranc e test (Diabetes Care, 1998a). Etiology and Classification Although the etiology of diabetes is still unknown, it is thought to be more heterogeneous in nature wit h both genetic and environm ent factors possibly involved. Factors such as family histor y, race, age, viruses, stress, and certain drugs or toxins are associated with it s development. In the 1970s, researchers found that patients with newly diagnosed insulin-d ependent diabetes had selfdirected antibodies aimed at destroying t heir own beta cells and insulin. During the 1980s, additional diabetes-related antib odies were discovered in patients prior to clinical diagnoses of the diseas e and thus act as a marker for subsequent diabetes. Some investigators feel that some type of infection or exposure to certain viruses may trigger the auto-i mmune process in genetically susceptible

PAGE 62

49 individuals. By introducing a foreign prot ein that resembles a beta-cell protein, the immune system is fooled into attacking its own beta cells. As more became known about the heterogeneity of diabetes, the old classification of diabetes based on age of onset was replaced by revised classifications developed by the Ex pert Committee on the Diagnosis and Classification of Diabetes Mellitus (Ex pert Committee, 1998a). Juvenile-onset diabetes or insulin-dependent diabetes mellitus was renamed type1 diabetes, and maturity-onset diabetes or non-insu lin-dependent diabetes mellitus was replaced with type 2 diabetes. Separat e categories were established for impaired glucose intolerance and gestational diabetes. Type 1 diabetes accounts for approx imately 10% of the total diabetic population. It is considered an autoimm une disease where inadequate insulin is secreted requiring daily injections of insu lin to stay alive. Patients with Type 1 diabetes are highly prone to ketoacidos is and other classic symptoms of polydipsia, excess urinati on, constant hunger with an intense craving for sweets, fatigue, blurred vision, and rapid weight loss. Type 1 diabetes generally occurs in childhood and is only second to cancer as the most chronic disorder affecting U.S. children (ADA, 1996; Bellen ir & Dresser, 1995; CDC, 1997). The most common form of diabetes is called Type 2 and accounts for over 90% of the cases of diabetes It generally occurs in adults over forty and is especially prevalent in peopl e who are sedentary and obese (Bellenir & Dresser,

PAGE 63

50 1995). Nearly 80% of the people diagnosed with Type 2 diabetes are overweight (ADA, 1996). The risk of developing diabet es also increases with age. Type 2 diabetes is more prevalent among Native American, Hispanic American and African-American populations. More fe males are afflicted by type 2 diabetes than males. Also, women who have had gest ational diabetes are at high risk of acquiring type 2 diabetes. Gestational diabet es is defined as diabetes that is discovered during pregnancy. Secondary diabetes is a less common form of the disease resulting from another condition or medication (Bellenir & Dresser, 1995). Epidemiology The prevalence of diabetes in the U. S. adult population is estimated at 20.8 million, with approximately half being unaw are of their disease (CDC, 2005). Insulin dependent diabetes mellitus (IDDM) or type 1 diabetes occurs in approximately 10% of all peo ple with diabetes. Alt hough the risk for developing Type 1 diabetes is higher in individuals wi th a genetic disposition, the onset of Type 1 diabetes generally results when env ironmental factors such as viruses, toxic substances, or intensive stress tr iggers a progressive destruction of the beta-cells that produce insu lin. Prevalence is highest among Caucasians and among young people under 20 years of age. Type 2 diabetes accounts for nearly 90% of all cases of diabetes in the United States. Generally due to the insipi d nature of the disease, many persons

PAGE 64

51 with type 2 diabetes are likely to have had the disease for several years by the time it is first diagnosed. Persons w ho are overweight or obese have a higher risk of acquiring type 2 diabetes. Diabetes Related Complications Diabetes is a serious disease associated with severe neurological, cardiovascular, ocular, and renal complic ations. It is the leading cause of blindness, amputations, and kidney disease in the U.S. (CDC, 2005). Recent studies have reported delay of onset and slowing of progression of diabetesrelated morbidity (DCCT, 1993; Fishbein & Palumbo, 1995; Ohkubo, et al., 1995; UKPDS, 1997). The most comprehensive diabetes study ever conducted, the Diabetes Control and Complications Tr ial (DCCT, 1993), demonstrated that intensive management of diabetes that stri ved to keep blood glucose levels as close to normal as possible successfully slowed the onset and progression of early vascular complications including eye, kidney, and nerve disease. The DCCT trial demonstrated that intensive therapy compared to conventional therapy resulted in a 76% reduction in retinopathy, a 36-56% reduction in nephropathy, and a 60% reduction in diabetic neuropathy. Although the DCCT was conducted among persons with type 1 diabetes, results are thought to benefit those with type 2 diabetes as well. Recently, studies (Ohkubo, et al., 1995; UKPDS, 1997) involving persons with type 2 diabetes have found that intensive control coupled with

PAGE 65

52 comprehensive diabetes self-management training successfully improved metabolic control. In another clinical tr ial aimed at the preventing or delaying the onset of diabetes, researchers found that preventing obesity could lower the risk of developing type 2 diabetes by nearly 50% (ADA, 1996a). Unfortunately, clinical practice lags behind scientific advancements mainly due to the resistance by third-party payers to reimburse for many prevention and educational services. Prevention of acute and chr onic complications relies on diabetes education that improves understanding and helps participa nts make appropriate behavioral changes to improve their metabolic control. Providing education that only sets out to improve the diabetes knowledge of the participant has had disappointing results (Bloomgarden, et al., 1987; Beeney & Dunn, 1990). For programs to effect desirable outcomes such as reducti ons in sick days, hospitalizations and length of stays, emergency visits, ac ute or chronic complications, and death rates, more comprehensive programs that employ a multidisciplinary team are essential. Programs that target behavioral changes improve skills through practice, and assist participants in over coming barriers to self-care have been proven more successful (Glasgow & Osteen, 1992). The objective of diabetes management is to keep glycemic levels in a range associated with few diabetes complicat ions. In the Kumamoto University Study (Ohkubo, et al., 1995), the glycemic threshold to prevent the onset and progression of diabetic micro-angiopathy was a Hgb A1c level less than 6.5% (4-

PAGE 66

53 6% being a normal Hgb. A1c). The DCCT (1993) found a Hgb A1c level of 7% or lower was associated with fewer micr o-vascular complications. A mean improvement of Hgb A1c of 1% to 2% achieved in the first 10 years of overt diabetes could reduce micr o-complications 30-70%. The American Diabetes Association (1997a) recommends that Hgb A1c should be maintained below 7% to lower the risk of diabetes complicati ons. Corrective action needs to be taken for Hgb A1c levels exceeding 8%. S urveys by Klein ( 1995 ) and Mazze, Etzurlla, & Strock ( 1994) report that persons diagnosed with diabetes have an average Hgb A1c level of 8.5%. Over 50% of persons with type 2 diabetes have Hgb A1c above 8%. A 1995 survey of medical patients with type 2 diabetes f ound that 84% of their 97,388 sample had not received a Hgb A1c measurement during the st udy period (July 1, 1990-June 30, 1991). Some researchers feel that persons with type 2 diabetes receive suboptimal medical care and inadequate diabetes education because health providers consider type 2 diabetes, especiall y those not requiring insulin, to have a mild form of the disease (Teza, Davis, & Hass, 1988).

PAGE 67

54 The Expanded Health Belief Model In search of practical solutions to pr oblems relating to the failure of people to accept preventive health measures and comply with medical regimens, the Health Belief Model (HBM) was developed to help explain why individuals did or did not engage in a wide array of health-re lated actions. Since that time, the HBM has generated a prolific amount of evidence to support its value in explaining why some people engage in healt h-related actions and others do not. The HBM has shown a degree of predictiv e value for preventive and compliance behaviors (Becker, 1976; Janz & Becker, 1984; Harrison, et al., 1992). HBM investigations in general have found that i ndividuals will usually take action to prevent, screen for, or to comply with m edical regimens if they feel they are susceptible to a condition that poses a se rious threat to them They are more likely to follow certain health recommendations if they believe that the treatment will be beneficial by reduc ing the perceived threat at an acceptable cost. Due to the model’s usefulness in expl aining and predicting acceptance of health and medical care recommendations the model has been utilized in a variety of setting and situations. T he HBM is the most frequently cited and applied social-psychological model used in Behavioral research (Wallston & Wallston, 1984). Originally the HBM wa s designed to explain an individual’s willingness to participate in preventive health practices such as immunizations and screening for diseases like hypertens ion and cancer. More recently the

PAGE 68

55 HBM has been extended to describe co mpliance behavior during acute and chronic illnesses as well as to predict utilization of health services. According to the original Health Be lief Model, behavior was explained by a combination of these four components: perceived susceptibility, perceived seriousness, perceived benefits, and perce ived barriers or costs (Rosenstock, 1974). “ Perceived susceptibility” refers to the individua l’s perception of his or her vulnerability to a condi tion or health threat. “Perceived severity” refers to the perception of the seri ousness of the threat; this may include the person’s evaluation of the consequences both clinical and social. “Perceived benefit” refers to the person’s perception of the ef fectiveness of the action in reducing the threat by reducing either t he severity of the disease or the likelihood of contacting the disease. “Perceived barriers or costs” refer to the person’s perceptions as to the various costs associated with fo llowing the recommended action. Costs or barriers can be physical, emotional and socioeconomic, such as fear of side effects of the treatment, discomfort and pain stemming from the action or condition, complexity or difficulty of the regimen, high financial outlay and inconvenience (Janz & Becker, 1984). Rosenstock (1974) felt that besides t he combined effects of susceptibility, severity, benefits and costs, a stimulus or cue to action was needed to trigger the decision-making process. Cues to acti on may be internal (e.g. symptoms) or external (e.g. mass media communications, reminder postcards, posters). In

PAGE 69

56 addition, there are other modifying vari ables that may influence one’s health beliefs or decision to act. Demographics (e.g., age, sex, race, ethnicity), sociopsychological (e.g., social class, peer in fluence), and structur al factors (e.g., knowledge about the disease, prior contact or experienc e with the disease) may affect a person’s perceptions and thus indirectly influence health-related behavior. Rosenstock posited that fact ors such as educational attainment influence the patient’s k nowledge and understanding of t he disease or treatment, and as a result affects the patient’s complia nce. Complexity of the treatment and the patient’s interpersonal relationships with the health care staff, their own families, or significant others may also have a major impact on the patient’s health beliefs and willingness to comply with recommended health actions. In 1974, Becker, Drachman, and Kirsch t (1974) added the construct health motivation to the HBM’s conceptual framew ork. Health motivation was defined as an individual’s degree of interest in or concern about health in general. Becker, et al. theorized that a desire to attain or maintain health and to avoid illness resulted in a willingness to co mply with health recommendations. In recent studies (Friedman, Nelson, Webb, Hoffman, & Baer, 1994; Hallal, 1982; Kaplan, Ries, Prewitt, & Eakin, 1994; Toshima, Kaplan, & Ries, 1990; Trotta, 1980) researchers have incorporated in to the HBM self-efficacy theory to improve its usefulness in prediction and overall utilit y. Self-efficacy theory (Bandura, 1977; Bandura, 1986; Bandura, 1992) focuses on one’s perception

PAGE 70

57 that he or she is capable of adopt ing the recommended behavior. Bandura (1977) stresses that self-efficacy is essential for behavioral change and points out that people who regard outcomes as personally determined but lack the skills to take action would experience low self-efficacy and view the recommended action with a sense of futility. The Health Belief Model has been em ployed extensively to explain compliance with following prescribed diabet es regimens (Alogna, 1980; Becker, 1976; Cerkoney & Hart, 1980; Harris & Linn, 1985; Harris, Skyler, & Linn, 1982; Pham, Forton, & Thibaudeau, 1996; Schlenk & Hart, 1984; Wooldridge, et al., 1992). Kirscht and Rosenstock (1977) obser ved that in illness behavior the perceived severity of the symptoms is the most salient factor for seeking care and complying with medical recommendations. This finding was supported by the subsequent work of Cauffman, Peters on, and Emrick (1967) who found that immediate demands from acute symptoms may override ordinarily observed barriers to care. With mil der cases of illness, the decision to act would be more affected by barriers such as cost, time constraints, or other competing problems (Bice & White, 1969; Richardson, 1970). In examining diet compliance am ong obese patients with diabetes, Alogna (1980) focused on the HBM dimension of perceived severity and found a significant difference in this vari able between compliant and noncompliant subjects. In their study among persons with type 2 diabetes, Cerkoney and Hart

PAGE 71

58 (1980) demonstrated that there was a pos itive association for all four HBM variables and a significant correlation was attained for perceived severity and compliance. Other investigators (Becke r, et al., 1977; Harris, et al., 1982) examining the use of the HBM to expl ain compliance behavior of persons with diabetes have reached similar conclusions. The Health Belief Model is useful for examining beliefs that are likely to affect compliance. A scale measuring heal th beliefs should look at perceptions of susceptibility; do they feel that thei r diabetes can result in complications if poorly controlled? Do they anticipate an early cure? Lo oking at costs/benefits or barriers can help determine if they feel t he regimen is too complex, interferes with their lifestyle too much, or whether t hey believe that following the regimen is effective at improving their cont rol of their health problems. Dunbar and Stunkard (1979) view t he regimen as the single most important determinant of pat ient adherence. In general, compliance is expected to improve with decreases in the regimen’ s complexity, and barriers such as cost, convenience, and incompatibilit y with current lifestyle. St rategies of the self-care training intervention should focus on simplif ying the diabetic regimen. Helping patient’s select easy behavioral tasks or recommending they tackle behaviors one at a time may foster self-confidence. Health professionals also need to reinforce whatever compliance is achiev ed, and proceed in a step-wise fashion tailoring the regimen to the special needs of each patient.

PAGE 72

59 The HBM has been extensively utilized to understand and predict diabetic self-care behavior (Kavanagh, 1993; Kingery & Glasgow, 1989; McCaul, Glasgow, & Schafer, 1987; Schafer, Gl asgow, McCaul, & Dreher, 1983). An early review of studies employing the HBM, Janz and Becker (1984), concluded perceived barriers as being the most pow erful HBM dimension followed by perceived susceptibility. In most chronic debilitating i llnesses such as diabetes or cancer, most persons perceive the disease as serious. To make lifestyle changes, the person with diabetes must feel the benefits of lifestyle changes outweigh the cost to perform and sustain them. Persons must feel competent that they can make the desired changes in behavior. Numerous studies suggest that self-effi cacy is a strong predi ctor of behavior change (Jeffery, 1984; Kavanagh, Gooley, & Wilson, 1993; Kingery & Glasgow, 1989; Lawrance, 1986; Littlefield, 1992; Strecher, et al., 1986). According to Bandura’s (1977) self -efficacy concept, behavior changes are a result of both efficacy expectati ons (beliefs that they can perform a particular task) and outcome expectations ( beliefs that by performing a certain task certain results will occur). Self-efficacy in adolescents with Type 1 diabetes was investigated by Grossman, et al. ( 1987). They found a significant correlation between blood glucose values and beliefs they could do the various selfmanagement task. Other studi es (Jeffery, et al., 1984; Kingery & Glasgow, 1989; Littlefield, et al., 1992; McCaul, Glas gow, & Schafer, 1987) on diabetes and self-

PAGE 73

60 efficacy, especially among children and adolescents, have found that high degree of self-efficacy is correlated with better self-care behaviors. McCaul, et al. (1987) found self-efficacy was the best pr edictor of compliance to the diabetic regimen in both adolescents and adults with diabetes. Kingery and Glasgow (1989) found self-efficacy and outcome expectancies moderately strong predictors of adherence to exercise plans but found self-efficacy a slightly weaker predictor of dietary and glucose testing in person with type 2 diabetes. Littlefield, et al. (1992) found that among adolescent s with diabetes, self-efficacy was the strongest predictor of adherence. In the Health Belief Model, the person makes a subjective judgment about the potential consequences or outcomes by weighing the benefits versus the costs. Both attitudes and beliefs are mu table, whereas other variables such as personal attributes (personality) or so cio-demographic characte ristics (age, sex, marital status) cannot be changed. The HBM is useful in explaining behavior that is in the person’s direct control. The concerns, attitudes, or beliefs are obtained from interviews and questionnaires from me mbers of the target group. Perceived barriers are the most influential determi nant of behavior. By determining the barriers most salient to the participants, interventions can be planned to focus on overcoming these issues, especially wh en program funding or resources are limited.

PAGE 74

61 Factors Affecting Metabolic Control The literature suggests t hat metabolic control is quite complex and is mediated by a number of variables. Patient and clinical characteristics (e.g. age, severity of their disease, type and m ode of therapy, co-morb idity, duration of disease, and ethnicity) can all influ ence glycemic control (Anderson, Wolf, Burkhart, Cornell, & Bacon, 1989; Dunn, 1990; Glasgow & Osteen, 1992; Glasgow, Toobert, & Hampson, 1992; Irvi ne, Saunders, Blank, & Carter, 1990; Jacobson, Adler, Wolfsdorf, Anderson, & Derby, 1990; Wing, Ep stein, Norwalk, Scott, & Gooding, 1987; W ooldridge, et al., 1992). Wooldridge, et al. (1992) proposed a t heoretical framework incorporating the expanded Health Belief Model to illustrate the relationships between the motivation to control diabetes, the self -management behaviors required to control diabetes, and metabolic control (see Figure 2). Numerous factors influence the person’s motivation to control diabetes in cluding health beliefs, personal values and priorities, life experiences, psycholog ical variables, knowledge of diabetes, and past medical experiences. Factors that determine whether a person will perform self-care management behaviors are influenced by the person’s motivation to control diabetes, financial re sources, emotional support, complexity of the regimen, disruption of one’s lif estyle, self-management skills, perceived barriers, locus of control, and cues to ac tion. Metabolic control is determined the persons’ diabetes self-care, medical interventions, severity of diabetes, and other

PAGE 75

62 unknown variables. Non-behavioral factors such as a person’s age, duration of disease, and other medical conditions a ffect metabolic control but are basically out of the direct control of the person with diabetes. The Health Belief Model has been used to help understand attitudes and beliefs that are related to diabetes compliance and metabolic control (Alogna, 1980; Becker, Haefner & Maiman, 1987; Ce rkoney & Hart, 1980; Harris, et al., 1987; Janz & Becker, 1984; Pham, Forton, & Thibaudeau, 1996) Studies have shown that health beliefs are positivel y correlated with compliance (Alogna, 1980; Kirscht, 1974) and both health beliefs and compliance are correlated with metabolic control (Harris & Linn, 1985). Investigators (Anderson, Funnell, et al., 1989; Fitzgerald, 1997) suggest that futu re research should identify beliefs, attitudes and behaviors that influence the ability of persons with diabetes to successfully carry out diabetes self-care. Educational efforts need to focus on developing effective strategies to help individuals overcome barriers and modify health beliefs and attitudes non-conductive to positive diabetes self-care behaviors.

PAGE 76

63 Figure 2. Theoretical Framework: Factors Infl uencing Metabolic Control of Diabetes

PAGE 77

64 Educational efforts need to focus on devel oping effective strategies to help individuals overcome barriers and m odify health beliefs and attitudes nonconductive to positive diabetes self-care behaviors. Health beliefs and health behaviors are influenced by ethnic, cultural, and socioeconomic factors (Friedman, 1990). Self-care plans reco mmended by health professionals need to accommodate these ethnic and cultural variations. The current research suggests that diabetes education effectiveness should focus on measuring mutable variab les such as knowledge of diabetes, self-management skills, health beliefs and attitudes, and program components to foster better glycemic control (Beeney & Dunn, 1990; Glasgow, et al., 1997; Glasgow & Osteen, 1992; W ooldridge, et al., 1992). Glasgow and Osteen (1992) illust rated suggested various psychobehavioral variables that may influence gl ycemic control either directly or indirectly through their influence on self-c are behaviors. The model in Figure 3 (revised from Glasgow & and Osteen, 1992) illustrates possible variables that affect glycemic control and those that are amenable to change by diabetes selfmanagement training. Most frequently ci ted barriers were related to dietary adherence, followed by exercise and gl ucose testing. Of the three HBM psychosocial constructs studied (barri ers, treatment effectiveness, and seriousness), treatment effectiveness wa s the strongest predictor of selfmanagement.

PAGE 78

65 Culminating evidence exists in the literature that comprehensive diabetes education can effectively reduce the mo rbidity and mortality associated with diabetes and its complications by improv ing glycemic control (Coonrod, et al., 1986; DCCT, 1993; Mazzuca, et al., 1986; Ohk ubo, et al., 1995; Pirart, 1978; UKPDS, 1995). Educationa l programs using a multidisciplinary team approach and innovative strategies to develop se lf-care skills and address psychosocial issues tend to have more positive longterm effects on patient 's well-being than traditional didactic approaches (Beeney & D unn, 1990; Brown, 1988; Kaplan, et al., 1987; Mazzuca, et al., 1986; Wooldridge, et al., 1992). Few clinical studies have repor ted the impact of diabetes selfmanagement training on intervening variable s of health beliefs, self-care ability, attitudes, and self-care behaviors whic h examining the effects on metabolic control (Anderson, 1986; W ooldridge, et al., 1992). Wooldridge, et al., (1992) explored in their study whether health beliefs and attitudes could be changed by diabetes education. They f ound that the health beliefs that were not conducive to performing diabetes self-care coul d be positively changed through a comprehensive diabetes self-management tr aining program. Their study showed significant improvements in Hgb A1c levels.

PAGE 79

66 Diabetes-Specific Health Belief Instruments The Health Belief Model (HBM) has undergone extensive conceptual refining and empirical testing (Becke r, 1974; Becker, Drachman, & Kirscht, 1974a; Becker, Maiman, Kirscht, & Haefner 1977; Janz & Becker, 1984; Kasl & Cobb, 1966; Maiman & Becker, 1974; Rose nstock, 1966; Strecher, et al., 1986). The re-conceptualization of the model adding the health motivation and selfefficacy constructs improved the model’s ab ility to explain and predict behavior. The Expanded Health Belief Model is parti cularly useful in examining whether perceptions and situations that are limiting self-care behavior can be changed through intervention. A limitation of the Exp anded Health Belief Model has been the absence of a reliable and valid diabetes-specific inst rument that measures all the major dimensions of the model Despite the development of numerous scales and questionnaires to measure the HBM, no consensus has been reached as ways to standardize measurement of the Expanded HBM variables.

PAGE 80

67 Figure 3. Model of Diabetes Education Environmental and Social Concept > Support > Social & Personal Factors Age, race, sex, education Program Characteristics > Complexity > Teaching method > Structural & process > Evaluation standards > Provider relations Process and Mediative Variables > Knowledge of diabetes > Attitudes > Health Beliefs > Self-efficacy/skills > Intentions > Motivation Diabetes-Mgt Behaviors > Diet Adherence > Self-care Adherence Health Outcomes > Hospitalizations > Quality of Life > Blood Pressure Metabolic & Physiologic Changes > Hbg A1c > % Ideal Body Weight or Body Mass Index Patient Characteristics > Comorbidity > Social Support Diabetes-Mgt Behaviors > Diet Adherence > Self-care Adherence Note: Modified from Glasgow & Osteen’s Model of Diabetes Education, 1992

PAGE 81

68 Early attempts to measure diabetes health beliefs suggested a relationship between the major HBM constr ucts of severity, susceptibility, benefits, and barriers to adherence to the diabetes regimen and metabolic control (Alogna, 1980; Cerkoney & Har t, 1980; Harris & Linn, 1985; Harris, Skyler, Linn, Pollack, & Teuskburg, 1982; Hurley, 1992; Lewis, Jennings, Ward, & Bradley, 1990). Conclusions that could be drawn from these earlier studies are inconclusive due to use of unreliable and invalid instruments, limited samples, and use of global rather than diabetes-specific measures. The Expanded HBM, although conceptually sound, has limited practical utility with diabetes due to the absence of a reliable and valid diabetes-specific instrument that measures all the major dimensions of the model. Although several diabetes health belief instrum ents have been developed based on the original HBM, none of the scales address all dimensions of the EHBM (Becker & Janz, 1985; Bradley, Lewis, Jennings, & Sandifer, 1987; Bradley, Lewis, Jennings, & Ward, 1990; Brownlee-Duffick, Peterson, Simonds, Kilo, & Hoetle, 1987; Davis, Hess, Harrison, & Hiss, 1987; Gwen, Gwen, Gallin, & Condon, 1983; Hurley, 1992). The Diabetes Health Belief Scale (DHBS) developed by Harris, Linn, Skyler, and Sandler comes closest to measuring the Expanded HBM by measuring the health motivation, structural elements, cues to action, as well as

PAGE 82

69 ways to measure the four original construc ts of perceived severity, susceptibility, benefits, and barriers or costs. Three experts working in the diabetes field, two nurse researchers, and thirteen indivi duals with type 2 diabetes evaluated the original version of the scale for face validity. The relia bility of the original scale used to measure health beliefs was not reported (Harris & Linn, 1985). Initially, 71 items were generated from HBM theoretical research and previously used scales (Becker, Haefner, Kasl, et al., 1977; Becker & Janz, 1985; Becker, Maiman, & Kirscht, 1977; Maiman, Becker, & Kirscht, 1977; Rosenstock, 1966; Rosenstock, 1978). In subsequent research, the DHBS was rigorously tested for reliability and validity (Harris, et al., 1987). Test-retest reliability was employed to select items. The developers retained items that had a .35 test-retest correlation (intra-class R’s) and items with moderate variance ( .40 standard deviations). Of the original 71 items, 38 items were retained. Data was collect ed from a sample of 280 men at Miami Veterans Administration Medical Center and factor analysis was used to verify the dimensions the scale was designed to measure. Seven factors were the obtained from the 38 items. The factor susceptibility consisting of eight items accounted for 8.6% with 0.83 test-retest reliability. The treatment beneficial factor consisting of eight items accounted for 7.3% of the variance with 0.51 testretest reliability. The seve rity factor included five items accounting for 6.7% of the variance with 0.74 test-rete st reliability. The struct ural element factor, which

PAGE 83

70 measured patient understandi ng and family support, consisted of five items accounting for 5.5% of the variance. Test-retest reliability was 0.81. The psychological barrier factor consisting of four items accounting for 5.5 variance with 0.78 test-retest reliabi lity. This factor pertained to the perceived psychological obstacles or costs involv ed in following the prescribed diabetes treatment. The health motivation factor consisting of four items accounted for 4.3% of the variance with 0. 49 test-retest reliability. Factor scores were intercorrelated and found to be relatively inde pendent and intra-class correlation were used to calculate test-retest reliabilities. Predictive validity was determined by us ing step wise multiple regression analysis with type of diabetes co-varied to predict compliance. The best predictor of compliance was the lack of p sychological barriers. The structural elements factor and the susceptibility fact or added significantly (p < 0.05; p<0.03, respectively) to predicting compliance. All seven factors were able to predict compliance accounting for 15% of the total variance. Multiple aggression analyzes was also used to predict metabolic control from the seven DHBS factor scores. Metabolic control was a composite score from four laboratory values, hemoglobin A1c, fasting blood sugar, 24-hour urine, and fasting triglycerides. Health belief fa ctors predicted control of diabetes better than compliance, accounting for 23% of the variance. The best predictor of metabolic control was perceived severity followed by treatment beneficial,

PAGE 84

71 susceptibility, and lack of psychological barriers. Metabolic control is the objective of compliance and therefore refl ects the goals of comprehensive selfcare training programs. Several limitations of the DHBS persi st. The dimension of self-efficacy was not part of the HBM but has sinc e been added to improve the predictive and explanatory powers of the expanded model. In addition current HBM literature suggests that diabetes self-care is affe cted by environmental barriers as well as psychological barriers. Individuals with di abetes vary in terms of the number of barriers that affect them and thei r sensitivity to each barrier. Previous scales designed to measure perceived barriers to diabetes selfcare (Ary, et al., 1983; Davis, et al., 1987; Schafer, et al., 1983) have confirmed the significant influence that barriers have on self-care behaviors. Most HBM scales are limited by the number of barri ers they measure and their inability to differentiate the level of severity of eac h barrier. The Environmental Barriers to Adherence Scale (EBAS) developed by Ir vine, Saunders, Blank, and Carter (1990) at the University of Virginia, addresses 60 barriers across the four major areas of diabetes self-care (diet, exercise, blood glucose testing, and medication). The scale is able to differ entiate how intensely each barrier affects the person. The scale has been rigorously tested for re liability and validity. Test-retest for the total scale was 0.80 (p < 0.001). Test-retest scores for each subscale

PAGE 85

72 ranged from 0.59 for both medication and ex ercise to 0.74 for glucose testing and diet barriers. A high Cr onbach alpha was obtained for the total scale (0.94). The coefficients for diet (0.91), exerci se (0.86), glucose testing (0.86), and medication (0.84, also indicated high inte rnal consistency. Va lidity was assessed by four measures. Content validit y was made by having five diabetes professionals review the scale. Concurr ent validity was assessed by correlating the scale with two barrier scales, t he Barriers to diabetes Adherence Scale (BAS), and the diabetes-Care Profile Barrier s (DCP-BAS) subscale (Davis, et al., 1987; Schafter, et al., 1983). The co rrelations (r = 0.63) with the BAS and (r=0.51) the DCP-BAS subscale reflected a m oderate level of concurrent validity. Statistical conclusion validity was det ermined by correlating the EBAS with a reliable measure of self-care behavior, the Diabetes Self-Care Behavior (DSCB) scale (Wilson, Ary, et al., 1986). The DSCB is a 24-item self-report scale that measures levels of adherence for diet, ex ercise, medication (oral and insulin), glucose testing, alcohol consumption, foot care, and carrying diabetes identification. Correlations ranged from r = -0.33 to –0.52 (p < 0.05) reflecting the negative relationship between barriers and adherence behaviors. The total EBAS scale was also found to be correlated with Hemoglobin A1c (r=0.28, p<.001). The scale was found to have di scriminate validity. Each of the EBAS subscales (diet, exercise, medication, and glucose testing) correlated well with corresponding measures of self-care behavior and less well with non-

PAGE 86

73 corresponding self-care behavior (foot ca re, alcohol consumption, carrying diabetes identification). Correlations between the SCBS and the total EBAS scores ranged from r=0.73 to r=0.86. The external validity of the EBAS is limited due to its validation with mainly a type 2 diabetes population. The se lf-report nature of the scale and the validation with other self-r eported adherence measures ca n be biased; therefore, future validation with other measures of self-care behaviors would be beneficial.

PAGE 87

74 Comprehensive Diabetes Self -Care Management Training For many years, diabetes education has been viewed as an important part of diabetes care. Since the discovery of insulin in the 1920s, leaders in the diabetes field have recognized the important role diabetes education plays in the patient's ability to manage the disease. Over the years, diabetes education has been the focus of hundreds of research studies Most of the earlier research that focused on knowledge acquisition report ed that depending on the combination of education interventions most showed beneficia l effects. Brown (1988) examined the effects of diabetes education in a meta-analysis study of 47 studies conducted from 1954 to 1986. Overall, Brown found that the mean unweighted effect size was .91 (SD 0.75) and the weighted mean effect size for controlled studies was .33 (SD 0.01) The strongest evidence of beneficial effects of diabetes education was the large weighted mean effect size of 0.84 calculated from the 13 studies that reported using glycosyla ted hemoglobin as the dependent variable. In the 1970s, new techniques became available to assess the metabolic control of diabetes. Glycosylated hemoglobin (GHB), a form of hemoglobin that reflects the average blood glucose conc entration over a three-month period, gave researchers and clinicians a tangible, reliable measure of glycemic control (Gabbay, et al., 1977; Koenig, et al., 1976). Technological advances in blood glucose monitoring allowed patients to perform glucose testing and make

PAGE 88

75 management decisions at home. These new techniques allowed researchers to measure the metabolic outcomes of diabetes education efforts. In a similar meta-analysis of diabe tes education, Padgett (1988) and her colleagues examined 93 studi es conducted from 1976 to 1986. Overall, the unweighted mean effect size of 0.51 pr ovided more support for the positive effects of diabetes education. The authors were quick to point out that many of the studies were of poor quality and suffe red weak study designs and numerous methodological problems. Brown crit icized numerous studies attempting to measure knowledge scores with instru ments that had not been tested for reliability or validity. Internal validity wa s a problem in several studies included in the meta-analysis that failed to examine t he effects of high attrition rates. Two earlier studies (Mulhauser, et al., 1983; Paulozzi, Norman, McMahon, & Connel, 1984) attempted to measure the effect s of diabetes education on metabolic control but without the use of a control. The possib ility that the improvement found could be a result of something besid es the education intervention could not be ruled out. Although the results of the DCCT were promising, the practical considerations of implementing in tensive management and comprehensive education programs in today's health care climate have been difficult to overcome. With the introduction of prospective payment and managed care, many health care providers are faced wit h disincentives to provide diabetes

PAGE 89

76 education. In the 1980s, a Federal surv ey (Leicher, 1986) found less than 10% of the hospitals had structured dia betes self-care management training programs. Coonrod et al (1994) found only 35% of the diabetes population reported having attended any diabetes cl asses or education programs. The major impediments to receiving a dequate diabetes education are lack of physician referral and inadequate reimbursement for education services. Whereas the effectiveness of diabetes self-management training or education to improve metabolic control controversy st ill exists, there is a danger that those pressed to reduce costs may choose to agree with those studies that favor their own biases. Will the number with severe knowledge deficits grow higher than the already high 50-80% levels previously reported (Coonrod, Betschart, & Harris, 1994; Watkins, et al., 1967; Williams, et al., 1967)? Despite the plethora of studies that have attempted to answer the question of whether diabetes educati on effectively improves met abolic control, the results are still inconclusive. Early studies ex amining the relationship between diabetes education and metabolic control were lar gely cross-sectional, correlational, or causal-comparative allowing no causal infe rences to be made. Little attention was paid to having adequate power to detect a difference if one truly existed. Sample sizes used were generally too small to detect realistic effect sizes that may actually exist between usual care and other educational interventions. Due to ethical considerations, educational in terventions could not be compared with a

PAGE 90

77 no education control group. In addition, most studies were criticized for employing weak research designs that lacked random selection and assignment, proper comparison with a control group, and inadequate contro l of extraneous factors that could infl uence glycemic control. Educational interventions varied in composition and comprehensiveness among studies making it difficult to a rrive at any conclusions about whether diabetes education was effective or not (Clement, 1995). Pr ompted by these concerns, Congress established the Nationa l Diabetes Advisory Board in 1983 to attempt to structure diabetes educati on and develop standards. Educational programs that meet the national standards receive re cognition from the American Diabetes Association. ADA approved programs follow a goal-oriented approach that has been found to be highly successful at improving metabolic control (Assal, et al., 1985). The goal-approach a ssumes that the patient is able to define or set goals and has the commitment and know-how to achieve the goals defined. The goals of diabetes self-m anagement education are to help the patient acquire the skills and knowled ge to manage the day-to-day issues of diabetes, handle crises, and prevent diabetes-related complications. The individual with diabetes must learn to test his or her own blood sugars, interpret the results, and balance their diet, exer cise and medication. Key elements of ADA comprehensive diabetes self-care management training include: an assessment of their educational needs; diabet es teaching individually tailored to

PAGE 91

78 address the needs identified in the a ssessment; and a follow-up assessment to identify deficiencies and plan strategies to meet the patient's goals (Street, et al., 1993). ADA recognized programs follow st andards for structure, process, and outcomes. Obviously, not all persons with diabetes are willing or able to perform the self-care behaviors needed to achieve goo d metabolic control. Benefits of diabetes education may be minimal or non-ex istent if patients are not motivated or unable to make improvements in self-c are. Is it fair to determine the effectiveness of all diabetes self -care management training programs on metabolic control when some of these programs do not include metabolic control in their objectives? In one of the few prospective studies that used a random, controlled design, Bloomgarden, et al. (1987) attempted to compare an educational intervention group with a usual-care group. The education intervention utilized a traditional didactic approach that stated metabolic control was not one of the program's objectives. The sample size a priori seemed adequate but failed to take into account t he high rate of attrition expected in a long-term trial. After randomization of 749 type 1 patients, 30% of the experimental group, and 26% of the c ontrol group refused to participate, jeopardizing the intent of randomization. Only 71 out of 165 subjects in the experimental group completed seven out of nine educational sessions. Bloomgarden was unable to detect a signifi cant difference in metabolic control

PAGE 92

79 between the experimental group and the c ontrol group. Mulhauser and Berger (1993) criticized the study for its unstr uctured, didactic educational approach and its lack of information on the insulin tr eatment used. They cautioned that diabetes education cannot compensate for deficiencies of inappropriate insulin treatment regimens. Another controlled study conducted by Korhonen, et al., (1983), failed to demonstrate the impact of diabetes educatio n on metabolic control using didactic educational methods. Their sample size was small (N = 72) and therefore the power was too low to detect significant differences between the experimental group receiving the 8 12 hours diabetes education intervent ion and a control receiving .5 hours of basic education. Their study was also criticized for inadequate insulin therapy confounding the re sults. The study did not measure glycosylated hemoglobin which is consi dered essential when attempting to measure metabolic control. In contrast, long-term prospective studies that incorporated behavioral change strategies demonstrat ed improvement in metabo lic outcome (Delamater, et al., 1991; Hartwell, Kaplan, & Wallace, 1 986; Kaplan, et al., 1987; Laitinen, et al., 1993; Raz, Soskolne, & Stein, 1988; Wo oldridge, et al., 1992). In a largescale randomized, controlled clinical trial, Mazzuca et al. (1984) found a significant difference between the exper imental group receiving comprehensive diabetes education and a control group receiving routine education. They

PAGE 93

80 studied the effects of educat ion on knowledge, skills and self-care behaviors on a study population of predomi nantly elderly black wom en with non-insulin treated type 2 diabetes. Garrard et al. (1990) f ound, using a quasi-experimental design, a statistically significant improvement in Hgb A1c values from preto postintervention in the experimental group receiving a five-day comprehensive diabetes self-care management training pr ogram. Because of their weaker design, these improvements in metabo lic outcome can only be associated with the treatment rather than c aused by the intervention.

PAGE 94

81 Continuous Glucose Monitoring A new glucose monitoring device approved by the FDA in 1999, the Continuous Glucose Monitoring System (CGMS) developed by MiniMed Inc., provides a novel minimally invasive tec hnique that continually measures glucose levels using a miniscule (0.5ul) sample of dermal interstitial fluid. The Continuous glucose monitoring system re cords tissue glucose levels at five minute intervals for up to 3 days. Information pr ovided from the continuous glucose monitoring system CGMS) has hel ped identify patient-specific patterns of glucose control. Previous studies hav e shown that dermal interstitial glucose measurements by the CGMS closely correlate to capillary glucose meter measurements and venous blood laboratory values (Jansson, Fowelin, Smith & Lonnroth, 1988; Service, O’Brien, Wise, Ness & LeBlanc, 1997; Bolinder, Hagstrom-Toft, Ungerst edt &Arner, 1997). Although capillary self-monitoring of blood glucose (SMBG) levels has been a mainstay in intensive management of diabetes, the true variations in blood glucose levels may not be detected by conventional SMBG recording. The glucose sensor data in conjunction with frequent SMBG measures can be helpful to identify glucose excursions incl uding asymptomatic hypoglycemia. The information obtained from the sensor can facilitate decisions concerning diabetes management and provide evaluation of treatm ent options. The sensor may also be used to help patients see directly t he glycemic effects of food intake and

PAGE 95

82 exercise and may motivate or act as cues to facilitate behavior change. A pilot study using the MiniMed continuous gl ucose monitoring system in nine poorly controlled type 1 subjects (mean HgbA1c 9.9%) showed a reduction of mean HgbA1c to 8.8 % in a 5 week period (B olinder, Hagstrom-T oft, Ungerstedt & Arner, 1997).

PAGE 96

83 Implications for Future Research A prospective study design is needed to assess the role of health beliefs, and the conditions under which they ma y be modified to improve the healthrelated outcome. The real promise of the Expanded Health Belief Model lies in its usefulness not only to explain a nd predict behavior but to apply this understanding to develop different intervent ion strategies to change these beliefs in order to direct their behav ior towards the right directio n. Certain variables in the model can be modified to produce the desired outcome. Some HBM variables have already been successfully manipulated to improve health-related actions. For example, action can be triggered by appropriate communications such as reminder postcards, follow-up phone calls, lab reports, and mass media campaigns. New technologies such as t he continuous sensor can be employed as cues to action by showing patients t he effects of their di abetes therapies. Barriers can be minimized by addressing t he salient issues of the individual involved such as cost of the treatment, insurance co verage, fear of adverse reactions, or complexity of the regim en. By knowing t he perception of the individual, the practitioner may dis pel any misconceptions, and improve the persons’ understanding of t he efficacy and relevant issues relative to the advocated health action. The scheme of t he Health Belief Model developed by Becker (1974) has been revised to illustrate the mutable variables that affect diabetes self-care (Figure 4). The highlighted variables reflect the variables that

PAGE 97

84 have potential for improvement from diabet es education intervention. Knowledge of the disease, perceived susceptibil ity, and seriousness of the disease, perceived threat of the di sease, perceived benefits minus barriers, as well as modifying factors such as attitudes, be liefs, skills, self-efficacy, and cues to action can be modified to improve t he likelihood of diabetes self-care management. In summary, the literature suggests that most diabetes education interventions have positive effects on knowledge; but in order for diabetes education to effectively improve metabolic c ontrol it must be able to motivate and enable the patient to adopt self-care behaviors In order to overcome many of the obstacles encountered in previous st udies, several conditions must be met. First, a prospective randomized, controlled trial is necessary to assure that demographic, psychosocial, and clinical fa ctors have an equal chance of being in the experimental or control group. A dditional follow-up studies are needed to examine the long-term effe cts of diabetes education on glycemic control, complication rates, and other important outcomes such as costs associated with increased health care utilization, disa bility, and quality-of-life issues.

PAGE 98

85 Figure 4. The Health Belief Model as Predictor of Diabetes Self-care Behavior Perceived susceptibility Perceived seriousness to disease "x", symptoms or complications (severity of disease "X") Demographic variables Sociopsychological varaibles Structural variables* Social-personal factors* Program characteristics* (age, sex, race, ethnicity, etc.) (personality, social class, peer and reference group pressure, educational attainment, support) ( prior contact with and experience with the disease, and duration of disease)Knowledge about the disease, (self-efficacy," skills, coping skills) (Complexity, competehce of instructions, ADA recognized teacher method, atmosphere) Perceived benefits of Perceived barriers to action minus Perceived threat of disease "X", symptoms or complications Intention Motivation Likelihood of performing Diabetes self-care (taking action) Health Behavior Maintained Cues to action External(Incentives/reinforcements) Mass media campaigns postcard from physician or dentist, phone follow up Illness of family member or friend Newspaper or magazine article Advice from others Reminder Internal Feedback from health providers/professionals, labs, symptoms MODIFYING FACTORS INDIVIDUAL PERCEPTIONS LIKELIHOOD OF ACTIONPhysical: discomfort, pain Emotional: fear of side effects, complexity, feels burdened by regimen Social economic: time constraints, financial outlay, incompatibility of life style Note: Revised from Becker, M.H., ET al ., 1974, Am. J. Public Health 64, 205216.

PAGE 99

86 Chapter Three: Methods This chapter describes the methods to be used in this study. It includes the purpose of the study, t he hypothesis, research questions, and the research plan consisting of the population and samp le, instrumentation, procedures, data collection and analysis.

PAGE 100

87 Purpose of the Study The purpose of this study was to determine if persons with diabetes participating in an ADA diabetes se lf-management education with intensive follow-up care achieve better metabolic control than the standard follow-up care. The intensive follow-up group differs from the standard fo llow-up group by utilizing a MiniMed continuous glucose sens or. This study sought evidence as to whether the data obtained from the sens or could lead to improved HgbA1c values by facilitating understanding of pati ent specific patterns of glucose control and in facilitating appropriate ch anges in diabetes management. The specific aims of this investigation were: 1. To determine whether there was a st atistically significant difference in mean Hgb. A1c values over time (week 1, week 12, and week 24) for the intensive follow-up group (SFG) com pared to a standard follow-up control group (IFG). 2. To determine if there were statisti cally significant changes in mean scores for health beliefs measured by the Revised Diabetes Health Belief Questionnaire subscales (perceived barriers, perceived severity, perceived susceptibility, perceived benefits, cues to action, health motivation, perceived self-efficacy abili ty, and structural elements) over time (from week 1 to w eek 24) for the intensive follow-up group compared to the standard followup control group.

PAGE 101

88 3. To assess whether persons with diabet es who participated in the intensive follow-up group had any significant di fferences in diabetes self-care practice scores compared to t he standard follow-up control group.

PAGE 102

89 Hypothesis This prospective study compared the effectiveness of an intensive followup group compared with standard care fo llow-up of participants of an ADA diabetes self-management educat ion program on metabolic control, diabetes self-care behaviors and health beliefs. The randomized pretest-posttest control group design tested the following null hypothesis: Ho: There is no significant difference in mean glycosylated hemoglobin (HgbA1c) value over time (week 1, week 12, and week 24) for participants of a ADA diabetes self-management education program assigned to the intensive follow-up group (Utilizing the MiniMed s ensor) compared to a standard follow-up control group.

PAGE 103

90 Research Questions This investigation also attempted to answer the following questions: 1. Is there any significant diffe rence in mean glycosylated hemoglobin values over time (week 1, week 12, and week 24) between those participants of the ADA diabet es selfmanagement education (DSME) randomly assigned to the intensive follow-up group or the standard follow-up control group? 2. Are there any significa nt profile differences in mean Health Belief Model subscale scores (perceived barriers, severity, susceptibility, benefits, self-efficacy, cues to action, motivation and structural elements) between the DSME partic ipants assigned the intensive follow-up and the standard followup control group over time? 3. Is the difference between t he week 1 and the week 24 mean perceived barrier scores of t he intensive follow-up group significantly different compared to the standard follow-up control group? 4. Is the difference between t he week 1 and the week 24 mean perceived benefits score of the intensive follow-up group significantly different compared to the standard follow-up control group? 5. Is the difference between t he week 1 and the week 24 mean

PAGE 104

91 perceived self-care ability scores of the intensive follow-up group significantly different compared to the standard follow-up control group? 6. Is the difference between t he week 1 and the week 24 mean perceived severity scores of t he intensive follow-up group significantly different compared to the standard follow-up control group? 7. Is there a significant differenc e between the week 1 and the week 24 mean perceived susceptibility scores of DSME participants randomly assigned to the intensiv e follow-up group compared to the standard followup control group? 8. Is there a significant differenc e between the week 1 and the week 24 mean perceived cues to acti on scores of DMSE participants randomly assigned to the intensiv e follow-up group compared to the standard followup control group? 9. Is there a significant differenc e between the week 1 and the week 24 mean structural elements scores of DMSE participants randomly assigned to the intensive followup group compared to the standard follow-up control group? 10. Is there a significant differenc e between the week 1 and the week 24 mean diabetes self-care practi ce scores of DSME participants

PAGE 105

92 randomly assigned to the intensiv e follow-up group compared to the standard followup control group? This study seeks evidence as to whet her type 2 diabetes participants of an ADA diabetes selfmanagement educati on program randomly assigned to the intensive follow-up group have an y significant differences in their HgbA1c values, health beliefs, and diabetes self-care behaviors over time compared to a standard follow-up group. The intensive follow-up group differs from the standard follow-up group by utilizing a cont inuous glucose sensor to determine blood glucose patterns and responses to ex ercise, medications, food intake, and other daily activities.

PAGE 106

93 Research Plan Population and Sample The sample was recruited from a di abetes population receiving care within the Tampa Bay area. The area has an et hnically diverse population. Numerous cultures and socio-economic backgrounds are represented, such as AngloAmericans and persons of other European he ritage, African-Americans, Asians, Hispanics, and Native Americans. The Tampa Bay area is comprised of both urban and rural communities. The study population was lim ited to literate English-speaking Adults ages 18 or older with type 2 diabetes. The study is further limited to consenting adults referred to the University Community Hospital by their physicians. Participants had to be referred to the program by the physicians who had evaluated their health st atus and determined that they were medically stable and able to participate in all diabetes self-care activities. Persons with medical problems that impair their ability to perform diabetes selfcare management were ineligible for the study. The case manager during the initial assessment validated t he health status of persons referred to the study. Pertinent medical information was doc umented on the baseline evaluation form (see Appendix 1). Any physica l and mental impairments t hat limit parti cipation in diabetes self-care were discussed with the referring physician who determined whether the person was able to participate in the study. Other exclusion criteria for the study are listed in Table 1.

PAGE 107

94 The recruitment strategies included advertising in newsletters published by the University Community Hospital. Letters and flyers were mailed to Tampa area physicians and persons with diabetes interested in attending DSME. Persons with diabetes who were on existi ng mailing lists also were sent flyers about the study, including how to be referre d by a physician. Flyers also were distributed at the USF Medical Clinics, University Square Mall and Citrus Park Mall Health Source Centers, and at diabet es special events such as University Community Hospital annual Sugarfest and at diabetes fundraiser events (see Appendix 2). Both physicians and pati ents were made aware of the purpose and design of the study. During the initia l assessment, patient informed consent forms were signed prior to random assi gnment (see Appendix 3). The study was reviewed and approved by the University of South Florida Inst itutional Research Board and the University Community Hospital Research Committee.

PAGE 108

95 Table 1. Exclusion Criteria Persons with type 1 diabetes gestational diabetes, or secondary diabetes Pregnant women Patients with concomitant condition s or in poor health that impairs ability to participate in self-care management activities or limits life expectancy as determined by their physician Persons with diabetes who have participated in other diabetes selfmanagement education program s within last year Persons with diabetes who are unwilling to attend scheduled sessions over a six-month period Persons with diabetes who have any form or history of hemoglobinopathy or hemolytic proc ess that could interfere with reliable assessment of Hgb A1c Persons with diabetes with HgbA1c less than 7.0% or greater than 13.0%

PAGE 109

96 Enrollment continued until a minimum of 159 participants completed the program to achieve adequate po wer (0.80) at a 0.05 alpha le vel. A priori sample size was based on minimum sample si ze calculations for a 2-group repeated measures design (Vonesh & Chinchilli, 1997) for a maximum correlation of .90 with a minimum standardized difference (T/s ) of 0.5 (see Table 2). The control group size of 78 allowed for a 30% attriti on rate and the intervention group size of 81, a 35% attrition rate. Self-addressed re minder cards were completed. Prior to appointments, reminder cards were mail ed or reminder phone calls were made to help minimize attrition.

PAGE 110

97 Table 2. Minimum Sample Sizes per Group for Te sting the Hypothesis of Equal Group Effects in a Two Group RM Design Number of groups Number of treatments (time) Maximum correlation Standardized difference: T/s 0.50 0.75 1.00 1.25 1.50 2.00 2 2 0.0 0.1 0.3 0.5 0.7 0.9 32 35 41 48 54 60 14 16 19 21 24 27 9 10 13 14 15 16 7 7 8 9 10 11 5 6 6 7 8 8 4 4 4 5 5 5 3 0.0 0.1 0.3 0.5 0.7 0.9 21 26 34 41 51 59 10 13 15 19 23 27 7 8 10 12 14 16 5 6 7 8 10 11 4 5 5 6 7 8 4 4 4 4 5 5

PAGE 111

98 Table 2 (continued). Number of groups Number of treatments (time) Maximum correlation Standardized difference: T/s 4 0.0 0.1 0.3 0.5 0.7 0.9 16 21 30 40 49 59 8 10 14 18 22 26 6 7 9 11 14 16 5 5 6 7 8 11 4 5 5 6 7 8 5 5 5 5 5 5 5 0.0 0.1 0.3 0.5 0.7 0.9 13 18 28 38 48 58 6 8 13 17 22 26 6 6 8 11 13 16 6 6 6 8 9 11 6 6 6 6 7 8 6 6 6 6 6 6 6 0.0 0.1 0.3 0.5 0.7 0.9 11 16 27 37 48 58 5 8 12 17 21 26 7 7 8 11 13 16 7 7 7 7 9 11 7 7 7 7 7 8 7 7 7 7 7 7 Note. Adapted from Rochon (1991, Table 3).

PAGE 112

99 Procedures Subjects with diabetes who were willing to participate and meet eligibility requirements signed the informed consent and completed the initial Revised Health Belief Questionnaire. A comp lete diabetes assessment (following American Diabetes Associat ion guidelines) and initial evaluation forms were completed by the case manager (see Appendi x 1 for initial evaluation and initial assessment forms). To reduce selecti on bias, both the interviewing case manager and participants were not aware of group assignment at the time of the initial assessment. To minimize selecti on bias, a program assistant not involved in patient assessment or t eaching randomly assigned patients to an intervention or control group. In the initial diabetes assessment a variety of demographic, psychosocial, and clinical information was compiled including: the patient's medical history; present health status ; previous education; health services utilization; diabetes knowledge; skills; attitudes; social and financial support systems; barriers to learning; self-care; and associated medical conditions or risk factors. During the initial visit, patients comp leted the Revised Diabetes Health Belief Questionnaire (see Appendix 4) comprised of the DHBS, EBAS, and DSES (Harris, et al., 1987; Hess, Davi s, & Van Harrison, 1986; Irvine, 1996).

PAGE 113

100 Each scale was tested rigorously and found to be valid and reliable. Content validity was examined during a pilot te st, and reliability was confirmed with the target diabetes population. The case manager scored the various subscales of the questionnaire and recorded them on the in itial evaluation form. The program assistant placed all forms in medical record s. To reduce potentia l for investigator bias, testing information was recorded s eparately and filed in medical records until compiled for data analysis. Physiological Instrumentation The primary variable selected to meas ure diabetes control is glycosylated hemoglobin or Hgb A1c. The Hgb A1c has been substantiated in the literature as a reliable biochemical test able to quantif y the average glycemic levels for an 812-week period. Glycosylated Hgb is also a useful surrogate marker for development of diabetes-related complicat ions (Gabbay, et al., 1977; Koenig, et al., 1976). The Hgb A1c employed a c apillary blood glucose sample analyzed using a Micro Optical Detection method (A 1cNow, Metrika, Inc, Sunnyvale, Ca). The normal range of the assay is 4.3 to 5.9. Comparisons with a National Glycohemoglobin Standardizati on Program (NGSP) cert ified reference method showed a within-laboratory pr ecision of 5%. Capillar y-drawn blood and venousdrawn blood have shown to be equivalent (R2 = 0.89). Cagliero, Levina, and Nathan (1999) found that immediate capill ary HgbA1c testing provided accurate and reliable results in office based settings. The capillary blood volume

PAGE 114

101 requirement for the test is 10 micr o-liters obtained by a standard lancet technique. The analytical method is able to detect abnormal hemoglobin (e.g. sickle cell, anemia, polythysemia), interfering substances (e.g. bilirubin, coumadin) and nonviable samples. A pilot test was initiated by the Diabet es Treatment Center at University Community Hospital in August 1996 to analyze the feasibility of using capillary Hemoglobin A1c measurements, preparticipation and post-participation in a diabetes self-management outpati ent program. Thirty vo lunteer participants who were tested prior to the program also were re-tes ted three months after completing the diabetes self-man agement program. The mean Hgb A1c before the program was 9.26% com pared to the mean Hgb A1c of 7.52% three months after completion of the program (Figure 5). Ten capillary-drawn samples were compared with simultaneously venous-drawn sample analyzed by the central lab method, which uses electrophoresis. A coef ficient of variation of .997 was found between the two methods. The inter-rater Kappa coefficient was .97.

PAGE 115

102 Figure 5. Pilot Study Results

PAGE 116

103 Psychosocial Instrumentation The revised Expanded Health Belie f Model Questionnaire (EHBMQ) was designed from three estab lished diabetes-specific in struments; the Diabetes Health Belief Scale (DHBS) (Harris, et al., 1987); the Envir onmental Barriers to Adherence Scale (EBAS) (Irvine, Saund ers, Blank, & Carter, 1990); and the Diabetes Self-Efficacy Scale (DSES) (Cr abtree, 1987). The scales selected had undergone rigorous reliability and validity testing. The DHBS was developed and validated by Harris, et al (1987). Test items we re generated from the HBM literature and previously developed scales. Test-retest correlates on a sample of 30 men were used to determine item reliabili ty. Items with greater than 0.35 testretest correlation and standard deviations of 0.40 were retained. A factor analysis was conducted on a larger sample of 280 male veterans, 7 interpretable factors representing the 7 dimensions of the HBM (s everity, susceptibility, predictive psychological ba rriers, treatment beneficial, cues to action, health motivation, and structural elements) were found providing evidence of construct validity. Convergent valid ity was established with a tested sample of 120 men. The developers found that health beliefs a ccounted for 15% of the total variance in compliance to diabetes self-care and 23% of the total variance in diabetes control. The EBAS was added to the questionnaire to overcome one of the major limitations of the DHBS. The scale only measures psyc hological barriers. In addition, Irvine et al. (1990) overcame other shortcomings of previous scales by

PAGE 117

104 increasing the number of ba rriers measured and being able to differentiate the sensitivity of each individual to the barri ers. The EBAS was rigorously tested for reliability and validity using 214 individual s with type 1 and type 2 diabetes. The test-retest reliability for the total scale was 0.80 (p < 0. 001). Test-retest reliability scores on the subclass were 0.59 for m edication barriers, 0.59 for exercise barriers, 0.74 for glucose testing barriers, and 0.74 for diet ba rriers (p < 0.001 for each subscale). Cronbach’s alpha for t he total scale was 0.94, with subscales ranging from 0.84 to 0.94, indica ting high internal consistency. Content validity was established by hav ing a panel of five diabetes experts review the scale for representativeness. Evidence of concurrent validity was established through the correlation with two other barrier scales, the Barriers to Adherence Scale (BAS) developed by Gl asgow, McCaul and Schafer (1986), and the Diabetes Care Profile – Barriers to Adherence Subscale (DCP-BAS) developed by Davis, Hess, Van Harrison and Hiss (1987). Moderate correlation were found ranging from r = 0.34 to 0.62 for the BAS and r = 0.30 to 0.51 for the DCP-BAS. Convergent validity was determined by the scale’s correlation with selfcare behavior, measured by the diabetes self-care behavior (DSCB) scale. Selfcare behaviors included in the scale we re diet, exercise, medication, glucose testing, alcohol consumption, foot care and carrying diabetes identification. The EBAS correlated well with each corre sponding adherence behavior (diet,

PAGE 118

105 exercise, medication, and glucose testi ng), and less well with non-corresponding behaviors (foot care, alcohol use, and carry ing identification). Correlation with corresponding adherence behaviors ranged from r = -0.33 to -0 .52 (p < 0.05), reflecting a consistent negative relations hips between behaviors. In addition, Hgb A1c was found to be correlated with the total EBAS (r = 0.28, p < 0.001). The EHBMQ also includes a diabetes self-efficacy subscale; a ten-item subscale was devised from the self-e fficacy theoretical literature (Bandura, 1977a; 1977b), and the Diabetes Self-E fficacy Scale (DSES) developed by Crabtree (1987) in collaborat ion with Bandura. The DSES work was tested on 143 persons with type 1 and type 2 diabetes yi elding internal consistency for the total scale (standardized alpha = 0.71). Using the DSES, Crabtree found 25% to 33% of the variance in diabetes self-c are behaviors could be explained by selfefficacy (after controlling for the effects of age, sex, marital status, types of diabetes, severity, and number of complicat ions) providing evidence of construct validity. The revised Expanded Health Belief Model Questionnaire comprised of the three established diabetes-specific health belief scales was designed to measure the following aspects of the model: Perceived susceptibility which m easures the person with diabetes perceptions of their own risk of deve loping diabetes or its complications, and measures the perceived threat of becoming ill or having a condition

PAGE 119

106 become worse from diabetes. Perceived severity is a person’s perception of the consequences of diabetes and it s associated complications (pain, death, disability) and how much interference in their lives they perceive are due to diabetes. Perceived benefits measure beliefs regarding the effectiveness of diabetes self-management in reducing the threat of diabetes and its complications. It also measures the person’s faith in curative powers of the physician and that the recommended action or intervention by their physician will result in a reduction in the health threat. Perceived barriers measures psycholog ical and situational obstacles that would have to be overcome in order for a person to follow the recommended diabetes regimen. It measures the potential negative aspects of the diabetes treatm ent perceived by the person. Cues to action assess the symptoms or motivating factors such as health messages that are likely to lead to seek ing medical care or follow positive health behaviors. Structural elements assess modifying factors such as a person’s locus of control, social support, and underst anding of diabetes and the treatment regimen. Health motivation measures general heal th motivation or concern for one’s own health.

PAGE 120

107 Perceived self-efficacy measures a person’s own assessment of their ability to perform diabetes self-care or engage in diabetes-specific health behaviors. Each item in the questionna ire was assigned to one of the eight subscales designed to measure the dimension. Responses follow a common 4-point Likert scale. The higher the scores within each subscale, the greater the influence of the belief or concept. Content validity of the revis ed EHBMQ was strengthened by use of established reliable and valid scales, a thorough literatur e review, personal diabetes education experience, and the use of a panel of experts, eight certified diabetes educators, four HBM research experts, and one statistical measurement expert. The diabetes expe rts evaluated the instrument with regards to diabetes self-care management, while the HBM expe rts evaluated the items in terms of the representative of the di mensions of the EHBM. T he statistical measurement expert reviewed the scales for overall measurement design. During pilot-testing of the revis ed questionnaire, ten individuals with diabetes who previously participated in diabetes self-care management training were asked to complete the questionnair e on two separate occasions,10-14 days apart. Comments about the questionnaire it ems were encouraged. Overall, the questionnaire was found to be easy to understand and to administer, taking between 10 and 20 minutes to complete Feedback from the individuals indicated that the wording and reading level were appropriate for the target

PAGE 121

108 population. The SMOG readability formula applied to the revised questionnaire indicated an 8th grade level. The SMOG is gener ally reliable within 1.5 grade levels (McLaughlin, 1969). Concurrent validit y is if the HBM fa ctors are logically associated with metabolic control. Low er barrier scores, higher severity, susceptibility benefits, self-efficacy, c ues to action, health motivation, and structural element scores are expect ed to account for improved Hgb A1c values. Expanded Health Belief Model Quest ionnaire Reliability and Validity Results Although previous studie s established the validity and reliability of the health belief subscales used to make up the revised Expanded Health Belief Model Questionnaire (EHBMQ), the instru ment was pilot tested in the type 2 diabetes target population. Ten patients were asked to comment on all aspects of the questionnaire: time it takes to comple te, ease of format, any questions that were unclear or ambiguous. During pilot testing, questions that were ambiguous or difficult for the target group to under stand were modified. The instrument was examined by a panel of diabetes experts for content validity, ease of use and SMOG testing to confirm an 8th-grade reading level. Reli ability was re-confirmed by test-retest measurement using ten additional patients with type 2 diabetes; the revised questionnaire was adm inistered twice within 10 to14 days to confirm testretest reliability. Statistical analysis of the revised questionnaire was performed using SAS procedures (SAS In stitute, Cary, NC). Table 3 shows which questions

PAGE 122

109 are included in each scale, the range of po ssible scores, and the interpretation of the score.

PAGE 123

110 Table 3: Revised Diabetes Health Belief Model Questionnaire Formulae Scale name Subscale name Scale equationa Range of possible scores on scale Interpretation of score Perceived susceptibility --9, 11, 19, 21, 23, 32, 35, 39, 43 9-36 >18 considered high Perceived treatment benefits --5, 6, 7, 8, 10, 13, 17, 18, 31, 40 10-40 >20 considered high Perceived severity --16, 20, 22, 25, 26, 36, 38, 41 8-32 >16 considered high Cues-to-action --15, 29, 37, 42, 44, 45 6-24 >12 considered high Perceived barriersb --4, 14, 57, 60, 63, 66 65-260 >126 considered high Perceived barriers Medication 57.1-57.15 15-60 >30 considered high Perceived barriers Diet 60.1-60.15 15-60 >30 considered high Perceived barriers Exercise 63.1-63.16 16-64 >32 considered high Perceived barriers Monitoring 66.1-66.17 17-68 >34 considered high

PAGE 124

111 Table 3 (continued). Scale name Subscale name Scale equationa Range of possible scores on scale Interpretation of score Structural elements (understanding, social support) --3, 12, 24, 34 4-16 >8 considered high Motivation --1, 2, 27, 28, 30, 33 6-24 >10 considered high Self-efficacy --46, 47, 48, 49, 50, 51, 52, 53, 54 9-36 >18 considered high Note: The range of scores per question is 1 to 4. Mi ssing item scores 1 (Not at all) on all scales except for Perceived Barriers questions 57 (15-60), 60 (15-60), 63 (16-64), and 66 (17-68) aQuestion numbers. bPerceived Barriers subscales: The following subscales are calculated and represent one "question" in the Perceived Barriers scale: M edication (Q57), Diet (Q60), Exercise (Q63), Monitoring (Q66). The results of the ANO VA confirmed good test-retest reliability for each Health Belief subscale (see Table 4) In Table 4, each row represents an ANOVA model. In the model, the score on the scale or subscale is the dependent variable, and the two independent variables introduced into the model

PAGE 125

112 are patient identificat ion number (“patient” in the Table 4), and time (1 or 2). The F-test value and the p-value on the F-test are reported. If the F-test is large enough so as to generate a small enough pvalue, then the ANOVA model is considered to fit well enough to allow interp retation of the mean square values for each covariate. All F-tests had an associat ed p-value of <0.05, so all models were interpreted. The patient mean squar e p-values were all <0.05, which would be expected. This suggests that patient identification explained a significant amount of variability in the m odel. If the instrument is reliable, the time covariate should not explain a significant amount of va riability in the model. In other words, it should not matter when the patient comp letes the instrument, the results should be similar. The time covariate was highly non-significant in all models, suggesting this was the case. The correlati on coefficients ranged from 0.65 for the perceived severity scale to 0.91 for the treatment benefit subscale.

PAGE 126

113 Table 4. Reliability of Health Belief Scales and Subscales Reliability of Scale from Time 1 to Time 2 (ANOVA) (n=10) Correlation Coefficient from Time 1 to Time 2 (n=10) Scale Name Subscale Name F Test value p-value on F test Patient Mean Square p-value Time Mean Square p-value r rsquared Perceived susceptibility --12.150.00040.00030.2339 0.86 0.74 Perceived treatment benefits --17.100.0001 <.00010.2259 0.91 0.83 Perceived severity --3.150.04930.03810.8321 0.65 0.43 Cues-to-action --12. 540.00040.00030.5911 0.88 0.78 Perceived barriersa --7.460.00290.00220.3452 0.79 0.62 Perceived barriers Medication 21.38 <.0001 <.00010.6618 0.93 0.87

PAGE 127

114 Table 4 (continued). Reliability of Scale from Time 1 to Time 2 (ANOVA) (n=10) Correlation Coefficient from Time 1 to Time 2 (n=10) Scale Name Subscale Name F Test value p-value on F test Patient Mean Square p-value Time Mean Square p-value r rsquared Perceived barriers Diet 5.350.00940.00710.5765 0.71 0.51 Perceived barriers Exercise 7.160.00340.00250.8777 0.81 0.66 Perceived barriers Monitoring 10.610.00080.00060.1965 0.85 0.72 Structural elements (understanding, social support) --10. 290.00090.00060.5203 0.84 0.70 Motivation --5.030.01160.00870.7577 0.73 0.54 Self-efficacy --12.14 0.00040.00030.7804 0.87 0.75 aPerceived Barriers subscales: The following subscales are calculated and represent one "question" in the Perceived Barriers scale: M edication (Q57), Diet (Q60), Exercise (Q63), Monitoring (Q66).

PAGE 128

115 Cronbach Alpha procedures were empl oyed to test the questions and subscale scores at each time period (see Table 5) Table 5. Cronbach Alpha Results for Health Belief Scales and Subscales at Times 1 and 2 Cronbach Alpha for Scale Time 1 (n=10) Cronbach Alpha for Scale Time 2 (n=10) Scale name Subscale name Raw Standard -ized Raw Standardized Perceived susceptibility --0.820.810.87 0.85 Perceived treatment benefits --0.720.770.86 0.88 Perceived severity --0.200.330.68 0.66 Cues-to-action --0.590.600.72 0.73 Perceived barriersa --0.450.540.56 0.68 Perceived barriers Medication 0.77b 0.68 b Perceived barriers Diet 0.87b 0.78 b Perceived barriers Exercise 0.62b 0.81 b Perceived barriers Monitoring 0.73b 0.54 b Structural elements (understanding, social support) --0.420.530.57 0.64

PAGE 129

116 Table 5 (continued). Cronbach Alpha for Scale Time 1 (n=10) Cronbach Alpha for Scale Time 2 (n=10) Scale name Subscale name Raw Standard -ized Raw Standardized Motivation ---0.040.060.55 0.62 Self-Efficacy --0.870.880.84 0.86 aPerceived Barriers subscales: The following subscales are calculated and represent one "question" in the Perceived Barriers scale: M edication (Q57), Diet (Q60), Exercise (Q63), Monitoring (Q66). bNot calculable due to lack of variance on several questions. In Table 4, p-value on the F-test for each subscale showed significance indicating that the difference between ti me one and two were due to individual differences rather than time differences. All time mean square p-values were not significant. Time two standardized Cr onbach alpha values ranged from 0.62 for motivation scale to 0.87 for perceived treatment benefits (see Table 5). The Cronbach alpha results for individual questions are in Appendix 11 Perceived self-efficacy measures a person’s own assessment of their ability to perform diabetes self-car e or engage in diabetes-specific health behavior. The 4-point Likert subscale scores range from 9 (low self-efficacy score) to 36 (highest score); the mean self -efficacy score for all participants was 30.31. The control and intensive groups mean at baseline was similar: 30.20

PAGE 130

117 compared to 30.41 reflecting moderatel y high self-efficacy scores before intervention. All nine questions of the subscale had high alpha values 0.85 or higher. The overall standardized Cronbach alpha score was 0.88 at time one, and 0.86 at time two. The perceived susceptibility subscale measures the person’s perception of his or her own risks associated with diabetes. The scale scores range from 9 to 36 (lowest to highest). The mean for all par ticipants at baseline was 27.94. Both groups had similar high mean susceptibi lity scores at baseline. The Cronbach alpha for each of the nine test items was 0.746 or higher. The overall standardized Cronbach alpha for the subsca le was. 0.81 at time one and 0.85 at time two. The perceived benefits subscale measures the diabetic’s belief in the effectiveness of the diabetes treatment regimen and their fa ith in their physician’s recommendations to reduce the threats pos ed by diabetes and its complications. The scale ranges from 9 (lowest score) to 36 (the highest). The mean for all participants at baseline reflected a high mean score of 36.32. The groups showed no pre-intervention differences The Cronbach alpha for the nine test questions were 0.746 or above. The ov erall standardized Cronbach alpha for the subscale was 0.77 at time one and 0.88 at time two. The severity subscale measures a per son’s perception of the burdens and consequences (e.g. pain, death, disability) associated with diabetes. The

PAGE 131

118 subscale ranges from 8 to 32 (lowest to highest score). The overall mean at baseline was 25.46. Both control and in tensive groups had similar high mean scores 25.52 and 25.40, respectively. T he Cronbach alphas for the eight test items were 0.542 or above. The overa ll standardized Cronbach alpha for the subscale was 0.33 at time one, and 0.66 at time two. The cues-to-action subscale assesses the person’s response to symptoms and modifying factors such as health messages that are likely to result in obtaining medical care or engaging in positive health behaviors. The subscale ranges 6 to 24 (low to high). The overall mean was 17.88 -both groups had similarly high mean scores at baseline: 18.23 for the control and 17.51 for the intensive group. The Cronbach alphas for t he six items were 0.66 or higher. The overall standardized Cronbach alpha for the subscale was 0.60 at time one, and 0.73 at time two. The structural element subscale asse sses the modifying factors such as a person’s locus of control, social s upport, and understanding of diabetes and the treatment regimen. The subscale ranges from 4 to 16 (lowest to highest). The overall mean was 10.88, with no significant pre-intervention differences. The Cronbach alphas for the four test items r anged from 0.33 to 0.74. The overall standardized Cronbach alpha subscale was 0.53 at time one and 0.64 at time two. The motivation subscale measures a person’s concern for his or her own

PAGE 132

119 health. The subscale ranges from 6 to 24 (low to high). The overall mean was high (18.01) as well as both control and intensive group, (17.84, 18.20, respectively). The Cronbach alphas for the six test items were 0.48 or above. The overall standardized Cronbach alpha was 0.62. The barriers subscale measures per ceived psychological and situational obstacles to overcome for a person to carry out the recommended diabetes regimen. The scale measures barriers to each pertinent diabetes treatment area, medications, diet, exercise and monitori ng. The scale ranges from 65 to 260 (lowest to highest). The mean for all parti cipants was 94.74. Both the control and intensive group had similar mean scores 93.41 compared with 96.15, respectively. The Cronbach alphas for the six test items (note that these questions have multiple parts) were 0. 60 or higher. The overall subscale standardized Cronbach alpha was 0.54 at time one and 0.68 at time two. The Diabetes Self-Care Practice Assessment was designed to measure diabetes self-care skills, knowledge, and adherence to each major diabetes regimen. The patient and di abetes educator who is res ponsible for the area of diabetes self-care education scores this scale. The health professional uses multiple tools such as 24-hour diet recall, food, blood glucose logs, and computerized glucometer memory data, (Appendix 5) written knowledge tests, and skill checks to rate individuals understanding and performance levels. The Diabetes Self-Care Practice Assessment (see Table 6) is used to

PAGE 133

120 assess patient’s ability to understand and carry out specific diabetes selfcare behaviors. The diabetes self-care asse ssment also helps ensure fidelity of program implementation and that educational objective s are met. Diabetes Educators are trained in the use of the instruments and scoring to ensure reliability and fidelity of implementation.

PAGE 134

121 Table 6. Diabetes Self-care Practice Assessment Date Date Date Date Objectives Objectives Learning and skill objectives Met (1)a Unmet (0)b Learning and skill objectives Met (1)a Unmet (0)b Diabetes overview understanding Able to state what diabetes is Able to state type of diabetes (Range 0-2) Subtotal Exercise and physical activity Able to state how exercise affects blood sugar Able to adjust food intake or medication for exercise Able to state precautions to take when exercising or when not to exercise Able to plan snacks for exercise when needed Able to follow personal exercise plan (Range 0-5) Subtotal

PAGE 135

122 Table 6 (continued). Learning and skill objectives Date Date Learning and skill objectives Date Date Objectives Objectives Met (1)a Unmet (0)b Met (1)a Unmet (0)b Nutrition Able to state reason for meal planning Eats meals/snacks at regularly scheduled times States reason for reaching or maintaining desirable weight Able to state rationale for eating less fat Able to identify fat in food Able to describe own personal meal plan Able to use meal plan Able to use meal plan when eating out Able to describe foods that are carbohydrates Medication Able to state medication type and dose Able to describe when it should be taken and its effects on blood glucose Able to describe side effects/ precautions (range 0-4) Subtotal (for insulin users only, add to above) Able to describe onset, peak and duration of insulin taken Able to determine when low blood sugars are more likely to occur

PAGE 136

123 Table 6 (continued). Learning and skill objectives Date Date Learning and skill objectives Date Date Objectives Objectives Met (1)a Unmet (0)b Met (1)a Unmet (0)b Able to use labels to determine which foods fit meal plan Able to explain benefits of fiber and list foods high in fiber (range 0-11) Subtotal Describe proper handling and how to store insulin Able to correct ly administer insulin (range 0-8) Subtotal Urine ketone testing (insulin users only) Able to describe w hen to test for ketone and able to take appropriate action for positive ketones (range 0-1) Subtotal

PAGE 137

124 Table 6 (continued). Learning and skill objectives Date Date Learning and skill objectives Date Date Objectives Objectives Met (1)a Unmet (0)b Met (1)a Unmet (0)b Behavioral change Able to set dia betic care goals Able to problem-solve and identify strategies to achieve goals (range 0-2) Subtotal Hypoglycemia Able to state what hypoglycemia is and possible symptoms Able to state correctly ways to prevent/and treat low BS States purpose of glucagon and has family/signific ant other able to give glucagon Blood glucose monitoring Demonstrates ability to test blood glucose Able to proper ly maintain meter and strips Able to state reason for blood glucose monitoring and record keeping Demonstrates via log that BS’s are recorded at times/frequency recommended for their diabetes type/situation Able to state normal blood sugar range and target range

PAGE 138

125 Table 6 (continued). Learning and skill objectives Date Date Learning and skill objectives Date Date Objectives Objectives Met (1)a Unmet (0)b Met (1)a Unmet (0)b Wears or carries diabetes identification (range 0-5) Subtotal Able to adjust food/medication/or activity based on results Able to obtain meter supplies Able to describe Hbg A1c or glycated hemoglobin test, its purpose and their personal value and goal (range 0-10) Subtotal Hyperglycemia Able to state what hyperglycemia is and possible symptoms Able to state courses of action (range 0-2) Subtotal

PAGE 139

126 Table 6 (continued). Learning and skill objectives Date Date Learning and skill objectives Date Date Objectives Objectives Met (1)a Unmet (0)b Met (1)a Unmet (0)b Complications States awareness of potential complications: Heart/symptoms Nerves/symptoms Eyes/symptoms Kidney/symptoms Circulation/peripheral vascular/symptoms Amputations/symptoms Sexual dysfunction/pregnancy Able to discuss how to prevent or delay complications (range 0-8) Subtotal Personal health Has regular eye exams (min yearly) Has regular dental care (min yearly) Has regular physical exams or doctor visits/medical care (min yearly) Has awareness of community and health professional resources (e.g. support groups, Endo’s, ADA, etc.) Able to inspect feet Does not go barefoot Able to trim toenails properly or goes regularly to podiatrist

PAGE 140

127 Table 6 (continued). Learning and skill objectives Date Date Learning and skill objectives Date Date Objectives Objectives Met (1)a Unmet (0)b Met (1)a Unmet (0)b Able to state effects of smoking on circulation Able to state effects of alcohol on BS (range 0-9) Subtotal TOTAL SCORE aIndicates patient met learning or skills objective. bIndicates patient was not able to met learning or skills objective

PAGE 141

128 The participant completes behavioral goals at week one. Progress toward goals will be assessed at 12 weeks. Partic ipants rate their progress on a scale of one to ten. Educational needs are r eassessed based on individual’s needs (Appendix 6). The ADA Diabetes Self-Care Ma nagement Training Program All participants received comprehensiv e diabetes self-care management education over the first 12-week trial per iod (for schedules s ee Appendix 6). The diabetes self-care management trai ning program is American Diabetes Association (ADA) approved and meets all national standards of structure, process, and outcome for diabetes educatio n. The ADA-approved diabetes selfcare management training consists of an interdisciplinary package of diabetes modalities comprising all aspects of diabetes management including social, medical, nutritional, and ex ercise management. A team of certified diabetes educators that includes a nurse, a diet itian and an exercise specialist is deployed. The aspects of care are prov ided to persons with diabetes through a case management system that individually assesses the needs of the patient or client. Education sessions focus on areas needing more attention. Each session has standardized curricula, objec tives, and evaluation. The program objectives are based on three critical components: 1. Optimal and intensive blood glucose monitoring and treatment; 2. Behavioral change, and

PAGE 142

129 3. Diabetes education which includes the person with diabetes, their families, and significant others. The ADA-approved curriculum and materials used in this study were adapted from Life with Diabetes: A Seri es of Teaching Outlines by the American Diabetes Association (1997b). The gui de was developed to assist health professionals in the education of patient s with diabetes mellitus in the required content areas. The guide also provides a supplementary section that addresses other diabetes information. Each outli ne includes a statement of purpose, prerequisites that should be known before participation in the session, materials needed, recommended teaching methods, a content outline, evaluation and documentation plan, and suggested r eadings related to each topic. The Diabetes Self-Care Practice Assessment was designed for diabetes educators to rate persons with diabetes on the understanding of diabetes selfcare, and then actual performance of se lf-care observed by the diabetes educator or obtained from glucometer memo ry data. This assessment based on the Living with Diabetes cu rriculum documents progress toward behavioral and educational goals and whether curriculum objectives are met (see Table 6). A variety of teaching strategies we re employed in the program. Their selection was based on an assessment of the patient's age, medical status, educational level, physical and emotional di sabilities, and readiness to learn. Teaching took place in the well-equipp ed classroom, the exercise room, or a

PAGE 143

130 counselor's office. Flip charts and pos ter size educational charts were used to illustrate key concepts during sessions. Each patient received a folder that included a patient educational manual, a m eal planning pamphlet, monitoring log books, and other handouts to supplement the core patient education materials. Portable VCRs, overhead projectors, blackboards and whiteboards, and food models are some of the educ ational tools used to enhance learning. The patient is asked personally, or by way of questionnai re, what they are most interested in learning and any perceived barriers to lear ning. The instructors incorporate this information into their lesson plans. For a listing of various handouts, tapes, posters, and book titles used, see Appendix 7. Printed materials are written at between the 6thand 8th-grade reading level. Pictogr ams, food models, and large print are used frequently for those with visi on problems. Skills are taught using a variety of techniques including verbal instruction, demonstration, problemsolving, and empowe rment training. In 1993, the national standards for diabetes self-management education programs were revised by an American Diabetes Association (ADA) Task Force (made up of various diabetes organizations and representatives) to reflect recent research and current health care trends. The purpose of the standards was to provide a guide for the establishment and maintenance of qua lity diabetes selfmanagement education. The term self-ma nagement training replaced previous traditional terms such as patient educat ion and self-management and is “to refer

PAGE 144

131 to the process whereby individuals learn to manage their diabetes” (ADA, 1998a). For a program to receive ADA re cognition, it must meet all of the national standards. Diabetes self-care training programs mu st meet stringent standards for structure, process, and outcome eval uation. The structure guidelines delineated the human and natural resour ces needed including the professional staff requirements, and space and budget requirements. The structural standards ensure that the curri cula meet specific criteria. The standards ensure that ADA-appr oved programs provide instruction in the content areas that are relevant to the pr ogram’s target diabetes population. The content areas included in the revised standards are: 1. Diabetes overview 2. Stress and psychosocial adjustment 3. Family involvement and social support 4. Nutrition 5. Exercise and activity 6. Medications 7. Monitoring and use of results 8. Relationships among nutrition, exercise, medication, and blood glucose levels 9. Prevention, detection, and treat ment of chronic complications 10. Prevention, detection, and treat ment of acute complications

PAGE 145

132 11. Foot, skin, and dental care 12. Behavioral change strategies, goal setting, risk factor reduction, and problem-solving 13. Benefits, risks, and management options for improving glucose control 14. Preconception care -pregnancy and gestational diabetes 15. Use of health care systems and community resources The standards acknowledge the unique ness of both the community as a whole and the individual with diabetes. The standards require a community needs assessment be done to define the target population and whether the program is designed to m eet the needs of this populat ion. In addition, an individual assessment for eac h participant in the program is required to identify personal educational needs. The individuali zed assessment takes into account a person’s age, disease processes, cult ure, and lifestyle. The assessment includes: relevant medical history, present health st atus, health services or resource utilization, risk factors, di abetes knowledge and skills, cultural influences, health beliefs and attit udes, health behavior and goals, support systems, barriers to learning, and so cioeconomic factors (see Appendix 4). Through this assessment, individual perceptions of susceptibility and seriousness of diabetes and its complicati ons can be determined. Pre-program

PAGE 146

133 assessment gathers information about the individual’s beliefs, attitudes, knowledge, and skill levels, as well as the individual’s environment and their social and physical situations to whic h the person must function. During the initial orientation session, the revised Health Belief Model Questionnaire and the Diabetes Needs Assessment are complet ed. Based on this information, individual program education plans and in tervention strategies are designed to focus on those perceptions that need to be changed and those skills and knowledge levels needing improvement. Barriers to learning and barriers to treatment modalities are a ssessed to help patients develop problem-solving and coping skills that overcome barriers. In the first education session, the diabet es educator attempts to provide information about diabetes, its pathophysiolog y, the risk factors for the disease, and who is susceptible to the disease. T he instructor also gives an overview of the benefits of various tr eatment modalities. Duri ng the needs assessment, the diabetes educators identify indivi duals who score poorly on perceived susceptibility, severity of diabetes or benefits, and attempt during the initial education to modify these beliefs. T he educator would emphasize that diabetes is a serious disease with shortand long-term complications. The educator would describe the various diabetes-rela ted complications such as nephropathy, retinopathy, and peripheral neuropathy. T he educator would also explain the relationship of hyperglycemia to other condi tions such as cardiovascular disease.

PAGE 147

134 Symptoms associated with high and low blood sugar are described. Abnormal blood sugars can act as cues to action. When persons with diabetes recognize the symptoms they can act to manage t heir blood glucose levels. Acute complications such as diabetes ketoac idosis (DKA) and hyperosmolar coma are described; the risk associated with these complications and actions to take to prevent their occurrence are described. Co sts of these complications in terms of illness, pain, financial bur dens from hospitalizations, increased medical care, disability or loss or work, and the benefits of preventio n are discussed. For those who have low scores on perceived benefits of care, the educator can emphasize the immediate and long-term rewards for controlling diabetes. Avoiding pain, financial expense, and other problems due to poor management are stressed. Benefits of increased sense of well being, feeling good, and having the energy to enjoy life are emphasized The ADA program recognizes the impor tance of identifying barriers to specific self-management behaviors. Du ring the assessment, the educator focuses on uncovering specific barrier s (e.g. financial cost, convenience, complexity of regimen, lack of suppor t) to each area of diabetes management. The nurse educator may focus on barriers to medical treatment; the dietitian may focus on barriers to following a meal plan; while the exercise specialist may work with the patient on barriers to exercise. For those identified as having low per ceived self-care abilities, focus is

PAGE 148

135 placed on developing skills and self-confiden ce. To foster self-confidence the diabetes educators instructs the patient on the correct se lf-care skills then allows the participants to return the demonstrati on. Use of glucose monitoring and log books act to reinforce the behavior but al so allows the diabet es educator to provide feedback and reinforcement to fac ilitate the participant ’s self-efficacy. The program not only focuse s on increasing knowledge of self-care activities, but provides time for practicing these skills ADA programs use empowerment such as goal setting techniques and contracting to enhance self-efficacy. A matrix of the educational sessions and the applicatio ns of the Expanded Health Belief Model concepts are illustrated in Table 7.

PAGE 149

136 Table 7. Matrix of the Educational Sessions and Applications of t he Expanded Health Belief Model Concepts Education/session/ content area Instructor Health Belief Model application Orientation Case Manager — CDE Baseline attitudes, beliefs, knowledge of diabetes self-care skills assessment, and adherence. (Hgb A1c) Overview of diabetes Certified Diabetes Educator (CDE) Nurse/Dietitian Seriousness of diabetes. Susceptibility to diabetes, and its complications. Benefits of treatment, and diabetes control including self-efficacy by patient able to identify chronic complications, symptom recognition, and self-care behaviors that will prevent complications

PAGE 150

137 Table 7 (continued). Education/session/ content area Instructor Health Belief Model application Medications CDE/Nurse Benefits of medications. Barriers to medication regimen. Strategies to overcome barriers of medication regimen: timing, action, and side-effects (for insulin users — skills are checked, proper methods demonstrated, and practiced to improve self-confidence. Self-efficacy enhanced through understanding the timing action, sideeffects, and administration of medications (insulin: demonstrate abilities, and proper administering of insulin).

PAGE 151

138 Table 7 (continued). Education/session/ content area Instructor Health Belief Model application Nutrition Dietitian Importance of nutrition in controlling blood sugar. Benefits of timing/scheduling of meals, carbohydrate consistency and balance of meals. Barriers to nutrition management identified and strategies discussed. Self-efficacy enhanced by participation in meal planning food records. Feedback and reinforcement acts to enhance and elicit family support and ensure an understanding of glycemic effects of food.

PAGE 152

139 Table 7 (continued). Education/session/ content area Instructor Health Belief Model application Behavior goal setting CDE/Dietitian Importance of behavior change. Barriers to behavioral change. Self-efficacy enhanced through small steps achieved toward goals, measured objectives, problem solving/coping skills to prevent relapse. Discuss personal incentives and rewards. Demonstrate commitment to change through written contract. Exercise CDE/Exercise Physiologist Enhance understanding of the effects of exercise on blood sugar control and general health. Benefits of exercise. Barriers of exercise discussed and strategies to overcome.

PAGE 153

140 Table 7 (continued). Education/session/ content area Instructor Health Belief Model application Enhanced self-efficacy through exercise practicum (practice on treadmill, stationary bike). Teach how to monitor pulse, heart rate (cues to action) and encourage family and social support. Self-behavioral goals development for exercise. Monitoring CDE/Nurse/ Exercise Physiologist/ Dietitian Importance/benefits of glucose monitoring and Hgb A1c testing. Barriers to monitoring discussed and strategies to overcome. Practicum to perform glucose monitoring and logging results to enhance self-efficacy. Blood sugar values/symptoms (cues to action).

PAGE 154

141 Table 7 (continued). Education/session/ content area Instructor Health Belief Model application Acute hypoglycemia/ hyperglycemia management Complications Chronic Complications/foot care CDE/Nurse Importance of prevention of complications. Benefits of self-management. Self-efficacy enhanced by patient able to demonstrate proper foot care, list ways to treat low/high blood sugars. Evaluate glucose monitoring log and enhance patient problem solving skills to raise and lower blood sugar values. Barriers discussed/strategies discussed. Self-behavioral goals developed. Recognize signs/symptoms of complications (cues to action). Discuss the tests and the assessments available for detecting and preventing complications.

PAGE 155

142 Table 7 (continued). Education/session/ content area Instructor Health Belief Model application Diet and exercise followup Barriers discussed for exercise/meal planning/food logs discussed/problem solving strategies for special situation discussed/exercise practicum/demonstrate skills learned, heart rate. Reinforcement by educators and cues to action. Stress/sick days rules disaster plan, special situations Progress of behavioral goals assessed for meal planing/exercise. Importance of stress management. Problem solving: travel, special situations, disaster, sick days. Reassessment of Hgb A1c, beliefs, attitudes, adherence. To ensure fidelity of impl ementation, all diabetes educators were certified diabetes educators with a minimum six months teaching experience using the Life with Diabetes curriculum. All di abetes educators had completed yearly

PAGE 156

143 diabetes self-care training compet encies and had documentation in their employee record files. The diabetes educ ators who participated in the research study were trained during staff meetings on the use and interpretation of scores obtained from the EHBMQ. Because the investigator and di abetes educators may represent “passionate workers” who want to see the ADA program succeed, efforts to minimize investigator bias were made. Instructors were randomly audio-taped to ensure equal treatment of participants. Taping of educational sessions also ensured that key diabetes content areas were being delivered uniformly. The diabetes self-care practice assessment co mpleted served to document that the objectives of the curriculum were bei ng met. Each participant completed a program satisfaction survey that eval uates educator performance and overall satisfaction with the program (see Appendix 7). A potential weakness of t he EHBMQ assessment is social desirability bias. Social desirability bias is where pat ients report levels of self-care that they want health care providers to believe they perform rather than actual performance. To offset th is potential bias, other measures of self-management were obtained and compared. Patients were asked to maintain diaries of selfcare behaviors (food records, blood glucose monitoring records). Using patients’ food records, glucose monitoring record s, exercise records, the diabetes educators rated patient’s understanding of diabetes and actual performance on

PAGE 157

144 the diabetes self-care practice assessment. This diabetes program is taught in small group and one-to-one counseling sessions that reduce the complexity duration of sessions, and amount of behavior change expected in each step through tailoring the sessions and behavioral goals to each person. Emphasis is placed on behavioral objectives that the client believes t hat he/she can achieve. The ultimate goal of diabetes education is to enable patients to achieve op timal control of their disease through personal involvement, motivation, and be tter understanding of diabetes and its effects on their health and well-being. Intensive Follow-up Care After completing ADA comprehensive se lf-care training, study participants were randomly assigned to standard follow-up care or intensive follow-up care. Standard follow-up care partici pants were mailed reminder letters to return to the Diabetes Care Institute 12 weeks from in itial HgbA1c value to obtain follow-up HgbA1c values. Behavioral goals st atus, and educational needs were reviewed with patients by certified diabetes educ ators and documented on the Behavioral Change Goals form (Appendix 7) and Educat ion Assessment fo rm (Appendix 8). (See also Appendix 9 for Division of Nu rsing Standards/Policies/Procedures.) Outcome data are entered into the com puter to summarize HgbA1c, education and behavioral goal outcomes. Pre-test/ post-test percent improvement and patient satisfaction survey results are al so collected to evaluate the education

PAGE 158

145 program. The intensive follow-up care group differs from standard follow-up care by utilizing the MiniMed continuous gluc ose sensor to provide a more comprehensive glucose pattern analysis to evaluate the effects of food, exercise, and medications on patient’s blood gl ucose levels. All participants were scheduled to return for their 12th week fo llow-up that consisted of HgbA1c and assessment of educational and behavioral goals. Those participants randomized to the intensive follow-up group also we re placed on the continuous glucose monitor system. The trained diabetes educator placed a MiniMed continuous sensor on the intensive follow-up pati ent and they were instructed in its use. The patient wore the device at home for 72 hours. Patients we re instructed to maintain their usual daily activities, medications, and check blood glucose levels four times a day while wearing the monitor. Key events were entered into the sensor such as eating, taking medication, exercise, usi ng a number code. Patients were given a logbook to record meals, activities, blood glucose test values, and medication dose and time. The MiniMed continuous glucose monitoring system (CGMS) is a Holter-style sensor that continuously and automatically records glucose levels (within 40-400 mg/dl range) in the patient’s subcutaneous tissue fluid, once every 5 minutes, for 72 hours. Information co llected by the CGMS was downloaded on a computer and reviewed by the diabetes care team. A clinical report identifying

PAGE 159

146 patterns of glucose excursions and graphs of the patient’s blood glucose levels (approximately 288 blood gluc ose readings per 24 hours) are generated. During the follow-up appointment with the diabetes team, the computer generated blood glucose data and the patient’s logbook were reviewed with the patient and therapy effects were explained (for po licies and procedures see Appendix 8). Both the intensive follow-up group and the standard follow-up groups returned at 24 weeks for review of behav ioral goals, complete post-intervention Health Belief Questionnaires, and HgbA1c determinations. Participants were able to discuss any diabetes self-care concerns with the certified di abetes educators. Determination of HgbA1c on three occa sions pre-education week one, posteducation week 12, and at week 24 follo w-up using a random assignment of follow-up care allowed examination of the effects of self-care training and collection of evidence to concerning t he efficacy of intensive follow-up care versus standard follow-up care. Documentation The diabetes educators received trai ning on proper documentation on the Diabetes Treatment Assessment and Educat ion Record, the Diabetes Self-care Practice Assessment, and the initial eval uation and follow-up evaluation forms. A program assistant filed ev aluation forms in the medical records. During postintervention, preevaluation Hgb A1c levels and initial EHBMS scores were not available for comparison, to minimize inve stigator bias. Post-intervention scores

PAGE 160

147 were recorded on the follow-up evaluation form and given to the program assistant to file in medical records. The diabetes educator documented each session completed by the patient on the diabetes assessment and education record. The date attended, the instru ctor who taught the session, and the content areas covered were recorded. As each objective of the Life with Diabetes curriculum was met, the diabetes educator documented the diabetes self-care practice assessment, the date, and the score obtained. Data Collection and Analysis Baseline clinical and demographic dat a were collected and documentation was kept in the participant’s medical reco rd. Descriptive statistics were used to describe the study population (see Table 8) Statistical analysis was performed using the SAS (1979) statisti cal program. Characterist ics of the study groups were compared for any pre-intervention differences. Chi-square analysis was used for nominal or ordinal variables such as gender, race, and education. Ttests were used for interval or ra tio variables, such as pre-Hgb. A1c, and EHBMQ scale scores. Tukey or Bonferroni pos t-hoc procedures were to be used as appropriate if differences were found. An analysis of variance (ANOVA) for repeated measures tested whether the groups differed in mean Hgb. A1c over time. The “between” variable was the group variable (control vs. intervention group) and the “within” factors ar e the various occasions Hgb. A1c was measured. If an overall difference was found between the mean Hgb. A1c over time, Tukey

PAGE 161

148 or Bonferroni post-hoc procedures were to be applied. Tukey procedures are preferred if the sphericit y assumption is met. The F statistic obtained in the repeated measure analysis was correlated using the average of the Huynh and Feldt and Greenhouse-Geisser epsilon (S) esti mators to maintain alpha close to the level of significance. Table 8. Characteristics of Study Populatio n Before Education Intervention Characteristic Female (%) Race (%): White Anglo-Amer ican or of European heritage Black African-American Hispanic Other-Pacific Islander Native Americans Asian Education (%): Did not complete or less than high school High school graduate or higher Age (yr.): Hgb A1c (%) Years since diabetes diagnosis (%) Less than 1yr 1 yr-5yr Over 5 yr. Physician Type (%): Endo Other

PAGE 162

149 Table 8 (continued). Characteristic Treatment Type (%): Insulin, or al medications, diet-controlle d Insulin and oral diabetes medications Expanded Health belief Model Subscale Mean Scores Perceived Susceptibility Score Perceived Severity Score Perceived Benefits Score Perceived Barriers Score Perceived Self-Efficacy Score Health motivation Score Cues to Action Score Structural Elements Score The use of random assignment to the intensive follow-up or control group strengthens the design’s ability to control syst ematic error such as selection bias. The repeated measure design reduces the within group variability among participants due to individual differences such as dur ation of disease, severity of disease, motivation, education level and social background.

PAGE 163

150 Summary In summary, an ADA-recognized comprehensive diabetes program incorporates several components of the Expanded Health Belief Model to enhance patient adherence to the diabetes self-care regimen and thereby improve glycemic control. The intens ive follow-up intervention is designed to evaluate blood glucose patterns and determine the effects of diabetes therapy on blood glucose. If persons receiving intens ive follow-up achieve significantly better outcomes, a new standard for disease m anagement will have received important evidence. DSME leads to lower Hgb A1 c levels by increasing the patient’s understanding of diabetes and its treatment, and by improving attitudes and beliefs that foster adherence to the diabetes regimen. The pr ogram fosters selfefficacy by providing reinforcement and per formance of self-care skills. Patient logs are kept of diet, exercise, gl ucose monitoring and medication schedules, and reviewed with instructors for feedba ck and reinforcement. Families and significant others are encour aged to participate in the sessions. Patients are encouraged to set self-management goals with guidance from the educator.

PAGE 164

151 Chapter Four: Results Introduction In this chapter, the population characte ristics, comparisons between those who completed the study with those droppe d from the study, an d the results of the analysis of a randomized repeated measure statistica l design are presented.

PAGE 165

152 Participant Characteristics Overall, 159 participants were random ized, 80 to the standard follow-up group (SFG) and 79 to the intensive fo llow-up group (IFG) after signing the consent form and blocking for insulin us e. A random number was assigned and group assignment was unknown to educat ors and participants. To reduce selection bias, a program assistant not involved in the assessment or education of participants, sent out follow-up notices and entered HgbA1c and behavioral scores into the outcome data files. Base line data were collected during the initial visit and recorded on the diabetes assessment and education record and initial evaluation form. Participants also comple ted the Revised Diabe tes Health Belief Model Questionnaire during this initial a ssessment and evaluatio n. All data were stored confidentially in t he patient’s medical record and computer databases. Characteristics of the study population were anal yzed for pre-intervention differences (see Tables 9 and 10). The two groups were compared on the categorical covariates using chi-squar e analysis and the continuous variables using T-tests. The groups were not si gnificantly different at baseline for the categorical variables: treatm ent type, ethnicity, education level, type of physician, and duration of diabetes. The control and the intensive groups had similar distributions for the four treatment types : diet; pills; pills and insulin and insulin only indicating the randomization blocking for insulin led to a similar distribution. At baseline, the gender covariate showed that more women (N=88) participated

PAGE 166

153 in the study then men (N=71). More women than men were randomized to the control than the intensive group (37 wom en in the control compared to 43 men, and 51 women were assigned to intensiv e group compared to 28 men, chisquare test value was 5.3902, p-value 0.0203).

PAGE 167

154 Table 9. Baseline Characteristics by Treatm ent Group: Categorical Covariates Group Control group Intensive group Level N % N % Pvaluea Total N % of Total All 80100.00%79100.00%--159 100.00% Treatment type Diet 810.00%911.39%17 10.69% Pills 5062.50%4455.70%94 59.12% Pills and insulin 1316.25%1822.78%31 19.50% Insulin 911.25%78.86%16 10.06% Unknown 00.00%11.27% 0.6461 1 0.63% Sex Female 3746.25%5164.56%88 55.35% Male 4353.75%2835.44% 0.0203 71 44.65% Ethnicity White 4860.00%5063.29%98 61.64% Black 2025.00%1417.72%34 21.38% Hispanic 810.00%1316.46%21 13.21% Other 22.50%11.27%3 1.85% Unknown 22.50%11.27% 0.5661 3 1.89%

PAGE 168

155 Table 9 (continued). Group Control group Intensive group Level N % N % Pvaluea Total N % of Total Highest level of education Grade School 11.25%45.06%5 3.14% High School 2733.75%3037.97%57 35.85% College 4758.75%4151.90%88 55.35% Unknown 56.25%45.06% 0.4804 9 5.66% Type of doctor Endocrinologist 2632.50%3037.97%56 35.22% Primary care 5366.25%4962.03%102 64.15% Unknown 11.25%00.00% 0.4876 1 0.63% Duration of diabetes <1 year 4252.50%3848.10%80 50.31% 1-5 years 56.25%911.39%14 8.81% >= 5 years 3240.00%2734.18%59 37.11% Unknown 11.25%56.33% 0.2189 6 3.77% aChi-squared test.

PAGE 169

156 Table 10. Baseline Age, Health Belief Scores and Hemoglobin (HgbA1c) by Group. Group Control IntensiveAll T-test Pvaluea All N 8079159--Age N 7574149 Mean 53.1352.3552.74 St. dev. 10.7111.3210.99 0.6656 Self-efficacy score N 7975154 Mean 30.230.4130.31 St. dev. 4.94.774.82 0.7872 Susceptibility score N 7975154 Mean 28.0527.8127.94 St. dev. 5.775.485.61 0.7941 Benefits score N 7975154 Mean 36.0936.5736.32 St. dev. 4.273.754.02 0.4559 Severity score N 7975154 Mean 25.5225.425.46 St. dev. 4.113.974.03 0.8554

PAGE 170

157 Table 10 (continued). Group Control IntensiveAll T-test Pvaluea Cue-to-action score N 7975154 Mean 18.2317.5117.88 St. dev. 4.163.743.97 0.2609 Structural score N 7875153 Mean 10.8110.9510.88 St. dev. 3.43.743.56 0.8102 Motivation score N 7975154 Mean 17.8418.218.01 St. dev. 2.363.013.14 0.4732 Barriers score N 7975154 Mean 17.8418.218.01 St. dev. 3.263.013.14 0.4802 HgbA1c N 8079159 Mean 8.58.58.5 St. dev. 1.31.31.3 0.9056 aT-test assumes equal variance between groups.

PAGE 171

158 T-test procedures applied to the cont inuous variables age, HgbA1c, and Health Belief subscale sco res at baseline showed no sign ificant pre-intervention differences (see Table 10). Box-plots and stem-leaf plots for all continuous variables are in Appendix 10. The cont inuous variables for the study sample were normally distributed except for H gbA1c, which was skewed. This skewness was due to the exclusion criteria for HgbA 1c (participant’s HgbA1c must be 7 or greater to participate). The mean age of the study par ticipants was 52.74 years. At baseline the two groups were similar. The control mean age of 53.13 years compared to the intensive group mean age of 52.35 years (p=0.6656, ns). The mean HgbA1c value for all subjects at baseline was 8.5%. The two groups had similar means at baseline, the control mean HgbA1c value was 8.49% compared to the intensive group mean HgbA1c value of 8.52% (p=0.9056, ns). The two groups had similar means for t he eight dimensions of the EHBMQ variables (self-efficacy, susceptibility, se verity, treatment benefits, cues to action, structural elements, motivation, and ba rriers). T-tests showed similar group means for the control and intensive gr oup at baseline (see Table 10). The univariate plots for the c ontinuous variables had norma l distributions (Appendix 10). However, treatment benefit showed for both groups a similarly high mean score trend at baseline. The box plot was evenly distributed despite being skewed toward the upper range.

PAGE 172

159 Drop Out versus Completers Overall, 82 patients completed all phas es of the study, and 77 dropped out or did not have complete data for the pr imary variable, HgbA1c or the secondary EHBMQ variables at week 24. These i ndividuals are referred to as “dropouts,” although they may have provided incomplete data that were kept for the analysis. Thirty-seven participants assigned to t he control group compared to 45 in the intensive follow-up group provided complete data for the study (see Figure 6). Overall, those who completed the study were not different than those who dropped out (see Table 11 and Table 12). The only difference that was statistically significant was for age, in that the dropouts (m ean age 49.89) were slightly younger than those who completed the study (mean age 55.49).

PAGE 173

160 Figure 6. Loss to Follow-Up

PAGE 174

161 Table 11. Characteristics of Completers vs. Dropouts: Categorical Covariates Complete Dropped Out Level N % of Completers N % of Dropouts N % of Total Pvaluea All 82100.00%77100.00%159100.00% --Group Control 3745.12%4355.84%8050.31% Intensive 4554.88%3444.16%7949.69% 0.1766 Treatment type Diet 910.98%810.39%1710.69% Pills 4959.76%4558.44%9459.12% Pills and insulin 1417.07%1722.08%3119.50% Insulin 1012.20%67.79%1610.06% Unknown 00.00%11.30%10.63% 0.6691 Sex Female 4858.54%4051.95%8855.35% Male 3441.46%3748.05%7144.65% 0.4036 Ethnicity White 5364.63%4558.44%9861.64% Black 1923.17%1519.48%3421.38% Hispanic 910.98%1215.58%2113.21% Other 11.22%22.60%31.89% Unknown 00.00%33.90%31.89% 0.3158

PAGE 175

162 Table 11 (continued). Complete Dropped Out Level N % of Completers N % of Dropouts N % of Total Pvaluea Highest level of education Grade School 11.22%45.19%53.14% High School 3441.46%2329.87%5735.85% College 4453.66%4457.14%8855.35% Unknown 33.66%67.79%95.66% 0.1894 Type of doctor Endocrinologist 3036.59%2633.77%5635.22% Primary care 5263.41%5064.94%10264.15% Unknown 00.00%11.30%10.63% 0.5574 Duration of diabetes <1 year 4048.78%4051.95%8050.31% 1-5 years 89.76%67.79%148.81% >= 5 years 3340.24%2633.77%5937.11% Unknown 11.22%56.49%63.77% 0.3044 aChi-squared test.

PAGE 176

163 Table 12. Characteristics of Completers vs. Dropout s: Continuous Covariates at Baseline Completed Dropped Out All T-test Pvaluea All N 8277159--Age N 7673149 Mean 55.4949.8952.74 St. dev. 10.0411.2710.99 0.0017 Self-efficacy score N 8272154 Mean 30.5430.0430.31 St. dev. 4.535.154.82 0.5268 Susceptibility score N 8272154 Mean 28.0427.8227.94 St. dev. 5.315.985.61 0.8116 Benefits score N 8272154 Mean 36.7235.8836.32 St. dev. 3.494.534.02 0.1939 Severity score N 8272154 Mean 25.6525.2525.46 St. dev. 3.684.424.03 0.5444

PAGE 177

164 Table 12 (continued). Completed Dropped Out All T-test Pvaluea Cue-to-action score N 8272154 Mean 18.1117.6117.88 St. dev. 43.953.97 0.4383 Structural score N 8271153 Mean 11.1210.5910.88 St. dev. 3.933.083.56 0.3598 Motivation score N 8272154 Mean 18.217.8118.01 St. dev. 3.592.543.14 0.4441 Barriers score N 8272154 Mean 94.0295.5694.74 St. dev. 23.0625.1523.99 0.6941 HgbA1c N 8277159 Mean 8.58.58.5 St. dev. 1.31.21.3 0.9859 aT-test assumes equal variance between groups.

PAGE 178

165 The drop out rate for the contro l group was 53.75% compared to the 43.04% for the intensive follo w-up group. The overall dist ributions of completers compared to dropouts on the categorical covariates treatment type, sex, ethnicity, education, physician type, di sease duration were not significantly different. ( value >0.05). There were no significant differences between the completers compared to the dropout s for the continuous health belief and HgbA1c covariates. Characteristics of completers compared with dropouts by treatment group are described in Table 13. There were no significant differences found for the categorical covariates in the control group. Fe wer diet-controlled patients in the standard fo llow-up group completed the study (8.11%) control compared with intensive follow-up group (13.33%). Participants treated with insulin had different participation rates. (SFG, 7/37, 18.92%; IFG 3/45, 6.67%) Those completing the study were evenly distributed between groups for treatment type. (P value = 0.40 in Table 16). More patients followed by primary care physicians (32/43, 74.41%) compared to endocrinologist dropped out of the control group ( 10/43, 23.26%). In the intensive follow-up group (18/34) 52.94% dr opped out from subjects referred by primary care physicians compared to endocrinologist referred (16/34) 47%. Those completing the study had similar dist ribution for physician type. The followup groups had no significant differenc es (p value = 0.26, ns). Patients dropping out of the control had diabetes of shorter duration (less

PAGE 179

166 than 1 year 25/43, 58.14%; 1-5 year, 4/ 43, 9.30%) compared with patients with diabetes over 5 years (13/43. 30.23%)? In the intensive group (44.12%, 15/34 < 1yr, 2/34, 5.88% for 1-5 year durat ion) more dropped out had diabetes for a shorter duration while only 38.24% (13/ 34) dropped out who had diabetes over 5 years. Dropouts were slightly younger (mean age 49.89) than those completing the study (mean age 55.49). The mean ages of those completing the study were similar (SFG mean 55.37; IFG m ean 55.59) with younger participants dropping out in both groups. Time commitm ents may be partially responsible for this trend secondary to hours the Diabetes Care Institute was open. Office hours were offered from 7:30 AM until 6:00 PM. More diet-controlled patients dropped out in the control group (5/43, 11.63% compared to the IFG (3/3 4 8.82%). Eight out of forty-five participants assigned to an intensive follow-up group treated with diet and pills completed the study compared six out of thirty-seven in the standard follow-up group. More women completed the study (15 / 37, 40.54%) than dropped out (22/43.51.16%). More African-Americans completed the study in the intensive group (11/43, 25.58%) than those assigned the control (9 /37, 24.32%). Hispanics dropped out more in the intensive group (9/34, 26. 47%) than the control (3/43. 7%). More college level participants comp leted the study in the intensive follow-up group (24/45, 53.33%) compared to the control. (20/37, 54.05%). A

PAGE 180

167 similar distribution was found for those co mpleted the study for high school level or below 43.23% for the control and 42. 22% for the intensive group. The difference between groups may be due to the unknown levels, two in the IFG and 1 in the control. (P value = 0.03). The small cell numbers may effect the interpretation of the p value but there s eems to be some potential for attrition bias. The main study continuous variabl es did not seem to be affected by the attrition.

PAGE 181

168 Table 13. Characteristics of Completers vs. Dr opouts by Treatment Group: Categorical Covariates Control Group Complete Dropped Out Level N % of Completers N % of Dropouts Pvaluea All 37100.00%43100.00%--Treatment type Diet 38.11%511.63% Pills 2156.76%2967.44% Pills and insulin 616.22%716.28% Insulin 718.92%24.65% Unknown 00.00%00.00% 0.2395 Sex Female 1540.54%2251.16% Male 2259.46%2148.84% 0.3421 Ethnicity White 2362.16%2558.14% Black 924.32%1125.58% Hispanic 513.51%36.98% Other 00.00%24.65% Unknown 00.00%24.65% 0.3597

PAGE 182

169 Table 13 (continued). Control Group Complete Dropped Out Level N % of Completers N % of Dropouts Pvaluea Highest level of education Grade School 12.70%00.00% High School 1540.54%1227.91% College 2054.05%2762.79% Unknown 12.70%49.30% 0.2901 Type of doctor Endocrinologist 1643.24%1023.26% Primary care 2156.76%3274.42% Unknown 00.00%12.33% 0.1199 Duration of diabetes <1 year 1745.95%2558.14% 1-5 years 12.70%49.30% >= 5 years 1951.35%1330.23% Unknown 00.00%12.33% 0.1698

PAGE 183

170 Table 13 (continued). Intensive Group Complete Dropped Out Level N % of Completers N % of Dropouts Pvaluea All 45100.00%34100.00%--Treatment type Diet 613.33%38.82% Pills 2862.22%1647.06% Pills and insulin 817.78%1029.41% Insulin 36.67%411.76% Unknown 00.00%12.94% 0.3812 Sex Female 3373.33%1852.94% Male 1226.67%1647.06% 0.0606 Ethnicity White 3066.67%2058.82% Black 1022.22%411.76% Hispanic 48.89%926.47% Other 12.22%00.00% Unknown 00.00%12.94% 0.1307

PAGE 184

171 Table 13 (continued). Intensive Group Complete Dropped Out Level N % of Completers N % of Dropouts Pvaluea Highest level of education Grade School 00.00%411.76% High School 1942.22%1132.35% College 2453.33%1750.00% Unknown 24.44%25.88% 0.1160 Type of doctor Endocrinologist 1431.11%1647.06% Primary care 3168.89%1852.94% Unknown 00.00%00.00% 0.1481 Duration of diabetes <1 year 2351.11%1544.12% 1-5 years 715.56%25.88% >= 5 years 1431.11%1338.24% Unknown 12.22%411.76% 0.1822 aChi-squared test.

PAGE 185

172 Table 14. Characteristics of Completers vs. Dropouts by Treatment Group: Continuous Covariates Control Completed Dropped Out T-test Pvaluea All N 3743--Age N 3540 Mean 55.3751.18 St. dev. 10.9610.23 0.0907 Self-efficacy score N 3742 Mean 29.9230.45 St. dev. 4.495.28 0.6323 Susceptibility score N 3742 Mean 29.1127.12 St. dev. 5.196.15 0.1272 Benefits score N 3742 Mean 36.7335.52 St. dev. 3.824.6 0.2122 Severity score N 3742 Mean 26.6524.52 St. dev. 3.114.63 0.0209

PAGE 186

173 Table 14 (continued). Control Completed Dropped Out T-test Pvaluea Cue-to-action score N 3742 Mean 18.7017.81 St. dev. 3.774.48 0.3446 Structural score N 3741 Mean 11.3810.29 St. dev. 4.292.27 0.1609 Motivation score N 3742 Mean 18.2717.45 St. dev. 3.832.65 0.2691 Barriers score N 3730 Mean 91.3095.97 St. dev. 18.9227.22 0.4206 HgbA1c N 3737 Mean 8.68.6 St. dev. 1.31.2 0.5326

PAGE 187

174 Table 14 (continued). Intensive Completed Dropped Out T-test Pvaluea All N 4534--Age N 4133 Mean 55.5948.33 St. dev. 9.3212.40 0.0054 Self-efficacy score N 4530 Mean 31.0429.47 St. dev. 4.564.99 0.1618 Susceptibility score N 4530 Mean 27.1628.80 St. dev. 5.35.69 0.2050 Benefits score N 4530 Mean 36.7136.37 St. dev. 3.234.46 0.6992 Severity score N 4530 Mean 24.8226.27 St. dev. 3.933.95 0.1237

PAGE 188

175 Table 14 (continued). Intensive Completed Dropped Out T-test Pvaluea Cue-to-action score N 4530 Mean 17.6217.33 St. dev. 4.153.09 0.7458 Structural score N 4530 Mean 10.9111.00 St. dev. 3.653.93 0.9204 Motivation score N 4530 Mean 18.1318.3 St. dev. 3.422.34 0.8163 Barriers score N 4530 Mean 96.2795.97 St. dev. 25.9727.22 0.9618 HgbA1c N 4534 Mean 8.48.6 St. dev. 1.41.2 0.5412 aT-test assumes equal variance between groups.

PAGE 189

176 To examine the effect on the di fferences between treatment groups imposed by the dropouts, the baseline characteristics were reanalyzed using only those 82 individuals for whom there were complete data. These results are shown in Table 15 and Table 16.

PAGE 190

177 Table 15. Baseline Characteristics by Treatment Group for Those with Complete Follow-up: Categorical Covariates Group Control group Intensive group Level N % N % Pvaluea Total N % of Total All 37100.00%45100.00% 82 100.00% Treatment type Diet 38.11%613.33%9 10.98% Pills 2156.76%2862.22%49 59.76% Pills and insulin 616.22%817.78%14 17.07% Insulin 718.92%36.67%10 12.20% Unknown 00.00%00.00% 0.3713 0 58.54% Sex Female 1540.54%3373.33%48 41.46% Male 2259.46%1226.67% 0.0027 34 64.63% Ethnicity White 2362.16%3066.67%53 23.17% Black 924.32%1022.22%19 10.98% Hispanic 513.51%48.89%9 1.22% Other 00.00%12.22%1 1.22% Unknown 00.00%00.00% 0.7243 0 0.00%

PAGE 191

178 Table 15 (continued). Group Control group Intensive group Level N % N % Pvaluea Total N % of Total Highest level of education Grade School 12.70%00.00%1 1.22% High School 1540.54%1942.22%34 41.46% College 2054.05%2453.33%44 53.66% Unknown 12.70%24.44% 0.7054 3 3.66% Type of doctor Endocrinologist 1643.24%1431.11%30 36.59% Primary care 2156.76%3168.89%52 63.41% Unknown 00.00%00.00% 0.2564 0 0.00% Duration of diabetes <1 year 1745.95%2351.11%40 48.78% 1-5 years 12.70%715.56%8 9.76% >= 5 years 1951.35%1431.11%33 40.24% Unknown 00.00%12.22% 0.0921 1 1.22% aChi-squared test.

PAGE 192

179 Table 16. Baseline Characteristics by Treatment Group for Those with Complete Follow-up: Continuous Covariates Group Control IntensiveAll T-test Pvaluea All N 374582--Age N 354176 Mean 55.3755.5955.49 St. dev. 10.969.3210.04 0.9270 Self-efficacy score N 374582 Mean 29.9231.0430.54 St. dev. 4.494.564.53 0.2658 Susceptibility score N 374582 Mean 29.1127.1628.04 St. dev. 5.195.35.31 0.0976 Benefits score N 374582 Mean 36.7336.7136.72 St. dev. 3.823.233.49 0.9810 Severity score N 374582 Mean 26.6524.8225.65 St. dev. 3.113.933.68 0.0242

PAGE 193

180 Table 16 (continued). Group Control IntensiveAll T-test Pvaluea Cue-to-action score N 374582 Mean 18.717.6218.11 St. dev. 3.774.154 0.2254 Structural score N 374582 Mean 11.3810.9111.12 St. dev. 4.293.653.93 0.5955 Motivation score N 374582 Mean 18.2718.1318.2 St. dev. 3.833.423.59 0.6304 Barriers score N 374582 Mean 91.396.22794.02 St. dev. 18.9225.9723.06 0.3345 HgbA1c N 374582 Mean 8.68.48.5 St. dev. 1.31.41.3 0.6184 aT-test assumes equal variance between groups.

PAGE 194

181 Of those 82 individuals on which data were complete, there were significant gender differences between t he control group and the intensive group (more women in the intensive group, p=0. 0027). Education (higher education in the intensive group, p=0.0340) and duration of diabetes (shorter duration in the intensive group, p=0.0020) were different. Also, baseline severity scores were higher in the control group (26.65) com pared to the intensive group (24.82, p=0.0242).

PAGE 195

182 Research Question 1 results The primary objective of this study was to determine if persons with type 2 diabetes participating in an ADA diabet es self-management education randomly assigned to an intensive follow-up group em ploying the continuous sensor would achieve lower mean HgbA1c values i ndicating better metabolic control than controls receiving standard control. A randomized, repeated measure design was employed to test the null hypothesis t hat there is no significant difference in mean HgbA1c values over time (week 1, week 2, and week 24), for participants of a ADA DSME randomly assigned to the intensive follow-up group compared to the standard follow-up group. Twelve week s from their initial assessment of HgbA1c and after completing the di abetes self-management education patients were scheduled to return for follow-up. Bo th groups were first informed by letter from the program assist ant the date their HgbA1c and behavioral and outcome follow-up was scheduled. The letter was sent well in advance of the scheduled time and place of their appointment. Those with a scheduling conflict were instructed to call immediately so that t heir schedule needs could be met. Those who were randomly assigned to the intensive group also received a letter that they had been assigned to the intensive group and given the number to schedule their appointment for follow-up employing the sensor. A complete description of the MiniMed continuous glucose monitor s ensor, and the time and effort required, were described in the letter and discu ssed on the phone on scheduling. The

PAGE 196

183 MiniMed CGMS used in this study was l oaned to the center as a courtesy. The principle investigator, t he diabetes educators and the Di abetes Care Institute had no affiliated with MiniMed and received no pa yment or incentives to use their product in this research. MiniMed did not solicit the center to conduct this research. MniMed had no involvement or control of the study .They implied no obligations for their donati on of the sensors used in th is research. The sensor was previously discussed at the initia l enrollment prior and during the informed consent process. The primary respons e variable, the HgbA1c value, was measured using the same capillary NGSP ce rtified micro-optical detection assay, A1cNow at week 1, week 12 and week 24 (see Figure 7). The mean HgbA1c values for the control were compared to the intensive fo llow-up group using a SAS ANOVA GLM procedure for repeated measures (see Appendix12 for ANOVA Summaries and SAS output). T he results used Tukey and Bonferroni adjustments for multiple comparis ons. The data analysis conducted found no significant statistical difference between the groups (see Table 17). The first arm of intervention reflec ted the largest drop of the mean HgbA1c. The overall mean HgbA1c value of 8.5% dropped to 7.1% at week 12. The mean baseline HgbA1c values for t he control (8.6, N=80) and intensive group (8.5,N=79 ) showed no pre-intervention differences, the F value was 0.16, p= 0.6880. At week twelve the standar d follow-up group mean HgbA1c was not significantly different than the intensive gr oup ( F> 0.00, p= 0.9776. At twenty-

PAGE 197

184 four weeks, HgbA1c means reflected no significant differences between the control and intensive follow-up groups (F > 0.02, p= 0.8798). Tukey and Bonferroni corrections showed no signifi cant differences between mean HgbA1c levels in each group at week 1, week 12, or week 24.

PAGE 198

185 Table 17. Results from Repeated Measures ANOVA A nalysis of Hemoglobin at Weeks 1, 12, and 24, Control vs. Intensive. Week 1 Week 12 Successc at Week 12, Y/N (%) Week 24 Successc at Week 24, Y/N (%) Control N 806227/35 (43.55%) 50 29/21 (54.72%) Mean 8.497.18--6.93 --Std Dev 1.261.29--1.40 --Intensive N 796033/27 (55.00%) 53 29/24 (58.00%) Mean 8.527.16--7.00 --StDev 1.281.40--1.34 --F Valuea 0.160.00--0.02 --p-valuea Unadjusted 0.68800.9776--0.8798 --Tukey Groupingb NSD NSD --NSD --Bonferroni Groupingb NSD NSD --NSD --Note: Considered in ANOVA Cont rol, N=48; Intensive, N=53 aRepeated measures ANOVA. bNo significant differences between groups. cSuccess = Hemoglobin at 7% or less.

PAGE 199

186 Table 17 also includes the proporti on of participants who successfully lowered their HgbA1c to the ADA standard of 7% or less. At we ek 12, 55% of the intensive group and 44% of t he control members were successful. At week 24, 58% of the intensive and 55% of the control group members were successful. A 1.0% lowering of HgbA1c is clinically significant to lower complications and costs of care. Overall, the m ean HgbA1c value dropped 1.6%. Figure 7. Mean Hemoglobin at Wee ks 1, 12, and 24 by Group Mean Hemoglobin at Weeks 1, 12, and 24 By Group0 1 2 3 4 5 6 7 8 9 Week 1Week 12Week 24 Week of MeasurementHemoglobin Level Control Intensive

PAGE 200

187 An annual follow-up visit was conducted wit h 51 of the participants. Table 18 shows their Hgb1ac measur es at this follow-up. Table 18. Hemoglobin (Hgb1ac) at 1-year Follow-up Group Control IntensiveAll T-test Pvaluea Hgb1ac at annual follow-up N 272451--Mean 6.787.156.95 St. dev. 1.701.411.28 0.3028 aT-test assumes equal variance between groups. .

PAGE 201

188 Research Questions 2-11 results A profile analysis was employed ut ilizing a repeated ANOVA to compare the mean Health Belief s ubscale scores for both gr oups for the eight major constructs of the expanded health belief model (self-ef ficacy, treatment benefits, susceptibility, severity, cues to action, structural elements, motivation and barriers preand post-intervention). A comparison of pre-intervention mean scores illustrates that the two groups were similar at baseline (see Figure 8). At week 24, study participants complet ed the revised expanded Health Belief Questionnaire for the second time to measure post-intervention beliefs (see Figure 9). Both groups showed a simila r response pattern. As previously reported there were no pre-interventi on differences. Figure 10 shows the differences at each time by group, and Figure 11 shows the overall profiles at weeks 1 and 24 by group. There were no significant EHBMQ subscale mean score differences found between the c ontrol group and the in tensive group at week 24, with the exception of perceiv ed severity scores. The mean for the control was higher (27.05) than for the int ensive follow-up group (p value 0.03). Tukey’s Studentized range test confirmed differences between the groups at week 24. When applying the GLM procedure to the mean score differences from week1 to week 24, no differences were found between groups. The persons in the control group with lower severity sco res tended to drop out more which may have inflated the severity mean scor es. The GLM procedure summary is

PAGE 202

189 summarized in Appendix 12. Figure 8. Baseline Health Belief Profiles by Group 0 20 40 60 80 100 120S e lf Eff i ca c y Susc e p ti bil i ty B e n e fi ts S e v e rit y C u e -t o -A c tio n Stru c tura l El e m e nt s M o ti v at io n Bar r ie r sHealth Belief SubscaleScore Intervention Control

PAGE 203

190 Figure 9. Week 24 Health Belief Profiles by Group 0 10 20 30 40 50 60 70 80 90 100S e lf Eff i ca c y Susc e p ti bil i ty B e n e fi ts S e v e rit y C u e -t o -A c tio n Stru c tura l El e m e nt s M o ti v at io n Bar r ie r sHealth Belief SubscaleScore Intervention Control

PAGE 204

191 Figure 10. Health Belief Profiles in Differences from Baseline to Week 24 by Group -10 -8 -6 -4 -2 0 2 4 6 8 10S el f -E ff ic a cy S u s c eptib il ity B e nefits Severi t y C u e -t o -A c tio n Stru c tura l El e m e nt s M o ti v at io n Bar r ie r sHealth Belief SubscaleScore Intervention Control

PAGE 205

192 Figure 11. Health Belief Profiles at Weeks 1 and 24 by Group 0 20 40 60 80 100 120Self-Efficacy Su sceptibility Benefits Severity Cue-to-Action Structural Elemen t s Moti vation BarriersHealth Belief SubscaleScore Intervention Wk1 Control Wk1 Intervention Wk24 Control Wk24 There was no significant difference bet ween groups from week 1 to week 24 in mean barrier scores. The base line mean for the control group went down from 96.15 to 90.60 at 24 weeks. T he intensive follow-up group mean remained fairly constant going from 93.4 to 94.1. The overall me an for barriers was 94.74. The scale ranged from lowest score of 65 to highest score 260. The mean score for all participants in both groups were lo w at baseline and after intervention.

PAGE 206

193 The mean differences from week 1 to week 24 showed treatment benefit scores increased slightly in both groups (S FG 36.7 to 37.2; IF G 36.7 to 37.2); however, no significant differences were found between groups. The scale range for treatment benefits was lowest score 9 to highest score 36. Both groups showed high treatment benefit score s before and after intervention. No significant differences were found between groups in the mean selfefficacy scores from week 1 to week 24. The difference from week 1 to week 24 showed a larger improvement in the SF G (30.21-33.15)) compared to the IFG (31.04 -30.74). The scale ranged from 9 (l owest) to 36 (hi ghest). Both groups showed a slight increase in mean scores at 24 weeks. Both pre-intervention and post-intervention self-efficacy scores re flected high self-efficacy beliefs. There was a significant difference at week 24 for the mean severity scores between the control and the intensive follo w-up group. The mean severity score increased from 25.4 to 27.3 for the cont rol; whereas, the intensive follow-up group declined from 25.5 to 25.0. The scale ranged from 8 to 32. The mean severity scores reflected high val ues for both groups before and after intervention. There was no significant difference between groups in mean susceptibility scores from week 1 to week 24. The control group means baseline score

PAGE 207

194 increased from 27.8 to 28.9 at 24 weeks. The intens ive follow-up group means score went from 28.1 to 27.8. The scale ranged from 9 to 36. Both group means indicated high-perceived susceptibility sco res before and after intervention. Cue to action mean scores for the control group remained high going from baseline 18.2 to 18.9 at week 24. The intensiv e follow-up group incr eased from 17.5 to 17.6 at week 24. Both groups showed improvement in structural element scores from baseline to 24 weeks. The control increas ed from 10.9 to 12.8. The intensive follow-up group increased from 10.9 to 12. 1. There was no significant difference between groups. The scale ranged from 4 to 16. Both groups had high scores pre-and post intervention. Mean motivation sco res for both groups were similar. The mean scores remained high at baseline and at week 24. (SFG 18.2 to 18.5; IFG 18.8 to 18.3) At week 12, both groups completed the diabetes self -care practice assessment (PCA) designed to measure diabetes self-care skills, knowledge, and adherence to each aspect of the diabet es regimen. Both the diabetes educator and the patient rated the tool. For example, when rating diet knowledge and compliance, the dietitian CDE, took in to account the patient’s ability to understand nutrition concepts covered by th e curriculum, but also reviewed food

PAGE 208

195 recalls and records. The rate r also utilized the patient and family self-reports: preand post-test; behavioral goal evaluations (Appendix 6), and when appropriate, observation of actual skills. Reading nutrition labels and selecting appropriate amounts of carbohydrates in a lunch m eal served are examples of ways the educator observed and rated patient perfo rmance (see Table 3, Table 4). The diabetes Self-Care Practice Assessment designed to meet the objectives of the Life with Diabetes curriculum was used to assure the fidelity of the program. Both groups had similar mean scores at 12 weeks. The mean of all persons completing the PCA was 95.5%’ the c ontrol group mean was 95.7% (N=55) compared to the intensive group mean score 95.4%.

PAGE 209

196 Chapter Five: Discussion This chapter includes: a summary of results; limitations; conclusions; implications for diabetes education, clin ical practice, and glycemic control; and lessons learned and recommendations for future research.

PAGE 210

197 Summary This study sought evidence as to w hether type 2 adults who participate in ADA approved DSME randomly assigned to an intensive follow-up would have significant differences in HgbA1c values, health belief and behavior scores compared to those assigned to the standard follow-up over time. A repeated measure ANOVA analysis show ed a statistically significant reduction in mean HgbA1c at each time period from week 1 to week 12 (SFG 8.6-7.1, N=79; IFG 8.5-7.1, N=80) and from week 12 to week 24 (SFG 7.1 to 6.9, N=48; IFG 7.1-7.0, N=53). There were no significant differences found between the groups (F=0.17 Pr>f ). The 8 major constructs of the expanded health belief model (self-efficacy, treatment benefits, susceptibi lity, severity, cues to ac tion, structural elements, motivation and barriers) were measured at baseline and at the six month (24 week) follow-up using the revised expanded health belief model questionnaire. A repeated ANOVA profile analysis found no signi ficant differences in health belief scores between the groups except for perceiv ed severity (p value 0.03). The mean severity scores for the standard fo llow-up group was higher (27.05) than for the intensive follow-up gr oup (25.00) at week 24. Bo th groups had similar high scores at baseline. The follow-up interv ention using the CGMS had resulted in a slight lowering of the pati ent’s perception of their severity; however, both groups perceived severity scored remained high. T-Test results comparing those who

PAGE 211

198 completed study (N=82, mean severity score=25.65) compared to those dropped from the study (N=77, m ean 25.25) showed no difference in mean severity (Pr >t 0.5444). There is a statistical diffe rence (p-value=0.02) when comparing the standard follow-up group mean scores at week 24 with the intensive group. The intensive follow-up group m ean declined slightly from 26.27 to 24.82. The SFG slightly increased from 24.52 to 26.64. Both groups had high means scores that improved from baseline to six month for treatment benefits, self-effi cacy, cuestoaction and structural elements. Barrier mean scores for the SFG slightly decreased (96.1 to 90.0) where as the IFG remained fairly constant (93.4 to 94.1). Both groups had similar low mean scores before and after in tervention (p-value > 0.05). Mean susceptibility scores slightly increased in the SFG (27.8-28.9) compared to the IFG lowering after intervention (28.1-27.8). Both groups had high mean scores preand post intervention. Motivation mean scores for both groups were constant remaining high before and after intervention. Mean barrier scores were similar for both groups with the mean scores improving in the SFG from baseline to post intervention (96.15-90.62) whereas the IFG remained fairly constant ( 93.40-94.08). The mean barrier scores reflected a low level of barriers befo re and after study interventions. Mean motivation scores remained consistently high before and after intervention with no differences found between groups. Cues to action scores remained

PAGE 212

199 consistently high from baseline to the si xmonth follow-up visit with no difference found between groups. The diabetes self-care practice score measures the patient’s understanding and ability to carr y out specific diabetes self-care behaviors. Both groups had similar high mean diabetes self-care practice scores (SFG=95.7, IFG=95.4). The diabetes self-care practi ce assessment developed from the ADA approved “Life with Diabetes “curriculum obj ectives helped ensure the fidelity of the education program. At baseline, the control and the int ensive follow-up groups had similar distributions of the cat egorical variables: treatment type, ethnicity, education level, and type of physician and duration of diabetes. At baseline, the intensive follow-up group had slightly more women than men (IFG N=51; SFG N=37). This slight gender difference had no known im pact on the primary variable under study. There is no conclusive evidence t hat suggest that HgbA1c or health belief scores would be different due to gender. The United Kingdom Prospective Study (1999) they found no differences in HgbA1c based on gender.

PAGE 213

200 Limitations The control and intensive follow-up group had similar distributions at baseline for the continuous variables: age; week 1 HgbA1c values; and the 8 Health Belief Model subscales. T he randomization worked well at evenly distributing the variables for a small sa mple size. The univariate SAS procedure was utilized to test the variables for no rmal distributions see Appendix 10). A normal distribution was confi rmed for the variables: self-efficacy, susceptibility, severity, cues to action, structural el ements, motivation, barriers and age. The box plot for treatment benefits showed equal distri bution: however, a ceiling effect towards the top of the scale indica ted that those who make the effort to attend diabetes classes are people who per ceive that education offers a high benefit. This finding is not unexpected si nce the exclusion cr iterion limits the study to participants willing to participat e in diabetes education. The finding of this study therefore may not be generalized to all diabetes patients. Patients unwilling or unable to attend diabetes classes may be different Due to the exclusion of patients with HgbA1c values less than 7%, the plot of the HgbA1c values reflected the upper hal f of the bell –shaped curve. At lower HgbA1c levels a significant clinical di fference due to effects of treatment would be difficult to measure without prohibitive sample sizes. From the pilot study and outcome data prior to the study we expec ted a drop in HgbA1c at 3 months but these outcomes not being obtained in a controlled study could only be

PAGE 214

201 considered an estimate. During the year s 2000-2004, the center’s outcome data for the years corresponding to the study showed a 1.14 to 2.27 percent reduction in mean HgbA1c levels per year (500 829 patients per year) for those who returned for follow-up. The results of this study cannot be generalized to all type 2 patients. This study is delimited to sample of health y type 2 adult volunteers referred to an outpatient ADA approved DSME employ ing an empowerment goal oriented approach. Other diabetes populations may have different results. The study results showed that participants of this study had high perceived health beliefs and were highly motivated prior to the DSME and follow-up intervention. Although at six months t here were slight improvem ent in mean health belief scores there no statistical difference found between the groups or differences over time except for perceived severity.. Patients not attending DMSE or those w ho did not complete the education interventions may be different and have di fferent outcomes. The participants in this study do not characterize all pers ons with type 2 who may seek or are in need of diabetes education. The DSME offered at the Diabetes Care Institute of University Community Hospital is an outpatient pr ogram providing services in the Tampa Bay area other types of programs in other geographic loca tions may be different. There is a potential for investigator bias because the investigator and Diabetes Educators

PAGE 215

202 may represent “passionate” workers. Us e of a Standardized curriculum, audiotaping, peer-evaluations and competencies, and employing the patient care assessment helped ensure the fidelity of the program. T he investigator and Diabetes Educators were blinded to the group assignment. The program assistant not involved in the education entered the data. Data was filed in medical records and databases until compiled for data analysis. Both groups had similar improvem ents in HgbA1 over time. The 1.4 percent HgbA1c reductions achieved in the education arm of the study was higher than expected.. This larger than expected drop in the HgbA1c made it more difficult to see significant ch anges between groups during the intervention phase. The smaller effect size of the in tervention would be difficult to detect in this sample size These findings provide further evidence that DSME meeting the national standards and follow-up interventio ns help achieve and maintain better glycemic control. Longer-term prospective studies are needed to determine whether these improvement s can be sustained. Several meta-analysis studies (Norris et. al., 2002, Brown et. al., 1992; Padgett et. al., 1998) confi rmed a beneficial effect of DSME on knowledge, improved diabetes self-care behaviors and im proved metabolic control. Brown (2002) stated that previous DSME studies showed an early moderate effect size on HgbA1c with a peak at 1 to 6 months. Similar results were found in this study. Participants in this study showed a conti nual reduction of HgbA1c from 12 weeks

PAGE 216

203 to 24 weeks. The study follow-up period wa s limited to six mont hs and therefore the longterm effects are not known. O ne-year follow-up data was available on early participants that showed a trend for the HgbA1c lowering to be sustained. Future studies are needed to confirm th is observation. Mean outcome data for all patients participating in DSME also s howed that patients returning in one year sustained HgbA1c improvements. Brown found that benefits on psychological outcomes improved more after 6 months. In this short-term study the expanded health belief model was not particularly helpful in determining changes in psychosocial outcomes. The groups showed no statistical differences in their Health belief scores except for severity. A highly motivated volunteer sample does not represent a cross-section of the diabetes population. Studies with other sample populations with less motivation or who score low initially on health beliefs may need to be explored to determine the utility of the model. In light of the study findings of no significant differences perhaps Diabetes Educators need to ex plore other theoretical models and frameworks that would be more us eful. The behavioral and health belief outcomes measured in this study by t he revised diabetes expanded health belief model questionnaire may be different than those measured with other instruments. Even though structural element questions in the EHBM questionnaire involved social support this study was limited by its lack of direct tracking of family involvement.

PAGE 217

204 One hundred and fifty-nine participant s had baseline HgbA1c values.. Fiftysix had missing HgbA1c values at week 24. Forty-three out of the fifty six were lost to follow-up (unable to contact by phone or mailings). Nine relocated out of the Tampa Bay Area. Four patients withdrew due to ti me constraints. One reason given for patient’s missing the 3 -m onth follow-up was that their physician ordered HgA1c and other labs pr ior to their 12-week follow -up. Of the remaining 103 who had an HgbA1c at week 24, twent y-three did not complete the second EHBM questionnaire. 82 0f the 159 had a ll primary and secondary variables at the end of the study period. Those that completed the study were similar to those who dropped out for all variables exce pt those completing the study were slighter older, the mean age 55 compared to 50. The ti me required for follow-up was limited to working hours 7:30 to 6pm, which may have been a factor for younger working adults. The final basel ine data analysis showed that the younger participants who dropped out were evenly distributed between the groups. Initially, the sample size was estimat ed for 159 participants to achieve adequate power (0.80) at a 0.05 alpha level. T he control group size was planned for 78 patients to allow for a 30% attrition rate. The int ensive follow-up group sample size was planned for 81 subjects to allow fo r a 35% attrition rate. Despite use of reminder letters, repeated phone calls and offe ring flexible follow-up times the attrition rate was higher than anticipated. Overall 35% (56/ 159) of patients did

PAGE 218

205 not return for the HgbA1c determination at week 24. After dropping from the analysis those with incomplete data, 82 patients had all primary and secondary variables. In the standard follow-up group fifty-four percent dropped out compared to forty-three percent of t he intensive group. The higher then expected attrition and the significant lo wering of the HgA1c following DSME made it difficult to detect small mean di fference in HgbA1 between the follow-up groups. The expected reduction in HgbA1c is less as patients obtain HgbA1c goals. The mean HgbA1c value for both fo llow-up groups were at the ADA goal of seven percent or less (SFG 6.9,,IFG 7.0). The proposed Hypothesis that the CGMS follow-up group would do better wa s not substantiated by the study results. The conclusions drawn that no differences exists between the groups should be interpreted with caution due to the loss of power to detect small differences between the groups. This new technology utilized in the intensive follow-up group resulted in similar reducti ons in HgbA1c as the standard followup. There was no statistically signific ant difference found between from the groups. Fifty-four percent of the patients in t he standard follow-up group achieved the ADA goal for HgbA1 (7% or le ss) compared to 58% in the intensive follow-up group. In light of these findings of no added b enefit, the standard follow-up is recommended over the intensiv e since it is less time consuming and more costeffective. The HgbA1c lowering effects of the CGMS reported in previous studies

PAGE 219

206 were mainly attributed to changes in medical management. Several studies reported effective loweri ng of HgbA1 when the sensor was used to adjust diabetes therapy. (Bode et al., 1999; Ludvigsson &Hanas, 2003 ;). This study controlled for medical management changes in order to examine the effects of DSME and t he follow-up interventions on the HgA1c outcomes. Those with recent medical changes were excluded in this study.

PAGE 220

207 Conclusions This study provides evidence that DSME programs that meet national standards effectively help type 2 adults improve their HgbA1 outcomes. Mean HgbA1 levels significantly dropped from week1 to week 12 following the DSME and from week 12 to week 24 following the follow-up intervention. There were No statistically differences found bet ween the SFG and the IFG. There were no statistically significant differences in mean Health belief scores and PCA scores between groups except for mean severity scores. Those in the IFG had significantly lower mean severity scores than the standard group.

PAGE 221

208 Implications for Diabetes Education, Clinical Practice, and Glycemic Control DSME should be offered to all persons with type 2 diabetes regardless of age, duration of diabetes or type of treatment. Diabetes Self-management. Education is needed to acquire the necessary knowledge, skills, and attitudes to perform self-care behaviors that foster good metabolic control. Follow-up interventions with certified di abetes educators were effective in measuring clinical improv ement through point of care HgbA1c measurement and to Identify whether patients were successful in achieving their diabetes selfmanagement goals. Those unsuccessful in obtaining the ADA goal need further evaluation of their barriers to control. The Certified Diabetes educator worked collaboratively with the patient to ident ify the barriers and help patients form strategies to help overcome these barri ers. This collabor ative relationship resulted in improvement regardle ss of the type of intervention. Norris (2002) concluded from his meta -analysis examining the effects of DSME that patients in these studies showed improved A1C results after DSME and these reductions in HgbA1 levels we re sustained with regular provider or professional contact. Pati ents not followed by health professionals regularly had relapses in their HgbA1 improvements. Recently, the American Association of Diabetes Educators (AADE, 2003) recommended that seven scientific-based di abetes self care behaviors (being active, eating healthy, medication, monitori ng, problem-solving, lowering risks for

PAGE 222

209 and living with diabetes) should be outco mes measured by DSME programs. The AADE suggested that DSME program s evaluate a continuum of outcome measures from immediate outcomes (tests observation of skills) to intermediate outcomes (lab test, self-report of goal progress, behavioral and health belief measurements) to post intermediate out comes (clinical improvement measures) to long term outcomes. These new recommendations pose considerable challenges for education programs that often lack the financial support and reimbursement needed to carry out these goal s. The paradigm of responsibility or improved diabetes outcomes is shifting again from provider to patient to a shared responsibility. Di abetes goals need to be formed by the patient with the active guidance of the health providers fo r them to achieve the desired effects. Evaluation of outcomes is necessary to determine which type of interventions is best for which populations. Questions regarding how often follow-up interventions should be offered and for what duration to prevent relapse need to be answered. At present the CGMS has received poor reimbursement for use in type 2 patients. Due to the cost s of these new technologic al interventions and the highly successful DSME outcomes utilizat ion of the sensor for educational and behavioral changes alone are not support ed. Further research is needed to determine which type 2 patients would bene fit when medical therapy adjustments are included in the intervention..

PAGE 223

210 The study findings support current re commendations that patients should first receive DSME. A basic level of kno wledge is necessary in order to carry out diabetes self-care skills and behaviors that lead to better clinical and psychological outcomes. Further studi es are needed to determine which patient characteristics are contributing to the successful and unsuccessful HgbA1c outcomes.

PAGE 224

211 Implications for Public Health According to the recent National Heal th and Nutrition Examination Survey (NHANES III) and data from the Am erican Association of Clinical Endocrinologists (AACE) survey, the ma jority of persons with type 2 diabetes have failed to reach A1C goals. The NHA NES survey (2000) reported only 37% reached goal levels, whereas the AAC E survey (2005) reported only 33% reached the AACE preferred A1C goal of 6.5%. In light of the culminating evidence that intensive diabetes managemen t lowers the burden and cost of diabetes Public Health policies are needed to assure that diabetes education is affordable and accessible to the public. Public Health initiatives are needed to continue the legislative efforts for adequate DMSE insurance coverage and reimbursement for diabetes supplies. The passage of the Balanced Budget Act in 1997 acknowledged the impor tance of DSME and led changes in the Medicare coverage of Diabetes. Prior to this legisl ation, persons with di abetes could only receive DSME within one year of their di agnosis or if they had 2 consecutive HgbA1c values over 8.5% or othe r new onset diabetes micro-vascular complication. Patients had to already present with major renal and retinal problems to qualify. Studies on the cost and healthcare utilization of Diabetes have shown that waiting unt il diabetes complications set in are considerably more costly. Although this legislation improved the Medicare guidelines it still does not do enough to prevent the diabetes prevalence from growing. The

PAGE 225

212 Center for Disease Control (2005) es timates 20.8 million has diabetes and another 54 million have insulin resistanc e of prediabetes. Epidemiological studies suggest that the di abetes epidemic is largely due to the person’s diet and lifestyle (Schulze & Hu, 2005). Medicare recognizes prediabetes and metabolic syndrome but still denies DSME and even mi nimal Medical Nutritional Therapy coverage to these patients. In the light of the high success rate (55% SFG, 58%) of patients who Attended DSME public health policies are needed to adapt the ADA DMSE to state and federal funded program s. Public Health policy needs to foster the various agencies to work towards t he common goal. Presently, ADA and AADE are recommending 3 sets of goals. Firs t, primary prevention needs to address the obesity and prediabetes population. These high risks patients need education to help them make difficult lifestyle changes such as diet and exercise. Secondary prevention focuses on those wit h the disease controlling the disease to prevent complications. Tertiary prevent ion goals are to prevent morbidity and mortality from the diabetes related complications. Although the CGMS has im portant clinical applications for physicians in medical therapy, policies regarding its use as an education intervention are not warranted due to the option of a follow-up intervention with the CDE that is as effective and less costly.

PAGE 226

213 Lessons Learned and Recommendati ons for Future Research Long–term prospective DSME outcome trials have had problems with high attrition rates. In addition, the health benefits attributable from education are difficult to measure separately from ot her therapies. These issues make it difficult to draw reliable conclusions about the effectiveness of DSME and followup interventions. In part this is due to ethical concer ns about use of a control group that withholds beneficial therapies. In order to detect sm all effect sizes, impractical sample sizes and needed. Increase in awareness and funding for educational outcome research is needed. A coalition made up of mem bers from the American Diabetes Association, Foundation for Account ability, The Health Care Finance Administration, the National Diabetes Quality Improvemen t Alliance, the American Academy of Family Physicians, The American College of Physicians and the Veterans Administration formed to develop an evidence-based consensus for outcome measurement goals. These endeavors have lead to the development of standards for DMSE and diabetes care. Outcome measures will continue to help evolve to answer the ques tions as to what is best practice. ADA approved DSME programs need to measure both clinical and psychosocial outcomes in order to determine how effective they are in empowering the patient to achieve thei r goals. Follow-up interventions are a

PAGE 227

214 critical time to focus on progress made by patients toward their goals and to help patients overcome the barriers they enc ounter. Numerous studies have provided evidence that psychosocial outcomes are important to measure because they affect clinical outcomes (Peyrot &R ubin, 1999, DCCT, 1996; Schafer, McCaul & Glasgow, 1986; Peyrot & Rubin 1995). Focus studies are needed to determine why patients do not follow-up and ways to increase return rates. Further studies are needed to determine costeffective ways to follow-up on patients, and at what interval are necessary to achieve and sustain diabetes self -care goals and behaviors. Follow-up intervention outcome measur es are needed to assess t he patient’s progress to their goals, measure their behavioral chang e and develop strategies to overcome any barriers encountered. Follow-up with the diabetes educators helped reinforce patient’s motivation and efforts toward their goals. Previous studies have found that patient’s did better when they communicated regularly with health professionals than when left on their own (Brown, 1992; Norris, 2002). The drop out rate was higher than planne d for estimated sample size. As a result the loss in sample size resu lted in a loss of power to detect small differences in groups. Since both groups showed significant drops in HgbA1c 1.6%, only very large samples could det ect a difference between the groups if one truly exist. Although unable to provi de evidence that the intensive follow-up was more effective, the study did indica te that HgbA1c values could improve

PAGE 228

215 significantly from both education and follo w-up interventions. The lower HgbA1c levels achieved in this study reflects a clinically significant difference, for every 1% lowering of the HgbA1c a 35% reduc tion of microvascular complication has been shown in type 2 population. The International Diabetes Center (2001) reported hav e that medical management can account for 0.5-4% lowering of HgA1c. This study controlled for medical management effect s by limiting sample to those who have had no recent change in medications. In light of the study findings, it is recommended that those who have not had DSME be offered training regardless of age, duration of the disease and other factors. Explorator y regression model research is needed to find out what elements of the ADA curriculum accounts for the most improvement in HgbA1c outcomes. There are a number of people who may fi nd attending ten hours of DSME and participating in follow-up too time consuming. By learning more about the ADA curriculums essential elements we can meet the needs of those the ADA program is not currently reaching and make it more appealing to those who feel the ten hour program is too time consuming. Additional studies are needed to dete rmine the type and duration of followup interventions that prevent relapse....D ue to time and financial constraints of this study the follow-up time period was limited to 24 weeks. Some patients with early enrollment came back for 1-year follow-up. The lower hgbA1c mean for

PAGE 229

216 those with 1 year followup, showed a trend of maintenance of the improved HgbA1c. N=51, 6.95: SFG N=27, HgA1 c mean 6.78; IFG N=24 mean HgbA1c 7.15). Future studies are needed to determine if long-term effects can be maintained and what type and frequency of intervention is needed to obtain and sustain diabetes goals.

PAGE 230

217 References AACE. State of Diabetes in America. May 2005. Available at: http://www.aace.com/public/awareness/ stateofdiabetes/DiabetesAmericaR eport.pdf .Accessed July 24, 2006. Abourizk, N.N., O’Connor, P. J., Crabtree, B.F., et al (1994). An outpatient model of integrated diabetes treat ment and education: functional, metabolic, and knowledge outcomes. The Diabetes Educator 20(5): 416421. Alogna, M. (1980). Percepti on of severity of disease and health locus of control in compliant and non-compliant diabetes patients. Diabetes Care, 3 533534. American Association of Diabetes Educators. (1992). The scope of practice for diabetes educators and the standards of practice for diabetes educators. Diabetes Educator, 18 52-56. American Association of Diabetes Educat ors. (2003). Position Statement: standards for outcome measurement of management education. Diabetes Educator, 29, 804-816. American Diabetes Association. (1990) Diabetes Mellitus and exercise (position statement). Diabetes Care, 13 804-805. American Diabetes Association. (1993). Im plications of the diabetes control and complications trial. Diabetes Care, 16 (11).

PAGE 231

218 American Diabetes Association, (2006). Clinical Practice Recommendations 2006. Diabetes Care, 29(1), S11-17. American Diabetes Association. (1996a). Diabetes: 1996 vital statistics McClean, VA: Author. (p. 1-102). American Diabetes Association. (1996b). National standards for diabetes selfmanagement educational programs and American Diabetes Association Review Criteria. Diabetes Care, 19 (Suppl. 1), 5114-5118. American Diabetes Association. (1997). Life with diabetes: A series of teaching outlines Alexandria, VA: Michigan Diabetes Research and Training Center. American Diabetes Association. (1998a). Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care, 21 ,514. American Diabetes Association. (1998b) Economic consequences of Diabetes Mellitus in the United States in 1997. Diabetes Care, 21 (2), 296-309. American Diabetes Association. (2001). Standards of medical care for patients with diabetes mellitus. Diabetes Care, 24 (Suppl. 1), S33-S43. American Dietetic Association. (1995). The diabetes care and education dietetic practice group. Journal of the American Dietetic Association, 95 (5), 607. Anderson, B. J., Wolf, F. M ., Burkhart, M. T., Cornell, R. G., & Bacon, G. E. (1989). Effects of peer-group interv ention on metabolic control of adolescents with IDDM: Random ized outpatient study. Diabetes Care, 12

PAGE 232

219 179-183. Anderson, L. A. (1990, Suppl. 2). Health-care communication and selected psychosocial correlates of adherence in diabetes management. Diabetes Care, 13 66-76. Anderson, R. M., Fitzgerald, J. T., Oh, M. S. (1993). The relationship between diabetes-related attitudes and pati ents’ self-reported adherence. The Diabetes Educator, 19 (4), 287-292. Anderson, R. M., Funnell, M. M., Butler, P. M., Arnold, M. S., Fitzgerald, J. T., & Feste, C. C. (1995). Patient empowerment re sults of a randomized controlled trial. Diabetes Care, 18 (7), 943-949. Ary, D. V., Toobert, D., & Wilson, W. (1986). Pati ent perspective on factors contributing to nonadherence to diabetes regimen. Diabetes Care, 9 168172. Ary, D. V., Toobert, D., Wilson, W., & Glasgow, R. E. (1986). Patient perspective on factors c ontributing to nonadherence to diabetes regimen. Diabetes Care, 9 (2), 168-172. Assal, J. M., Mulhauser, I., Pernet, A. A ., Feller, R., Jorgens, V., & Berger, M. (1985). Patient education as the bas is of diabetes care in clinical practice. Diabetologia, 28 602. Atkinson, M. A., Maclaren, N. K., Scharp, D. W., Lacy, P. E., & Riley, W. J. (1990). 64000 M1 autoantibodies as predictors of insulin-dependent

PAGE 233

220 diabetes. Lancet, 335 1357-1360. Bandura, A. (1977a). Self -efficacy: Toward a unifying theory of behavioral change. Psychological Review, 84 191-215. Bandura, A. (1977b). Social learning theory Englewood Cliffs, NJ: PrenticeHall. Bandura, A. (1986a). Scope of self-efficacy theory. Journal of Social and Clinical Psychology, 4 359-373. Bandura, A. (1986b). Social foundations of thought and action: A social cognitive theory Englewood Cliffs, NJ: Prentice-Hall. Bandura, A. (1992). Self-e fficacy mechanism in psychobiologic functioning. In R. Schwarzer (Ed.), Self-efficacy: Thought control of action (pp. 355-395). Washington, D. C.: Hemisphere. Barnard, R. I. (1938). The functions of the executive Cambridge, Massachusetts: Harvard University Press. Bayer Corporation. (1997, April). Glucometer Dex Diabetes Care System User Guide Elkhart, IN: Author, Diagnostic Division. Becker, M. H. (Ed.). (1974). T he health belief model and personal health behavior. Health Education Monographs, 2 324-473. Becker, M. H. (1976). So ciobehavioral determinants of compliance. In D. L. Sackett & R. B. Haynes (Eds.), Compliance with therapeutic regimens Baltimore, MD: Johns Hopkins University Press.

PAGE 234

221 Becker, M. H. (1985). Patient adherence to prescribed therapies. Medical Care, 23 (May), 539-555. Becker, M. H., Drachman, R. H., & Kirscht, J. P. (1974a). A new approach to explaining sick role behavior in low-income populations. American Journal of Public Health, 64 204-216. Becker, M. H., Drachman, R. H., & Kirscht, J. P. (1974b). A field experiment to evaluate various outcomes of continuity of physician care. American Journal of Public Health, 64 (November), 1062-1070. Becker, M. H., & Maiman, L. A. (1975). Sociobehavioral determinants of compliance with health and medical care recommendations. Medical Care, 13 10-24. Becker, M. H., Haefner, D. P., Kasl, S. V., et al. (1977). Selected psychosocial models and correlates of individ ual health-related behaviors. Medical Care, 15 (Suppl), 27-46. Becker, M. H., Maiman, L. A ., Kirscht, J. P., Haefner, D. P., & Drachman, R. H. (1977). The health belief model and pr ediction of dietary compliance: A field experiment. Journal of Health and Social Behavior, 18, 348-366. Becker, M. H., Radius, S. M., Rosenstock, I. M., et al. (1978). Compliance with a medical regimen for asthma: A te st of the health belief model. Public Health Rep, 93 268-277. Becker, M. H., & Janz, N. K. (1985). The health belief model applied to

PAGE 235

222 understanding diabetes r egimen compliance. Diabetes Education, 11 4147. Beeney, L. F., & Dunn, S. M. (1990) Knowledge improvement and metabolic control in diabetes education: Approaching the limits? Patient Education and Counseling, 16 217-229. Bellenir, K., & Dresser, P. D. (Eds.). (1995). Diabetes sourcebook Detroit, MI: Omnigraphics. Bice, T., & White, K. (1969). Factors re lated to the use of health services: An international comparative study. Medical Care, 7 124. Biglan, A., Glasgow, R. E. (1991). The social unit: An important facet in the design of cancer control research. Preventative Medicine, 20 292-305. Bloomgarden, A. T., Karmally, W., Metz ger, J., Brothers, M., Nechemias C., Bookman, J., Faierman, D., GinsbergFellner, F., Rayfield, E., & Brown, W. V. (1987). Randomized contro lled trial of diabetic education: Improved knowledge wit hout improved metabolic status. Diabetes Care, 10 263-272. Bode, B., Gross, T., Thor nton, K., & Mastrototaro, J. (1999). Continuous glucose monitoring used to adjust di abetes therapy improves glycosylated hemoglobin: a pilot study. Diabetes Research and Clinical Practice, 46, 183-190. Boehm, S., Schlenk, E. A., Funnell, M. M ., Powers, H., & Ronis, D. (1997).

PAGE 236

223 Predictors of adherence to nutrition recommendations in people with noninsulin-dependent diabetes mellitus. The Diabetes Educator, 23 (2), 157164. Bognetti, E.R.M., Zoja, A., Ciralli, F ., Chiumello, G. (1995). HbA1c Determination on Capillary Blood Sample : Validity, Stability, and Potential Usefulness. Diabetes Care, 18 (9), 1305-1306. Bolinder, J., Hagstrom-Toft, E., Ungerstedt, U., Arner, P. (1997) Self-monitoring of blood glucose in type 1 diabeteic pat ients: comparison with continuous microdialysis measurements of gl ucose in subcutaneous adipose tissue during ordinary life conditions. Diabetes Care 20 (1), 64-70. Brown, S. (1988). Effect s of educational interventions in diabetes care: A metaanalysis of findings. Nursing. Research, 37 223-230. Brown, S. A. (1990). Studies of educational interventions and outcomes in diabetic adults: A meta-analysis revisited. Patient Education Counseling, 16 189-215. Brown, S.. A. (1992). Meta analysis of diabetes patient education research: variations in interventi on effects across studies. Research in Nursing & Health, 15:409-419. Bryer-Ash, M., Granda-Ayala, R ., Clejan, S., Jaspan, J. (1997, January). Use of a Home Collection Kit to determine Hemogloblin A1c. ADA Abstract Burge, M. (1993). Lack of Compliance with HBGM predicts hospitalizations in

PAGE 237

224 diabetes care. Diabetes Care 28 (8), 1502-1503. Burns, A. C. (1992). The expanded health belief model as a basis for enlightened preventive health ca re practice and research. JHCM, 12 3245. Cagliero, E., Levina, E.V., Nathan, D.M. (1999). Imm ediate Feedback of HgbA1c levels improves glycemic control in type 1 and Insulin-treated type 2 diabetic patients. Diabetes Care 22, 1785-1789. Carey, M. P., Jorgensen, R. S., Weinstock, R. S., Sprafk in, R. P., Lantinga, L. J., Carnrike, C. L. M., Jr., Ba ker, M. T., & Meisler, A. W. (1991). Reliability and validity of the Apprai sal of Diabetes scale. Journal of Behavioral Medicine, 14 14-43. Carter Center of Emory University. (1985). Closing the gap. The problem of diabetes mellitus in the United States. Diabetes Care, 8 391-406 Cauffman, J., Peterson, E., & Emrick, J. (1967). Medical care of school children: Factors influencing the outcome of refe rral from a school health program. American Journal of Public Health, 57 60. Center of Disease Control and Prevention (CDC). (1997). Center of Disease Control and Prevention: Diabetes surveillance Atlanta, GA: U.S. Dept. of Health and Human Services. Center of Disease Control and Prevention. Statistics. The public health burden of diabetes melitis in the United States Diabetes Surveillance Report.

PAGE 238

225 1997. Online ed. Available at: http://www.cdc.gov/diabetes/statisti cs/index.htm. Accessed March 15, 2001. Center of Disease Control and Prevention (CDC). (2005). Center of Disease Control and Prevention Na tional diabetes fact sheet: general information and national estimates on diabetes in the United States. Atlanta, GA: U.S. Dept of Health and Human Services. Cerkoney, K. A. B., & Hart, L. K. (1980). The rela tionship between the health belief model and compliance of persons with diabetes mellitus. Diabetes Care, 3 594-598. Clement, S. (1995). Diabete s self-management education. Diabetes Care, 18 1204-1214. Cohen, J. (1988). Statistical power analysis for the behavioral sciences New York: Academic Press. Coonrod, B. A., Betschart, J., & Harri s, M. I. (1994). Frequency and determinants of diabetes patient education among adult s in the U.S. Population. Diabetes Care, 17 (8), 852-858. Crabtree, M. K. (1987, March). Self-efficacy, social support, and social incentives as predictors of diabetic self-care, type II Paper presented to the Society for Behavioral M edicine, Washington, DC. Cronbach, L. J. (1975). Beyond the two disciplines of scientific psychology.

PAGE 239

226 American Psychologist, 30 116-127. Davis, W. K., Hess, G. E., Harrison, R. V., & Hiss, R. G. (1987). Psychosocial adjustment to and control of diabetes mellitus: Differences by disease type and treatment. Health Psychology, 6 1-14. Davis, W. K., Hess, G. E., & Hiss, R. G. (1988). Psychosocial correlations of survival in diabetes. Diabetes Care, 11 538-545. Delamater, A. M., Bubb, J., Davis, S. G., Smith, J. A., Schmid t, L., White, N. H., & Santiago, J. V. (1991). Random ized prospective study of selfmanagement training with newly di agnosed diabetic children. Diabetes Care, 13 492-498. The Diabetes Control and Complications Tr ial Research Group (DCCT). (1993). The effect of intensive treatment of diabetes on the development and progression of long-term complicati ons in insulin-dependent diabetes mellitus. New England Journal of Medicine, 329 977-986. Dorman, J. S., McCarthy, B. J., O’Leary, L. A., & Koehl er, A. N. (1995). Risk factors for insulin-dependent di abetes. In M. I. Harri s, C. C. Cowie, M. P. Stern, E. J. Boydo, G. E. Re iber, & P. H. Bennett (Eds.), Diabetes in America (pp. 165-178) (2nd ed). Bethesda, MD: National Institutes of Health, National Institutes of Diabet es and Digestive and Kidney Diseases (NIH publ. No. 95-1468). Dunbar, S. M., & Stunkard, A. J. (1979). Adherence to diet and drug regimes.

PAGE 240

227 In R. Levy, B., Rifkind, B. Dennis, & N. Ernst (Eds.), Nutrition, lipids, and coronary heart disease New York: Raven Press. Dunn, S. M., Bryson, J. M., Hoskins, P. L., Alford, J. B., H andelsman, D. J., & Turkle, J. R. (1984). Developmen t of the diabetes knowledge (DKN) scales: Forms DKNA, DKNB, and DKNC. Diabetes Care, 7 (1), 36-41. Dunn, S. M., Smart, H. H., Beeney, L. J., & Tuttle, J. R. (198 6). Measurement of emotional adjustment in di abetic patients: Va lidity and reliability of ATT39. Diabetes Care, 9 (5), 480-489. Dunn, S. M. (1990). Re thinking the models and modes of diabetes education. Patient Educationa l Counseling, 16 281-286. Edelman, S., Balonga, J. (1977). Prelim inary evaluation of a novel real-time blood glucose monitor. Paper pr esented at the American Diabetes Association Annual Meeting, New Orleans, LA. Everhart, J., Knowler, W. C., & Bennett, P. H. (1985). Incidence and risk factors for non-insulin-dependent diabetes. In M. I. Harris, & R. F. Hamman (Eds.), Diabetes in America (Chap. 10). Washi ngton, D.C.: U. S. Department of Health and Human Services. Expert Committee on the Diagnosis and Cla ssification of Diabetes Mellitus. (1998). Report of the Expert Committee on the Diagnosis and Classification of Diabetes Mellitus. Diabetes Care, 21 (Suppl. 1). Fishbein, H., & Palumbo, P. J. (1995) Acute metabolic complications in

PAGE 241

228 diabetes mellitus (p. 283-292). In M. I. Harris, C. C. Cowie, et al., (Eds.), Diabetes in America 2nd ed. Bethesda, MD: Na tional Institute of Health, National Institutes of Diabetes and Digestive Diseases (NIH publ. no. 951468). Fishburn, M., & Ajz en, I. (1975). Belief, attitude, intention, and behavior: An introduction to theory and research Reading, MA: Addison-Wesley. Fitzgerald, J. T., Davis, W. K., Connell, C. M., Hess, G. E., & Funnell, M. M. (1996). Development and validation of the Diabetes Care Profile. Evaluation of Health Professions, 19 (2), 208-230. Ford, E. & Herman, W. (1995) Leisure-time physical actity patterns in the U.S. diabetic population: Findings from the 1990 National Health Interview Survey: Health promotion and disease prevention supplement. Diabetes Care, 18, 27-33. Franz, M. J. (1988, May/ June). Exercise and diabet es: Fuel metabolism, benefits, risks and guidelines. Clinical Diabetes 58-70. Frederickson, L., Graff, M., Gaithe r-Wallace, S., Yang, R. (2000). MiniMed Continuous Glucose Monitoring System trainer manual. MiniMed Inc. Sylmar, CA; Friedman, M. M. (1990). Transcultural family nursing: Application to Latino and black families. Journal Pediatric Nursing 5 (3), 214-222. Friedman, L. C., Nelson, D. C., Webb, J. A., Hoffman, L. P., & Baer, P. E.

PAGE 242

229 (1994). Dispositional optimism, se lf-efficacy, and health beliefs as predictors of breast self-examination. American Journal of Preventive Medicine, 10 130-135. Freinkel, N. (Ed.). (1985). Su mmary and recommendations of the Second International Workshop Conference on Gestational Diabetes Mellitus. Diabetes, 34 (2), 123-126. Funnell, M. M., Anderson, R. M., Arnold, M. S., Ba rr, P. A., Connelly, M., Johnson, P. D., Taylor-Moon, D., & Wh ite, N. H. (1991). Empowerment: An idea whose time has come in diabetes education. Diabetes Education, 17 37-41. Gabbay, K. H., Hasty, K., Breslow, J. L., Ellison, R. C., Bunn, H. F., & Gallop, D. M. (1977). Glycosylated hemogl obins and long-term blood glucose control in diabetes mellitus. Journal of Clinical Endocrinology Metabolism, 44 859-864. Garrard, J., Mullen, L., Joynes, J. O., Mc Neil, L., & Etzwiler, D. D. (1990). Clinical evaluation of the impact of a patient education program. Diabetes Educator, 16 (5), 394-400. Gilden, J. L., Hendryx, M. S., Clar, S., Casia, C., & Singh, S. P. (1992). Diabetes support groups improve health ca re of older diabetic patients. Journal of the American Geriatrics Society, 40 (2), 147-150. Given, C. W., Given, B. A ., Gallin, R. S., & Condon, J. W. (1983). Development

PAGE 243

230 of scales to measure the beliefs of diabetic patients. Registered Nursing Health, 6 127-141. Glasgow, R. E., McCaul, K. E., & Schafer, L. C. (1987) Self-care behaviors and glycemic control in type 1 diabetes. Journal of Chronic Diseases, 40 (5), 399-412. Glasgow, R. E., & Toobert, D. J. ( 1988). Social environment and regimen adherence among type 2 di abetic patients. Diabetes Care, 11 377-386. Glasgow, R. E., & Osteen, V. L. (1992). Evaluating di abetes education: Are we measuring the most important outcomes? Diabetes Care, 15 (10), 14231432. Glasgow, R. E., Toobert, D. J., & Hampson, S. E. (1992). Participation in outpatient diabetes educat ion programs: How many patients take part and how representative are they? Diabetes Education, 17 376-380. Glasgow, R. E., Toobert, D. J., & Hampson, S. E., Brown, J. E., Lewinsohn, P. M., Donnelly, J. (1992). Improvi ng self-care among older patients with type II diabetes: The “sixty something . “ study. Patient Education and Counseling, 19 61-74. Glasgow, R. E., Hampson, S. E., Strycker, L. A., Ruggiero, L. (1997). Personalmodel beliefs and social-env ironmental barriers rela ted to diabetes selfmanagement. Diabetes Care, 20 (4), 556-561. Graham, C., & Lasko-McCarthey P. (1990). Exercise options for persons with

PAGE 244

231 diabetes complications. Diabetes Educator, 16 212-220. Greene, D. S., Beaudin, B. P., & Bryan, J. M. (1991). Addressing attitudes during diabetes education: Suggestions from adult education. Diabetes Educator, 17 (6), 470-473. Greenfield, S., Kaplan, S. H. Ware, J. E., Yano, E. M ., & Frank, H. (1990). Patients’ participation in medical ca re: Effects on blood sugar control and quality of life in diabetes. Journal of General Internal Medicine, 3 448457. Greenfield, S., Roger, W., M angotich, M, et al. (1995) Outcomes of patients with hypertension and non-insulin-dependent diabetes mellitus treated by different systems and specialists. Journal American Medi cal Association, 274 1436-1444. Grey, N., Maljanian, R., St aff, I., Cruzmarino de Aponte, M. (2002). Improving care of diabetic patients through a collaborative care model. Connecticut Medicine. 66, (1): 7-11. Grossman, H. Y., Brink, S. J., & Hauser, S. T. (1987). Self-efficacy in adolescent girls and boys with insulin-dependent diabetes mellitus. Diabetes Care, 10 324-329. Haire-Joshu, D. (1988). Motivation and diabetes self -care: An educational challenge. Diabetes Spectrum, 1 279-282. Hallal, J. C. (1982). The relationship of health beliefs, health locus of control,

PAGE 245

232 and self concept to the practice of cont rol, and self concept to the practice of breast self-examination in adult women. Nursing Research, 31 (3), 137142. Hanefeld, M., Fischer, S., Schmechel, H., Roth, G., Schul ze, J., Dude, H., Schwanebeck, U., & Julius, U. (1 991). Diabetes intervention study: Multi-intentional trial in newly diagnosed NIDDM. Diabetes Care, 14 308317. Hanson, C. L., Henggeler, S. W., Harris, M. A., Burghen, G. A., & M oore, M. (1989). Family system variables and t he health status of adolescents with insulin-dependent diabetes mellitus. Health Psychology, 8 239-253. Harris, M., Cowie, C., & Howi e, L. (1993). Self-monitoring of blood glucose by adults with diabetes in the U.S. population. Diabetes Care 16, 1116-1123. Harris, M., Eastman, R., Co wie, C., Flegal, K., & Eberhar dt, K. (1999). Racial and ethnic differences in glycemic control of adults with type 2 diabetes mellitis Diabetes Care 22, 403-413. Harris, R., & Linn, M. W. (1985). Health beliefs, compliance and control of diabetes mellitus. Southern Medical Journal, 78 162-166. Harris, R., Skyler, J. S., Li nn, M. W., Pollack, L. W., & Tewksbury, D. (1982). The relationship between the health belief model and compliance as a basis for intervention in diabetes mellitus. Pediatric Adolescent Endocrinology, 10 123-132.

PAGE 246

233 Harris, R., Linn, M. W., Skyl er, J. S., & Sandifer, R. (1987). Development of the diabetes health belief scale. Diabetes Education, 13 292-297. Harrison, J. A., Mullen, P. D., & Green, L. W. (1992). A meta -analysis of studies of the health belief model with adults. Health Education Research Theory and Practice, 7 (1), 107-116. Hartwell, S. L., Kaplan, R. M., & Walla ce, J. P. (1986). Comparison of behavioral interventions for contro l of type 2 diabetes mellitus. Behavioral Therapy, 17 447-461. Haynes, R. B., Taylor, D. W. & Sackett, D. L. (1979). Compliance in health care Baltimore: Johns H opkins Press, pp. 1-2. Herman, W. H., & Dasbach, E. J. (1994). Diabetes, health insurance and health--care reform. Diabetes Care, 17 (6), 611-613. Herman, W. H., Teutsch, S. M. & Geiss, L. S. (1987). In closing the gap: The burden of unnecessary illness (Amler, R. W., & Dull, H. B., Eds.). New York: Oxford Univ. Press. Hess, G. E., Davis, W. K ., & Van Harrison, R. (1986) A diabetes psychosocial profile. Diabetes Education, 12 135-140. HgbA1cNow, Metrika,Inc., Professiona l-use product insert. SunnyDale, CA. Hirsch, I., Farkas-Hirsch, R. & Skyler, J. ( 1990). Intensive insulin therapy for type 1 diabetes (Review) Diabetes Care 13, 1265-1283,. Hurley, A. C. (1990). The Health Belie f Model: Evaluation of a diabetes scale.

PAGE 247

234 The Diabetes Educator, 16 (1), 44-48. Hurley, A. C. (In press) Development of the insu lin management diabetes selfefficacy scale. In Waltz, C., & Strickland, O (Eds), Measurement of clinical and educational nursing outcomes New York: Springer Publishing. Ingersoll, G. M., Orr, D. P., Herrold, A. J ., & Golden, M. P. (1986). Cognitive maturity and self-management among adolescents with insulin-dependent diabetes mellitus. Journal of Pediatrics, 108 620-623. International Diabetes Center (2001). Practical applications of diabetes management using staged di abetes management. Staged Diabetes Management Approach. Minneapolis MN.. Irvine, A. A., Saunders, J. T., Blank, M. B. & Carter, W. R. (1990). Validation of scale measuring environmental barrier s to diabetes regimen adherence. Diabetes Care, 13 705-711. Jacobson, A. M., Adler, A. G., Wolfsdorf, J. I., Anderson, B., & Derby, L. (1990). Psychological characteristics of adul ts with IDDM: Comparison of patients in poor and good glycemic control. Diabetes Care, 13 375-381. Jansson, P., Fowelin, J., Smith, U., & Lonnr oth, P. (1988). Characterization by microdialysis of intercellular gluc ose level in subcutaneous tissue in humans. Am J Physiol 255, E218-E220. Janz, N., & Becker, N. (1984). T he health belief model: A decade later. Health Education Quarterly, 11 (1), 1-47.

PAGE 248

235 Jeffery, R. W., et al. (1984) Correlates of weight loss and its maintenance over two years of follow-up among middle-aged men. Preventive Medicine, 13 155-168. Jiwa, F. (1997, Jan-March). Diabet es in the 1990s — An overview. Statistical Bulletin 2-8. Johnson, S. B., Pollak, R. T., Silverstei n, J. H., Rosenbloom, A. L., Spillar, R., McCallum, M., & Harkavy, J. (19 82). Cognitive-behavioral knowledge about insulin-dependent diabetes among children and parents. Pediatrics, 69 708-713. Kaplan, R. L., Ries, A. L., Prewitt, L. M., & Eakin, E. (1994). Self-efficacy expectations predict survival fo r patients with chronic obstructive pulmonary disease. Health Psychology, 30 (4), 366-368. Kaplan, R. M., & Hartwell, S. L. (1987). Differential ef fects of social support and social network on psychological and social outcomes in men and women with type 2 diabetes mellitus. Health Psychology, 6 387-398. Kaplan, R. M., Hartwell, S. L., Wilson, D. K., & Wallace J. P. (1987). Effects of diet and exercise interventions on c ontrol and quality of life in non-insulindependent diabetes mellitus. Journal of General Internal Medicine, 2 220-227. Kaplan, R. M., & Toshima, M. T. (1989). The func tional effects of social relationships in chronic illness and dis ability. In I. G. Sarason, B. R.,

PAGE 249

236 Sarason, & G. R. Pierce (Eds.), Social support: An interactional view. New York: Wiley. Kaplan, R. M. (1990). Behavior as the central outcome in health care. American Psychology, 45 1211-1220. Kasl, S. V., & Cobb, S. (1966). Health behavior, illness behavior and sick role behavior. Archives of Environmental Health, 12 (February, 246-266. Kavanaugh, D. J., Gooley, S., W., & Wilson, P. H. (1993). Prediction of adherence and control in diabetes. Journal of Behavioral Medicine 16, 509-522 Kenny, S. J., Augert, R. E., & Geiss, L. S. (1995). Prevalence and incidence of non-insulin-dependent diabetes. In M. I. Harris, C. C. Cowie, M. P. Stern, E. J. Boydo, G. E. Reiber, & P. H. Bennett (Eds.), Diabetes in America (pp. 47-68) (2nd ed.). Bethesda, MD : National Institutes of Health, National Institutes of Diabetes and Digestive and Kidney Diseases (NIH publ. No. 95-1468). Kingery, P. M., & Glasgow, R. E. (1989). Self-efficacy and outcome expectations in the self-regulati on of non-insulin dependent diabetes mellitus. Health Educator, 20 13-19. Kirscht, J. P. (1974). The health belief model and illness behavior. Health Education Monographs, 2 387-408.

PAGE 250

237 Kirscht, J. P. (1976). Resear ch related to the modification of health beliefs. In Becker, M. H. Ed., The health belief model a nd personal health behavior New Jersey: Charles B. Slack Co. Kirscht, J. P., & Rosenstock, I. M. ( 1977). Patient adherence to antihypertensive medical regimens. Journal of Community Health, 3 115-124. Klein, R., Klein, B. E., Mo ss, S. E., Davis, M. D., & DeMets, D. L. (1984). The Wisconsin epidemiologic study of di abetes retinopathy, III. Prevalence and risk of diabetic retinopat hy when age at diagnosis is 30 or more years. Arch Ophthalmology, 102 527-532. Klein, R. (1995). Hyperglycemia and mi crovascular and macrovascular disease in diabetes. Diabetes Care, 18 258-268. Koenig, R. J., Peterson, C. M., Jones, R. L., Saudek, C., Lehman, M., & Cerami, A. (1976). Correlation of gl ucose regulation and hemoglobin A1c in diabetes mellitus. (1976). New England Journal of Medicine, 285 417421. Korhonen, T., Huttunen, J. K., Aro, A., H entimen, M., Ihalainen, O., Majander, H., Siitonen, O., Uusitupa, M., & Pyorl, K. (1983). A controlled trial on the effects of patient education in the treat ment of insulin-dep endent diabetes. Diabetes Care, 6 256-261. Koro,C.E., Bowlin, S.F., Bourgeois, N., Fedder, D.O..(2004). Glycemic control from 1998 to 2000 among U.S. adults diagnosed with type 2 diabetes.

PAGE 251

238 Diabetes Care. 27,17-20. Krall, L. P. (1985). Educ ation : A treatment for diabetes In Marble A., Krall, L.P., Bradly, R. F., Christliebar Soeldaer, J. S. (Eds.), Joslin's diabetes mellitus (12th ed) (pp. 465-484). Philadelphia: Lea & Febiger. Kurtz, S. M. S. (1990). Adherence to diabetes regimens, empirical status and clinical applications. Diabetes Education, 16 50-56. Laitinen, J. H., Ahola, I. E., Sarkkinen, E. S., Wi nberg, R. L., HarmaakorpiHvonen, P. A., & Usitupa, M. I. (1 993). Impact of intensified dietary therapy on energy and nutrient intake s and fatty acid composition of serum lipids in patients with rec ently diagnosed non-insulin-dependent diabetes mellitus. Journal of American Diet Association, 93 276-283. LaPorte, R. E., Matsushi ma, M., & Chang, Y. F. (1995). Prevalence and incidence of insulin-dependent diabetes. In M. I. Harris, C. C. Cowie, M. P. Stern, E. J. Boydo, G. E. Reiber, & P. H. Bennett (Eds.), Diabetes in America (pp. 37-46) (2nd ed.). Bethesda, MD: National Institutes of Health, National Institutes of Diabet es and Digestive and Kidney Diseases (NIH publ. No. 95-1468). Lawrance, L., & McLeroy, K. R. (1986) Self-efficacy and health education. Journal of School Health, 56 (8), 317-321. Leicher, S. (1986). Diabetes edu cation in hospital settings. Diabetes Educator, 12 277.

PAGE 252

239 Lewis, K. S., Jennings, A. M., Ward, J. D ., & Bradley, C. (1990). Health belief scales developed specifically fo r people with tablet-treated type 2 diabetes. Diabetic Medicine, 7 148-155. Littlefield, C. H., et al. (1992). Rela tionship of self-efficacy and binging to adherence to diabetes regi men among adolescents. Diabetes Care, 15 90-94. Litzelman, D. K., Slemenda, C. W., Langefeld, C. D., Hays, L. M., Welch, M. A., Bild, D. E., Ford, E. S., & Vinicor, F. (1993). Reduction of lower extremity clinical abnormalities in patients wit h non-insulin-dependent diabetes. Ann Internal Medicine, 119 36-41. Lockington, T. J., Powles, S., Meadows, K. A., & Wise, P. H. (1989). Attitudes, knowledge and blood gl ucose control. Diabetes Medicine, 6 309-313. Ludvigsson, J>, & Hanas, R. (2003) Continuous subcutaneous glucose monitoring improved metabolic control in pediatric patients with type 1 diabetes: a controlled crossover study. Pediatrics, 111, 933-938. Maiman, L. A., & Becker, M. H. (1974) The health belief model: Origins and correlates in psychological theory. Health Education Monographs, 2 336353. Maljanian, R., Grey, N., Sta ff, I., Conroy, L. (2005). Intensive Telephone followup to a hospital-based disease management model for patients with diabetes mellitus. Disease Management 8 (1): 15-25.

PAGE 253

240 Manson, J. E., Rimm, E. B., Stampfer, M. J., Colditz, G. A., Willett, W. C., Krolewski, A. A., Rosner, B., Hennekens, C. H., & Speizer, F. E. (1991). Physical activity and incident of non-in sulin-dependent diabetes mellitus in women. Lancet, 338, 774-778. Marquis, K. H., & Wa re, J. E. (1979). Measures of diabetic patient knowledge, attitudes, and behavior regarding self-care (CDC Publication Nos. R-2480HEW, N-1151-HEW, N-1152-HEW, & N1153-HEW). Washington, DC: U.S. Government Printing Office. Mausner, J. S., & Bahn, A. K. (1974). Epidemiology: An introductory text Philadelphia: W. B. Saunders. Mazze, R. S., Etzwiler, D. D., Strock, E., et al. (1994) Staged diabetes management: Toward an integrated model of diabetes care. Diabetes Care, 17 (suppl 1), 56-66. Mazzuca, S. A., Moorman, N. H., Wheeler, M. L. Nort on J. A., Fineburg, N. S., Vinicor, F., Cohen, S. J., & Clark, C. (1986). The diabetes education study: A controlled trial of the effe cts of diabetes patient education. Diabetes Care, 9 1-10. McCaul, K. D., Glasgow, R. E., & Schaf er, L. D. (1987). Diabetes regimen behaviors: Predicting adherence. Medical Care, 29 868-881. McLaughlin, G. H. (1969). J Reading, 12 639-646. In Doak, C. C., Doak, L. G., & Root, J. H., (1985), Teaching patients with low literacy skills

PAGE 254

241 Philadelphia, PA: JB Lippincott. McNeal, B., Salisbury, A., Baumgar dner, P., & Wheeler, F. (1984). Comprehension assessment of diabetes education pr ogram participants. Diabetes Care, 7 232-235. Miller, V., & Goldstein, J. (1972). More efficient care of diabetic patients in a country hospital setting. New England Journal of Medicine, 286 13881391. Miller, V., Goldstein, J., & Nicolaisen, G. (1978). Evaluation of patient's knowledge of diabetes self-care. Diabetes Care, 1 275-280. Moos, R. H. (1987). The social climate scales: A user’s guide Palo Alto, CA: Consulting Psychology Press. Mulhauser, I., Jorgens, V., Berger, M., Graninor, W., Gurtler, W., Hornke, L., Kunz, A., Schernthaner, G., Scholz, V., & Voss, H. E. (1983). Bicentric evaluation of a teaching and treatment program for type 1 (insulindependent) diabetic patients: Improvement of metabolic control and other measures of diabetes care for up to 22 months. Diabetologia, 25 470476. Mulhauser, I., Bruckner, I., Berger, M., Cheta, D., Jorgens, V., LonescuTirgoviste, C., Scholz, V., & Mineu, L. (1987). Evaluation of an intensified insulin treatment and teaching program as routine management of type 1 (insulin-dependent) diabetes. Diabetologia, 30 681-690.

PAGE 255

242 Mulhauser, I., & Berger, M. (1993). Diabetes education and insulin therapy: When will they ever learn? Journal of Internal Medicine, 233 321-326. Nathan, D. M., Singer, D. E., Godine, J. E., Harrington, C. H., & Permuter, L. C. (1986). Retinopathy in older type 2 di abetics: Association with glucose control. Diabetes, 36 797-801. National Diabetes Adviso ry Board. (1983). The prevention and treatment of five complications of diabetes (DHHS Publication No. 83-8392). Atlanta, GA: Center for Disease Control. Norris, S.L.Lau, J., Smith, S.J., Schmid, C.H., Engel gau, M.M. (2002). Selfmanagement education for adults with type 2 diabetes: a meta-analysis of the effect on glycemic control. Diabetes Care, 25, 1159-1171. Ohkubo, Y., Kishikawa, H., Araki, E., Miya ta, T., Isami, S., Mo toyoski, S., Kojima, Y., Furuyoshi, N., & Shichiei, M. (1995). Intensive insulin therapy prevents the progression of diabetic microvascular complications in Japanese patients with non-insulin -dependent diabetes mellitus: A randomized prospective 6-year study. Diabetes Research and Clinical Practice, 28 103-117. Padgett, D., Mumford, E., Hynes, M., & Cart er, R. (1988). Meta-analyzes of the effects of educational and psychologi cal interventions on management of diabetes mellitus Journal of Clinical Epidemiology, 41 1007-1030. Paulozzi, L. J., Norman, J. E., McMahon, P., & Connel, F. A. (1984). Outcomes

PAGE 256

243 of a diabetes educ ation program. Public Health Report, 99 575-579. Peters, A., Legorreta, A., O ssorio, R., & Davidson, M. (1996). Quality of outpatient care provided care to di abetic patients: a health maintenance organization experience. Diabetes Care. 19, 601-606. Peters, A. L..,Davidson, M.B. (1998). Applic ation of a diabetes m anaged care program: the feasibility of using nurses and a com puter system to provide effective care. Diabetes Care 21 (7): 1037-1043. Peyrot, M., & Rubin, R. R. (1988). Insulin self-regulation predicts better glycemic control. Diabetes, 37 (suppl 1), 53. Peyrot, M., & Rubin, R. R. (1994). Modeling the effe ct of diabetes education on glycemic control. Diabetes Education, 20 (2), 143-148. Pham, D. T., Forton, F., & Thibaudeau, M. F. (1996). The role of the health belief model in a patient’s self-eva luation of adherence to diabetes selfcare behaviors. The Diabetes Educator, 22 (2), 126-132. Pierce, C. (1995). Advancing the frontie rs of diabetes research and treatment. In K. Bellenir, & P. D. Dresser (Eds.). Diabetes sourcebook (pp. 561567). Detroit, MI: Omnigraphics. Pirart, J. (1978). Diabetes mellit us and its degenerative complications: A prospective study of 4400 patients observed between 1947 and 1973. Diabetes Care, 1 168-188, 252-263. Polonsky, W. H., Anderson, B. J., Lohrer P. A., Welch, G., Jacobson, A. M.,

PAGE 257

244 Aponte, J. E., & Schwartz, C. E. (1995). Assessment of diabetes-related distress. Diabetes Care, 18 (6), 754-760. Posavoc, E. J. (1980). Evaluations of patient education programs: A metaanalysis. Evaluation and the Health Professions, 3 47-62. Ray, N. F., Willis, S., & T hamer, M. (1993). Direct and indirect costs of diabetes in the United States in 1992. Alexandria, VA: American Diabetes Association. Ray, N. F., Thamer, M., Taylor, T., F ehrenbach, S. N., & Ra tner, R. (1996). Hospitalization and expenditures for t he treatment of general medical conditions among the U.S. di abetic population in 1991. Journal of Clinical Endocrinology and Metabolism 1 (10), 3671-3679. Raz, I., Soskolne, V., & Stein, P. (1988). Influence of small-group education sessions on glucose homeostasis in NIDDM. Diabetes Care, 11 67-71. Reitman, J. S., Vasquez, B., Kilmes K., & Naguelesparan, M. (1984). Improvement of glucose homeostasis after exercise in NIDDM. Diabetes Care, 7 434-441. Rettig, B. A., Shrauger, D. G., Recker, R. R., Gallagher, T. F., & Wiltse, H. (1986). A randomized study of the e ffects of a home diabetes education program. Diabetes Care, 9 173-178. Richardson, W. (1970). M easuring the urban poor’s use of physicians’ services in response to illness episodes. Medical Care, 8 132.

PAGE 258

245 Rodin, J. (1990). Physiological effects of exercise. In Willi am, R. S., & Wallace, A. G., Eds., Biological effects of physical activity Champaign, IL: Human Kinetics. Rosenstock, I. M. (1960) What research in motivation suggests for public health. American Journal of Public Health, 50 295-302. Rosenstock, I. M. (1966). Why people use health services. Milbank Memorial Fund Q, 44 94-127. Rosenstock, I. M. (1974). Historica l origins of the health belief model. Health Education Monographs, 2 328-335. Rotter, J. B. (1966). Gener alized expectancies for internal and external control of reinforcements. Psychological Monographs, 80 1-28. Rubin, R. J., Altman, W. M., & Mendelson, D. N. (1994). Health care expenditures for people with diabetes mellitus. Journal of Clinical Endocrinology & Metabolism, 78 809A-809F. Rubin, R. R., Peyrot, M., & Saudek, C. D. (1989). Effects of diabetes education of self-care, metabolic contro l, and emotional well-being. Diabetes Care, 12 673-679. Rubin, R. R., Peyrot M., & Saudek, D. C. (1991). Differential effect of diabetes education on self-regulation and life-style behaviors. Diabetes Care, 14 335-338.

PAGE 259

246 Rubin, R. R., Peyrot, M., & Saudek, D. C. (1993). The effect of a diabetes education program incorpor ating coping skills training on emotional wellbeing and diabetes self-efficacy. Diabetes Education, 19 210-214. Ruggiero, L., Glasgow, R. E., Dryfoos, J. M., Rossi, J. S ., Prochaska, J. O., Orleans, C. T., Prokhorov, A. V., Rossi S. R., Green, G. W., Reed, G. R., Kelly, K., Chrobanian, L, & Johnson, S. (1997). Diabetes selfmanagement: Self-reported recommendations and patterns in a large population. Diabetes Care, 20 (4), 568-576. Salazar, M. K. (1991). Comparison of four behavioral t heories: A literature review. AAOHN Journal, 39 (3), 128-135. SAS user’s guide (1979). Raleigh, NC: SAS Institute Schafer, L. C., Glasgow, R. E., McCaul, K. D., & Dreher, M. (1983). Adherence to IDDM regimens: Relationship to psychosocial variables and metabolic control. Diabetes Care, 6 493-498. Schafer, L. C., McCaul, K. D., & Glas gow, R. E. ( 1986). Supportive and nonsupportive family behaviors: Relationships to adherence and metabolic control in persons with Type 1 diabetes. Diabetes Care, 9 (2), 179-185.

PAGE 260

247 Schneider, S., Knachadurian, A., Amorosa, L., Clemow, L. & Ruderman, N. (1992). Ten-year experience with ex ercise-based outpatient life-style modification program in the tr eatment of diabetes mellitis. Diabetes Care 15, 1800-1810. Schlenk, E. A., & Hart, L. K. (1984) Relationship between health locus of control, health value, and social support and compliancy among people with diabetes mellitus. Diabetes Care, 7 566-574… Schulze, M.B.,& Hu, F.B. (2005). Prim ary prevention of diabetes :What can be done and how much can be prevented. Annual Review of Public Healh, 26,445-467. Service, F., O’Brien, P., Wise, S., Ne ss, S., & LeBlanc, S. (1997). Dermal interstitial glucose as an i ndicator of ambient glycemia.. Diabetes Care ,20 (8), 1426-1429. Skyler, J. S., Seigler, D. E., & Reeves, M. L. (1982). A comparison of insulin regimens in insulin-de pendent diabetes mellitus. Diabetes Care, 5 (suppl 1), 11-18. Snyder, (1996). Group Practice Journal July-Aug, 11. Stamler, J., Vaccaro, O., Neaton, J. D., & Wentworth, D. (1993) Diabetes, other risk factors, and 12-yr cardiovascular mortality for men screened in the Multiple Risk Factor Intervention Trial. Diabetes Care, 16 434-444. Steed, R.D., Davidson, P. ( 1988, May). Therapeutic strat egies in the patient with

PAGE 261

248 uncontrolled diabetes mellitus. Atlanta Diabetes Associates Strecher, V. J., DeVillis, B. M., Becker, M. H., & Rosens tock, I. M. (1986). The role of self efficacy in achi eving health behavior change. Health Education Quarterly, 13 (1), 73-91. Street, R. L., Pizia, V. C., Carpentier, W. S., Herzog J., Hegil, J., Skinner, G., & McLellan, L. (1993). Patient pr ovider communication and metabolic control. Diabetes Care, 16 714-721. Strine T.W., Okoro, C.A., Chapman, D.P., Beckles, G.L., Balluz, L., Mokdad, A.H. (2005). The impact of formal diabetes education on the preventive health practices and behaviors of persons with type 2 diabetes. Preventive Medicine 41(1): 79-84. Teza, S. L., Davis, W. K ., & Hiss, R. G. (1988). Patient knowledge compared with national guidelines for diabetes care. Diabetes Education, 14 207211. Toobert, D. J., & Glasgow, R. E. (1991) Problem solving and diabetes selfcare. Journal of Behavioral Medicine, 14 71-86. Toshima, M. T., Kaplan, R. M., & Ries, A. L. (1990). Experimental evaluation of rehabilitation in chronic obstructive pulmonary disease: Short-term effects on exercise endurance and health status. Health Psychology, 9 237-252. Trotta., P. (1980). Breast self-exami nation factors influencing compliance. Oncology Nursing Forum, 7 13-17.

PAGE 262

249 Trovati, M., Carta, W., Cavalot, F., Vitali, S., Banaudi, C., Greco Lucchina P., Fiocchi F., Emmanuelli, G., & Lenti, G. (1984). Influences of physical training on blood glucose control, glucos e tolerance, insulin secretion, and insulin action in NIDDM. Diabetes Care, 7 416-420. UK Prospective Diabetes Study Gr oup (UKPDS). (1998). Intensive bloodglucose control with sulphonylureas or insulin compared with conventional treatment and risk of complications in patients with type 2 diabetes (UKPDS33). Lancet, 352 837-852. U.S. Goverment Printing Office. (1997). Balanced Budget Act of 1997. (publ no.869-033-00034-1). 115-116 Vanninen, E., Uusitupa, M., Siitonen, O., Laitinen, J., & Lansimies, E. (1992). Habitual physical activity, aerobic capacity and metabolic control in patients with newly-diagnosed type 22 ( non-insulin-dependent) diabetes mellitus: Effect of 1-year di et and exercise intervention. Diabetologia, 35 340-346. VIA Medical Cor poration. (1998). Blood Chemistry Moni tor for Glucose VIA VGLU1 Operator’s Manual San Diego, CA. 1-46. Vinicor, F., Cohen, S. J., Mazzuca, S. A. Moorman, N., Wheeler, M., Kuebler, T. Swanson, S., Ours, P., Fineburg, S. E., Gordon, E. E., Duckworth, W., Norton J. A., Fineberg, N. S. & Clark, C M. (1987). DIABEDS: A randomized trial of the effects of physician and/or patient education on

PAGE 263

250 diabetes patient outcomes. Journal of Chronic Diseases, 40 345-356. Vonesh, E. F., Chinch illi, V. M. (1997). Linear and nonlinear models for the analysis of repeated measurements New York: Marcel Dekker. Voss, E.M., Sembrowski, G.S., Clasen, B.L., Spencer, M.L., Ainslie, M.B., Haig,B. (1992). Evaluation of capi llary collection system for HgA1c specimens. Diabetes Care, 15, 700-701. Wagner, E..H., Sandhu, N., Newton, K.M., McColloch, D.K., Ramsey, S. D., Grothaus, L.C. (2001). Ef fect of Improved Glyce mic Control on Health Care Costs and Utilization. JAMA, 285: 182-189. Waller, D. A., Chipman, J. J., Hardy, B. W., Hightower, M. S., North, A. J., Williams, S. B., & Babick, A. J. (1986). Measuring diabetes-specific family support and its relation to metabo lic control: A preliminary report. Journal of American Academy of Child Psychiatry, 25 (3), 415-418. Wallston, B. S., & Wallston, K. A. (1984). Social p sychological models of health behavior. An examination and integration. In Baum A., Taylor, S. E., & Singer, J. E. (Eds.). Handbook of psychology and hea lth; Vol. 4: Social Psychological aspects of health (pp. 23-53). Hillsdale, NJ: Lawrence Erlbaum. Watkins, J. D., Williams, T. F., Martin, D. A., Hogan, M. D., & Anderson, E. (1967). A study of diabetic patients in the home. American Journal of Public Health, 57 452-457.

PAGE 264

251 Wdowik, M. J., Kendall, P. A., & Harri s, M. A. (1997). College students with diabetes: Using focus groups and interviews to determine psychosocial issues and barriers to control. The Diabetes Educator, 23 (5), 558-562. Weiner, J. P., Parente, S. T., Garnick, D. W., Fowles, J., Lawthers, A. G., & Palmer, H. (1995). Vari ation in office-based qualit y: A claim-based profile of care provided to medicare patients with diabetes. JAMA, 273 (19), 1503-1508. Westerhall, S. F., Olson, D. R., & Destafano, F. (1992) Trends in diabetes and diabetes complications. 1980-1987. Diabetes Care, 15 960-967. Wierenga, M. E. (1980). The interrelati onship between multidimensional health locus of control, knowledge of dia betes, perceived social support, selfreported compliance and therapeutic outcomes 6 weeks after the adult patient has been diagnosed wit h diabetes mellitus. Dissertation Abstracts International, 40 5610B. Williams, T. F., Martin, D. A. Hogan, M. D., Watkins, J. D ., & Ellis, E. V. (1967). The clinical picture of diabetes control studi ed in four settings. American Journal of Public Health, 57 441-451. Wilson, W. W., Ary, D. V., Biglan, A., Glasgow, R. E., Toobert, D. J., & Campbell, D. R. (1986). Psychosocial predictors of self-care behaviors (compliance) and glycemic control in non-insulin-dependent diabetes mellitus. Diabetes Care, 9 614-622.

PAGE 265

252 Winburger, M., Cowper, P. A., Kirkman, M. S., & Vinicor, F. (1990). The economic impact of diabetes mellitus on the elderly. Clinical Geriatric Medicine, 6 959-970. Wing, R. R., Epstein, L. H., Norwalk, M. P., Scott, N., & G ooding, W. (1987). Family history of diabetes and its e ffect on treatment outcome in type 2 diabetes. Behavior Therapy, 18 283-289. Wing, R. R., Blair, E. H., Cononi, P., Ma rcus, M., Watanabe, R ., & Bergman, R. (1994). Caloric restriction per Schsi sches Edelstahlwerke is a significant factor in improvements in glycemic control and insulin sensitivity during weight loss in obese NIDDM patients. Diabetes Care, 17 30-36. Wooldridge, K. L., Wallston, K. A., Graber, A. L., Brown, A. W., & Davidson, P. (1992). The relationships between health beliefs, adherence and metabolic control of diabetes. The Diabetes Educator, 18 (6), 495-501.

PAGE 266

253 Appendices

PAGE 267

254 Appendix 1: Assessment and Education Records and Baseline Evaluation Form Assessment and Education Records BASELINE EVALUATION FORM IDENTIFYING INFORMATION Date of Evaluat ion: _____/_____/_____ (mm/dd/yy) Patient's Physician type: ________ (1) Endocrinologist ________ (2) Primary Care/Other Patient Surname: Patient's ID/SS Number: Date of Birth: ____/_____/______ (mm/dd/yy) If age <35> 70, patient is ineligible Gender: ______ (1) Male ______ (2) Female Race: ______ (1) White ______ (2) Black ______ (3) Hispanic ______ (4) Other. Please specify: ____________ Education: ______ (1) Less than high school education ______ (2) Graduated high school or higher education

PAGE 268

Appendix 1: (Continued) 255 PERTINENT MEDICAL HISTORY Type of Diabetes: _______Type I (ineligible) _______Type II-non-insulin _______Type II-insulin _______Secondary diabetes (ineligible) History of Gestational diabetes ________ (ineligible) Duration since Diabetes Diagnosis: ________ (years) History of any hemoglobinopathy or hemolytic process: ________ (1) Yes (If yes, ineligible) ________ (2) No ________ Height (cm) ________ Weight (Kg) ________ BMI Other illnesses (check all that apply): ________ Heart Disease ________ Lung ________ Cancer ________ Hypertension ________ Stroke ________ Alzheimer’s ________ Other Specif y _____________________

PAGE 269

Appendix 1: (Continued) 256 DOCUMENTED DIABETES COMPLICATIONS Eyes ________ Yes ________ No Specify: Amputation ________ Yes ________ No Specify: Kidney ________ Yes ________ No Specify: Nerves ________ Yes ________ No Specify: Circulation ________ Yes ________ No Specify: ER Visits ________ Yes ________ No Specify: Hospitalization within last 2 years ________ Yes ________ No Specify: Vision Impaired ________ Yes (ineligible) ________ No Specify:

PAGE 270

Appendix 1: (Continued) 257 Mobility Impaired ________ Yes (ineligible) ________ No Specify: On dialysis ________ Yes (ineligible) ________ No Specify: Medical Clearance for Exercise ________ Yes (ineligible) ________ No Specify: TYPE OF TREATMENT FOR DIABETES _____ (1) Diet only. Specify type ________________________ __________ _____ (2) Diet and oral hypoglycemic agent (OHA) pills. Specify type/amount _____ (3) Oral hypoglycemic agents (Diabetic pills) only. Specify type/amount _____ (4) More than 2 doses of OHA's pills or more than 1 brand of OHA pills/day and diet. Specify type ________________________ __________ ______(5) Insulin Therapy Specify ________________________ __________

PAGE 271

Appendix 1: (Continued) 258 _____ (5) Date started on OHA ___/____/____ (mm/dd/yy) _____ (6) Date started on insulin ___/____/____ (mm/dd/yy) _____ (7) Date of last diabetes medicine change ___/____/____ (mm/dd/yy) Glycosylated Hemoglobin Value: ______ % and date ___/____/____ (mm/dd/yy) Expanded HBMQ Scales: Date ____/____/____ (mm/dd/yy) Expanded HBM Subscales Mean Scores Severity _____________ Susceptibility _____________ Benefits _____________ Barriers _____________ Health Motivation _____________ Self-efficacy _____________ Cues to Action _____________ Structural El ements _____________

PAGE 272

Appendix 1: (Continued) 259 For returning controls (8 week) Hgb A1c _______% Date ___/___/___ (mm/dd/yy) Last appointment with physician who manages your diabetes Date ___/___/___ (mm/dd/yy) Any changes in medical management over past 8 weeks? ________ Yes ________ No Specify: Date ___/___/___ (mm/dd/yy) Have you changed any aspect of your diabetes management over the past 8 weeks? ________ Yes ________ No Specify: Date ___/___/___ (mm/dd/yy) Previous Diabetes Education ________ Yes ________ No Specify: Date ___/___/___ (mm/dd/yy)

PAGE 273

260 Appendix 2: Sample Flyer for Advertising Program Purpose of Study To determine if persons with type 2 diabetes participating in diabetes self-management training assigned to an intensive follow-up group achieve better diabetes control than those assigned to the standard followup group. A finger stick Hemoglobin A1c will be performed to assess blood glucose control and eligibility for the study (> 7.0 < 13). Hemoglobin A1c is a measurement of long-term blood glucose control that assesses how your blood glucose control has been over the past 8-12 weeks. Previous studies have shown that by keeping HgbA1c level near normal can prevent or delay complications of diabetes such as blindness, nerve damage, kidney disease, and amputations. Who Can Participate? People age 18-70 with a diagnosis of Diabetes Mellitus, Type 2 People who have not attended a diabetes education program in the past year People willing to attend approximately 10-14 hours of education sessions over a 24 week period People who are physically able to care for their diabetes. People willing to use a glucose monitor and or sensor Proposed Benefits of Participating in Study: Improved blood glucose control Decreased risk of diabetes complications Learning new skills for diabetes management Your Hemoglobin A1c level will be tested. Education Will Cover Meal planning Exercise Medication Behavioral change Coping with diabetes/stress management Prevention of complications Blood glucose monitoring Special situation Trouble shooting Convenient scheduling Diabetes Education and supplies covered by most insurance Small group and individualized counseling No charge for follow-up Blood glucose monitoring, HgbA1c and follow-up sessions.

PAGE 274

Appendix 2: (Continued) 261 FOR MORE INFORMATION Call Jodee MeisenhelderSmith, RD @ The Diabetes Care Institute @ University Community Hospital 813-615-7751 DIABETES CARE INSTITUTE 3100 EAST FLETCHER AVENUE TAMPA, FLORIDA 33613-4688 Need Help Managing Your Diabetes? Volunteers Are Needed to Participate in the Diabetes Care Institute’s Research Study

PAGE 275

262 Appendix 3: Informed Consent Form INFORMED CONSENT FORM (ADULT) Study Title: The Effects of ADA Comp rehensive Diabetes Self-Management Training and Continuous Glucose Monito ring on Diabetes Health Beliefs, Behaviors and Metabolic Control Protocol No. Sponsor: Primary Investigator: Jodee Meisenhelder-Smith, MNS, CNS, CDE, RD, LD Institution (s): Univer sity of South Florida University Community Hospital, Diabetes Care Institute

PAGE 276

Appendix 3: (Continued) 263 Introduction: You are being invited to participate becaus e you have Type 2 Diabetes Mellitus The study will focus on reducing Hemoglobin A1c levels, a measurement of blood glucose that assesses how your blood gl ucose control has been over the past 8 – 12 weeks. Previous studies have shown that by keeping Hgb A1c levels near normal, diabetes complications can be delayed or prevented. Purpose: The primary purpose is to determine if persons with Type 2 diabetes participating in ADA diabetes self-car e management training program who have intensive follow-up achieve better control over their blood glucose levels than those receiving standard follow-up.

PAGE 277

Appendix 3: (Continued) 264 INFORMED CONSENT FORM Adult Informed Consent Diabetes Care Institute @ University Community Hospital and/or University of South Florida Information for People Who Take Part in Research Studies The following information is being present ed to help you decide whether or not you want to be a part of a research study. Please read carefully. Anything you do not understand, ask the doctor. Title of Study: The Effects of ADA Comprehensive Diabetes SelfManagement Training and Continuous Glucose Monitoring on Diabetes Health Beliefs, Behaviors and Metabolic Control Primary Investigator: Jodee Meisenhelder-Smith, MNS, CNS, CDE, RD LD Study Location(s): Diabetes Care Institute @ University Community Hospital.

PAGE 278

Appendix 3: (Continued) 265 General Information about the Research Study The purpose of this research study is to determine if persons with type 2 diabetes participating in a diabetes self-care management training progr am with intensive follow-up achieve better control over their blood glucose levels than those in the standard follow-up control group. You are being asked to consent to participate because the study will focus on reducing hemoglobin A1c values which show the average blood sugar over the past three months in adults aged 35-70 with type 2 diabetes. In order to test the effectiveness of the diabetes education, you will be randomly selected to receive one of two or more different treatment plans. Randomization is similar to a flip of a coin. Your chanc es of being put in either or any group is about the same. All participants wil l receive the education program. You may want a friend or family member to read the form and talk to the study doctor with you. You can also talk to your personal doctor about what you should do. Talking things over can help you make the right choice. The time you will need to spend in this re search study will be about: 16 hours. At

PAGE 279

Appendix 3: (Continued) 266 the initial visit and after completing t he education sessions, you will be asked to complete a questionnaire. The questionnaire takes about 30 minutes to complete and will ask questions about your diabetes care and feelings about your health and diabetes. The number of other people that might take part in this st udy at this local site is: approximately 159. Plan of Treatment Your regular medical treatment will not actually be part of the research study but will prepare you for the study. The experimental treatment that you will receive by taking part in this research study is: You will be seen as an outpatient at t he clinic several times over a 24 week period. This program differs from standard care for this disorder in the following way(s): Participants will receive 10 hours of education sessions with diabetic educators covering such topics as diabetes overview, nutrition and exercise gu idelines, medications, acute and chronic complications, glucose and ot her monitoring, stress management, and behavioral goal setting skills. During your participation, a capillary blood sample will be removed from you and analyzed, and used by the

PAGE 280

Appendix 3: (Continued) 267 investigators and/or the sponsors. By signing this consent form, you agree to allow the investigators to analyze and use your blood as described and for the purpose stated. Benefits of Being a Part of this Research Study We cannot tell whether you will benefit from taking a Diabetes self-care management training program because its effe cts on your disease are not totally understood. On the other hand, by taking par t in this research study, you may increase our overall knowledge of your dis ease and how to treat future patients. In addition, the program is designed to hel p patients improve t heir overall glucose control and in so doing reduce their risk of diabetic complications. Risks of Being a Part of this Research Study You may have side effects from regular treat ment of your disease. This treatment would be medication to control blood sugar levels, blood glucose monitoring, and diet. The possible side effects of this regular treatment ar e listed below along with your chances of having them and their seriousness compared to your disease.

PAGE 281

Appendix 3: (Continued) 268 There are no known risks associated with participating in a diabetes self-care management training program. The testing of Hgb A1c and blood sugars requires one (1) drop of blood drawn from the s ubject using standard capillary lancets used in every day care of diabetes. Alternatives of Being Part of this Research Study Alternatives of being part of this resear ch study is/are: Routine care. Another alternative is to not participate in this study. Payment for Being a Part of this Research Study You will not receive cash or other gifts for taking part in this research study. However, Florida law requires most insu rance to cover the cost of diabetes education and supplies. All eligible patients will rece ive the diabetes self-care management training program Costs of Being a Part of this Research Study The diabetes self-care m anagement training program and diabetes supplies such as lancets, needles and glucose monitori ng strips are covered by the patient’s insurance. Participants in the study will not incur any additional cost for follow-up care. Hgb Aic are perform ed free of charge. You will be provided a writt en list of procedures required because of the research

PAGE 282

Appendix 3: (Continued) 269 study. One of the pers ons in charge of the research will discuss these procedures with you, and will assist you in making sure prior to participation that your insurance covers your costs fo r regular diabetes care. Most costs mentioned above are covered by your insu rance. The DTC @ UCH is available to help you with insurance coverage info rmation. You may contact the primary investigator, Jodee Meisenhelder-Smith at (813) 632-7751; Beeper (813) 2270904. In Case of Illness or Injury Call your referring physician and the pr imary investigator Jodee MeisenhelderSmith, at 813-632-7751 in the event you get sick or injured while on this research study. If you have an emergency, go to the closest emergency room or clinic for treatment. Available Medical Treatment This study is being conducted by Jodee Meisenhelder-Smith, MNS, CNS, CDE, RD, LD on patients from University Comm unity Hospital. University Community Hospital reviews research studies through its Institutional Review Board but is not an investigator in this study and does not supervise or direct the study. If you experience a side effect or injury, and if emergency treatment is required, immediately contact Jodee Meisenhelder-S mith at (813) 632-7751, Beeper (813)

PAGE 283

Appendix 3: (Continued) 270 227-0904. University of South Florida Injury Statement In the event that you sustain an injury or illn ess as a result of participating in this research, please be aware that medical tr eatment for the inju ries or illness may not be available from the University of South Florida (USF). USF does not maintain an emergency department nor does it provide medical treatment in all disciplines of medicine. If you become ill or sustain an injury which you believe is related to participation in this research, immediately contact one of the persons listed on page 1 of this form, and if emergency care is needed seek emergency attention from your nearest local hospital. If injury results from y our participation in research, money damages are not automatically available. Money damages are only av ailable to the extent specified in Florida statut e, 768.28. A copy of this statute is available upon request to the Division of Compliance Servic es USF. This statute provides that damages are available only to the extent that negligent conduct of a University employee caused your injuries, and are lim ited by law. If you believe you are injured as a result of par ticipation in this research and the negligent conduct of a University faculty member, you may not ify the USF Self Insurance Programs (813) 974-8008, who will in vestigate the matter.

PAGE 284

Appendix 3: (Continued) 271 Confidentiality of Your Records Your research records will be kept confident ial to protect your privacy to the full extent of law. However, authorized res earch investigators, agents of the United States Food and Drug Administration, the Department of Health & Human Services and the USF Institutional Review Board may inspect your records from this research project. Doctors, nurses and others involved with your care will also be able to see the research info rmation in your medical record. The results of this research study may be published, but they will not include your name or any other information that may identify you. Volunteering to Be Part of this Research Study You should only take part in this resear ch study if you want to and not because you are afraid of losing medical benefit s. If you decide you want to stop taking part in the study tell a study monitor as soon as possible. They will want to tell you if there are any dangers in stopping treatment. If you decide to stop, any other suitable treatment for your disease that may exis t will be offered to you. You may be removed from the study wit hout your consent for non-compliance with the diabetes education sessions.

PAGE 285

Appendix 3: (Continued) 272 Questions and Contacts If you have any questions about this resear ch study, contact the investigator listed on the first page at USF/DTC @ UCH at (813) 632-7751, Beeper (813) 227-0904. If you have questions about your rights as a person who is taking part in a research study, you may contact a me mber of the Divis ion of Compliance Services of the University of South Florida at 813-631-4498. Your Consent—By signing this form I agree that: I have fully read or have had explained to me in my native language this informed consent form describing a research project. I have had the opportunity to question one of the persons in charge of this research and have received satisfactory answers. I understand that I am being asked to participate in research. I understand the risks and benefits, and I freely give my consent to participate in the research project outlined in this form, under the conditions indicated in it. I have been given a signed copy of this informed consent form, which is mine to keep.

PAGE 286

Appendix 3: (Continued) 273 ________________________ _____________________ ____________ Signature of Participant Date Printed Name of Participant ________________________ _____________________ ____________ Signature of Witness Date Printed Name of Witness Investigator Statement I have carefully explained to the subjec t the nature of the above protocol. I, hereby, certify that to the best of my k nowledge the subject signing this consent form understands the nature, demands, risks and benefits involved in participating in this study and that a m edical problem or l anguage or educational barrier has not precluded a clear understanding of the participant’s involvement in this study.

PAGE 287

Appendix 3: (Continued) 274 ________________________ _____________________ ____________ Signature of Primary Investigator Date Printed Name of Pr imary Investigator I may contact Jodee Meisenhelder-Smith at (813) 632-7751, Beeper (813) 2270904 if I have any further questions or suffer a research related injury. I understand that if I become pregnant I s hould notify the researcher and I should not participate in any other diabete s research while on this study. ________________________ _____________________ ____________ Signature of Participant Date Printed Name of Participant ________________________ _____________________ ____________ Signature of Witness Date Printed Name of Witness

PAGE 288

Appendix 3: (Continued) 275 Institutional Approval of Study and Informed Consent The research study and Informed Consent Form were reviewed and approved by the University of South Florida Institut ional Review Board for the protection of human subjects. This approval is valid until the date provi ded below. The board may be contacted at (813) 631-4498. Approval Consent Form Expiration Date: (Stamp date here.) Revision Date:_______

PAGE 289

Appendix 3: (Continued) 276 Appendix 4: Study Forms REVISED DIABETES HEALTH BELIEF AND ENVIRONMENTAL BARRIER SCALE NAME_____________________ ____DATE___/___/___ Directions: We are interested in your re sponses to the following health-related questions. Circle one of the numbers 1-4 after each question that best describes how you feel. 1. How likely are you to take y our temperature when you feel sick? 1. not at all likely 2. somewhat likely 3. likely 4. very likely 2. How important do you think it is to get a health checkup even when you feel ok? 1. not at all important 2. somewhat important 3. important 4. very important

PAGE 290

Appendix 4: (Continued) 277 3. How much do you feel you underst and your treatment for diabetes? 1. not at all 2. a little 3. somewhat 4. very much 4. How much would you say your diet gets in the way of your daily living? 1. not at all 2. a little 3. somewhat 4. very much 5. How helpful to you is information about your diet? 1. not at all 2. a little 3. somewhat 4. very much

PAGE 291

Appendix 4: (Continued) 278 6. How helpful is testing your own bl ood sugars at home for diabetes control? 1. not at all helpful 2. a little helpful 3. somewhat helpful 4. very much helpful 7. To what extent do you believe that exercise can lower your blood sugar levels? 1. not at all 2. a little 3. somewhat 4. very much 8. How helpful is a meal plan for c ontrol of your blood sugar levels? ? 1. not at all helpful 2. a little helpful 3. somewhat helpful 4. very much helpful

PAGE 292

Appendix 4: (Continued) 279 9. How likely is it that di abetes will shorten your life ? 1. not at all likely 2. a little likely 3. somewhat likely 4. very much likely 10. How much do you think your docto r can help you to achieve a longer, healthy with your diabetes? 1 not at all 2 a little 3 some 4 a lot 11. How likely are persons with diabetes to have numbness or tingling in their arms or legs? 1 not at all 2 somewhat 3 easily 4 very likely

PAGE 293

Appendix 4: (Continued) 280 12. How much do your family and close friends help you to stay on your diet? 1 not at all 2 a little 3 somewhat 4 a lot 13. How helpful would you say an educ ational program is for persons with diabetes? 1 not at all helpful 2 a little 3 somewhat 4 very much 14. How much do you worry about what you eat? 1 not at all 2 a little 3 some 4 a lot

PAGE 294

Appendix 4: (Continued) 281 15. How likely would sym ptoms of low blood sugar such as cold sweats, weakness, and headache lead you to seek medical help? 1 not at all likely 2 somewhat likely 3 likely 4 very likely 16. To what extend do you believe diabet es-related kidney disease interferes with your everyday activities? 1 not at all 2 a little 3 somewhat 4 a lot 17. How much of a problem would you have with your diabetes if you did not take your medications? 1 not at all 2 a little 3 somewhat 4 a lot

PAGE 295

Appendix 4: (Continued) 282 18. To what extent do you believe that the benefits of taking care of your diabetes is worth the effort? 1 not at all 2 a little 3 somewhat 4 a lot 19. To what extent do you worry about future health problems due to your diabetes? 1 not at all worried 2 somewhat worried 3 usually 4 very much 20. To what extent do you feel your diabetes is a serious disease? 1 not at all 2 a little 3 somewhat 4 very much

PAGE 296

Appendix 4: (Continued) 283 21. Do you think it is likely that poorly managed diabetes will lead to health problems affecting the nerves, kidneys, eyes, or heart? 1 not at all likely 2 somewhat likely 3 likely 4 very likely 22. To what extent do diabetes-relat ed skin problems interfere with everyday living? 1 not at all 2 a little 3 somewhat 4 very much 23. How likely are you to have circ ulation problems due to your diabetes? 1 not at all likely 2 somewhat likely 3 likely 4 very likely

PAGE 297

Appendix 4: (Continued) 284 24. How often are spec ial meals prepared for you? 1 not at all 2 occasionally 3 usually 4 nearly always 25. To what extent do you believe that di abetes treated with a pill or diet is a less serious form of diabetes than diabetes treated with insulin? 1 not at all 2 a little 3 somewhat 4 very much 26. To what extent do you believe po orly managed diabetes will lead to serious health problems in the future? 1 not at all 2 a little 3 somewhat 4 very much

PAGE 298

Appendix 4: (Continued) 285 27. Compared to other priorities in your lif e such as family, work, or religion, how important your health? 1 not at all 2 a little 3 somewhat 4 very important 28. How likely is it that you will take be tter care of your health in the future? 1 not at all likely 2 somewhat likely 3 likely 4 very likely 29. If you were having symptoms of hi gh blood sugars (e.g. blurred vision, increased thirst or urination), would you be likely to seek medical help? 1 not at all likely 2 a little likely 3 somewhat likely 4 very likely

PAGE 299

Appendix 4: (Continued) 286 30. Do you hesitate to tell newly ma de friends that you have diabetes? 1 not at all 2 a little 3 somewhat 4 very much 31. How much do you think medical tr eatment will reduce your chances of developing complications from diabetes? 1 not at all 2 a little 3 somewhat 4 very much 32. How likely are you to have diabetes-related skin problems. 1 not at all likely 2 somewhat likely 3 likely 4 very likely

PAGE 300

Appendix 4: (Continued) 287 33. How would you rate you health? 1 poor 2 fair 3 good 4 excellent 34. To what extent do you feel you have control over your diabetes? 1 not at all 2 a little 3 somewhat 4 very much 35. How likely is that you will have kidney problems due to your diabetes: 1 not at all 2 a little 3 somewhat 4 very

PAGE 301

Appendix 4: (Continued) 288 36. How much would diabetes-related ki dney disease get in the way of your daily living? 1 not at all 2 a little 3 somewhat 4 very much 37. If you were shor t of breath, would you be likely to see a doctor? 1 not at all likely 2 somewhat likely 3 likely 4 very likely 38. How much would diabetes-related eye disease get in the way of your daily living? 1 not at all 2 a little 3 somewhat 4 very much

PAGE 302

Appendix 4: (Continued) 289 39. How likely do you think it is that you will have numbness and tingling in your arms and legs due to your diabetes? 1 not at all likely 2 somewhat likely 3 likely 4 very likely 40. How much do you think your doctor can help if you have diabetes-related tingling and numbness in your arms and legs? 1 not at all helpful 2 a little helpful 3 somewhat helpful 4 very helpful 41. To what extent would numbness and tinglin g in your arms or legs get in the way of your daily living? 1 not at all 2 a little 3 somewhat 4 very much

PAGE 303

Appendix 4: (Continued) 290 42. If you were sick to your st omach would you seek treatment? 1 not at all likely 2 somewhat likely 3 likely 4 very likely 43. How likely do you think it is that you will have sexual problems (impotence, vaginal dryness or yeast infe ctions) due to your diabetes? 1 not at all likely 2 somewhat likely 3 likely 4 very likely 44. If you were unable to think concentrat e clearly, how likely would you be to seek help? 1 not at all likely 2 somewhat likely 3 likely 4 very likely

PAGE 304

Appendix 4: (Continued) 291 45. To what extent do your family and friends remind you to follow your diabetes treatment plan (take medications, follo w diet, exercise, check blood sugars, etc.)? 1 not at all 2 sometimes 3 usually 4 too much 46. How confident are you in your abili ty to eat meals at regularly scheduled times, every 4-5 hours? 1 not at all 2 a little 3 somewhat 4 very 47. How confident are you in your ability to follow a calorie controlled diabetes meal plan? 1 not at all 2 a little 3 somewhat 4 very

PAGE 305

Appendix 4: (Continued) 292 48. How confident are you in y our ability to read food labels? 1 not at all 2 a little 3 somewhat 4 very 49. How confident are you in your ab ility to select healthy foods when eating away from home? 1 not at all 2 a little 3 somewhat 4 very 50. How confident are you in your ability to test your blood sugars daily using a home blood sugar meter? 1 not at all 2 a little 3 somewhat 4 very

PAGE 306

Appendix 4: (Continued) 293 51. How confident are you in your ability to exercise (e.g. walk, bike, swim) 3-5 times a week for at leas t 15-20 minutes at a time? 1 not at all 2 a little 3 somewhat 4 very 52. How confident are you in your abilit y to follow your diabetes treatment plan (medication, diet, exercise, home bl ood sugar monitoring, foot care)? 1 not at all 2 a little 3 somewhat 4 very 53. How confident are you in your ability to cope with stress and your feelings (worry, fear, denial, etc.) about diabetes? 1 not at all 2 a little 3 somewhat 4 very

PAGE 307

Appendix 4: (Continued) 294 54. How confident are you in your ability to test your blood sugars and adjust your eating, exercise, or medications to improve your blood sugar levels? 1 not at all 2 a little 3 somewhat 4 very Environmental/Situational Barriers Answer the questions about your diabetes care YES or NO. If YES, circle the number 1-4 that best describes the extent each barrier affected you (1 not at all to 4 very much) over the past 3 months. 55. Do you take pills to control your diabetes? YES _____ NO _____ 56. Do you take insulin? YES _____ NO _____

PAGE 308

Appendix 4: (Continued) 295 57. To what extent did the situations bel ow keep you from taking your diabetes medication/s (pills/insulin) as you should? Not at all Very Much ____ time or schedule problems 1 2 3 4 ____ inconvenience (carrying supplies with me, finding good place, etc.) 1 2 3 4 ____ side effects of medication 1 2 3 4 ____ forget to take 1 2 3 4 ____ health problems (e.g. trouble seeing, hands shaky) 1 2 3 4 ____ too sick or weak to care for self 1 2 3 4 ____ too painful 1 2 3 4 ____ too costly 1 2 3 4 ____ being away from home 1 2 3 4 ____ special occasions (e.g. birthdays) 1 2 3 4 ____ feeling depressed or other negative emotions (anger, frustration, denial) 1 2 3 4 ____ do not believe the medication ordered is helpful 1 2 3 4 ____ takes too much of my effort 1 2 3 4 ____ interferes with daily activities 1 2 3 4 58. Have you been given a meal plan or diet to follow? YES _____ NO _____

PAGE 309

Appendix 4: (Continued) 296 59. Check any that apply to your meal plan ________ limits added sugar and sweets ________ calorie controlled ________ low fat or cholesterol ________ low salt ________ protein controlled

PAGE 310

297 60. To what extent did the situations below keep you from staying on your meal plan over the past 3 months? Very Little Very Much ____ time or schedule 1 2 3 4 ____ inconvenience (preparing food, etc.) 1 2 3 4 ____ feeling hungry a lot 1 2 3 4 ____ health problems (e.g. vision, shaky hands) 1 2 3 4 ____ too costly 1 2 3 4 ____ eating away from home 1 2 3 4 ____ special occasions 1 2 3 4 ____ negative emotions, feeling stressed, depressed, etc 1 2 3 4 ____ feeling diet is ineffective in controlling my diabetes 1 2 3 4 ____ too difficult 1 2 3 4 ____ takes too much effort 1 2 3 4 ____ interferes with daily activities 1 2 3 4 ____ too few foods I like that are on my diet 1 2 3 4 ____ no one else eats like I have to 1 2 3 4 ____ rely on others to prepare meals 1 2 3 4 61. Has your doctor advised you to exercise? YES _____ NO _____

PAGE 311

298 62. How often do you exercise for 15 minutes or longer at a time? _____ less than once a week _____ once a week _____ 2-3 times a week _____ 4-6 times a week _____ daily

PAGE 312

299 63. To what extent did the situations below keep you from exercising over the past 3 months? Very Little Very Much ____ too busy 1 2 3 4 ____ inconvenience 1 2 3 4 ____ feeling bad during or afterwards 1 2 3 4 ____ health problems 1 2 3 4 ____ too costly 1 2 3 4 ____ away from home 1 2 3 4 ____ special occasions 1 2 3 4 ____ too difficult 1 2 3 4 ____ requires too much effort 1 2 3 4 ____ interferes with other daily activities 1 2 3 4 ____ feeling too tired 1 2 3 4 ____ bad weather 1 2 3 4 ____ fear of low blood sugars 1 2 3 4 ____ too sick 1 2 3 4 ____ hate to exercise alone, no one was available 1 2 3 4 ____ do not like to exercise 1 2 3 4 Do you test your blood sugars with a meter at home YES ____ NO_____

PAGE 313

300 64. Over the past 3 months did you test your blood sugars at home _____ less than once a week _____ 1-2 times a week _____ every other day _____ daily _____ 2 times a day _____ 3 or more times a day

PAGE 314

301 65. To what extent did the situations below keep you from testing your blood sugars over the past 3 months? Very Little Very Much ____ too busy 1 2 3 4 ____ never shown how to use a meter 1 2 3 4 ____ too painful 1 2 3 4 ____ problems with health 1 2 3 4 ____ forgot to test 1 2 3 4 ____ too costly 1 2 3 4 ____ away from home 1 2 3 4 ____ too difficult 1 2 3 4 ____ requires too much effort 1 2 3 4 ____ meter not working 1 2 3 4 ____ rely on others to test, no one available 1 2 3 4 ____ blood sugars have been okay lately, so no need to test 1 2 3 4 ____ inconvenient, interferes with other activities 1 2 3 4 ____ testing makes me too nervous or frustrated by results 1 2 3 4 ____ dislike testing 1 2 3 4 ____ I can tell if my blood sugar is high or low without testing 1 2 3 4 ____ ran out of supplies 1 2 3 4

PAGE 315

302 Appendix 5: Follow-up Evaluation Form a nd Dex Glucometer Mo nitoring System Summary FOLLOW-UP EVALUATION FORM PART I Date: ____/____/____ (mm/dd/yy) Patient ID#_________ __________________ ____________ Attendance Dates of Program: ____/____/____ to ____/____/____ (mm/dd/yy) (mm/dd/yy) Attended more than 5 hours of sessions: (1) ______ Attended less than 4 hours of sessions: (2) ______ Missed sessions (# ) __________________________

PAGE 316

Appendix 5: (Continued) 303 Date: ____/____/____ (mm/dd/yy) Post Education: Expanded HBM Questionnaire Mean Scores Overall Attitudes Scores: _______________ Overall Health Motiva tion Scores: _______________ Overall Severity Scores _______________ Overall Susceptibi lity Scores _______________ Overall Barriers Scores: _______________ Overall Benefits Scores: _______________ Overall Self-efficacy Scores _______________ Overall Cues to Action Scores _______________ Overall Structural El ements Scores _______________

PAGE 317

Appendix 5: (Continued) 304 PART II Final Hgb A1c post 8 weeks intervention: __________ % Date: ____/____/____ 16 week FU ___________ % (mm/dd/yy) __________ Weight (Kg) Date: ____/____/____ __________ Height (cm) (mm/dd/yy) __________ Body Mass Index (Kg/m2) Current Type of Therapy Date Started on Therapy ______ (1) Diet only specif y ____________________ ___/___/___ (mm/dd/yy) ______ (2) Diet and OHA specify ________________ ___/___/___ (mm/dd/yy) ______ (3) OHA specify _______________________ ___/___/___ (mm/dd/yy) ______ (4) Insulin specif y ______________________ ___/___/___ (mm/dd/yy)

PAGE 318

Appendix 5: (Continued) 305 Any change in type therapy over pas t 3 months? No ______ Yes ______ If yes, specify ____________ _____________________ __________ Any hospitalizations or ER visits in last 3 months? No _____ Yes _____ If yes, specify ____________ _____________________ __________ Date of last physician visit (f or diabetes m anagement) ____/____/____ (mm/dd/yy) Patient ID#: ________ __________________ ______________ Part I is completed upon completion of program and placed in patient's file. Part II is a separate follow-up form used fo r all patients for 3 months follow-up by trained investigator performing Hgb A1c determinations. Patients are notified of follow-up by administrative assistant and asked to refrain from discussing whether they attended program or not; and to identify themselves only by ID number and showing appointment card.

PAGE 319

Appendix 5: (Continued) 306 Dex Glucometer Monitoring System Summary Summary View Table 5. Statistical summary of blood glucose results. Summary – mg/dL Average 176 mg/dL Standard Deviation 123 mg/dL Number of Readings 88 Days Covered 26 Average Readings per Day 3.3 Deleted Readings 6 Readings/Range – mg/dL Number Percent HI 0 0% Very High (300-600) 10 11% High (156-299) 30 34% Target (65-155) 30 44% Low (41-64) 3 3% Very Low (10-40) 6 6% LO 0 0% Control Readings 9 Deleted Control Readings 0 Lowest Blood Glucose 15 mg/dL High Blood Glucose 600 mg/dL ( ) – Values in parentheses include at least one HI or LO value. Notes: Indicative of patient comp liance and frequency of testing

PAGE 320

Appendix 5: (Continued) 307 Provides an overview assessment of general level of glycemic control. ( WinGlucofacts Help Manual, Ba yer Corporation, p. 6.)

PAGE 321

308 Appendix 6: Diabetes Self-management Education Schedule DIABETES SELF-MANAGEMENT EDUCATION SCHEDULE SESSION TIME TOPIC DISCIPLINE SPECIALIST Part 1 2 hours Orientation/Case Assessment Hgb A1c (EHBMQ) Case Manager Diabetes Educator Part 2 1 hour Diabetes Overview/Medication Nutrition Guidelines Behavior Goal Setting Diabetes Educator Dietitian Dietitian Part 1 2 hours Exercise/Monitoring Exercise Practicum & Glucose Monitory Practicum Goal Setting Exercise Specialist Exercise Specialist Part 2 2 hours Acute Complications Hypo/hyperglycemia Management Chronic Complications/foot care/foot exam Goal Setting – Glucose Monitoring Log Diabetes Educator

PAGE 322

Appendix 6: (Continued) 309 SESSION TIME TOPIC DISCIPLINE SPECIALIST Part 1 1 hour Exercise Follow-up Diet Follow-up/Special Situations Food Log Follow-up Exercise Specialist Dietitian Part 2 2 hours Diabetes and Stress/Sick day Rules Goal Setting/Taking Charge/Disaster Plan and Travel Evaluation Hgb A1c/Post EHBMQ Assessments Diabetes Educator Dietitian Case Manager/Diabetes Educator Core sessions (part 1 and 2) will be scheduled in week one. Advanced classes in areas identified in behavioral goals and follow-up. Education sessions times/dates may vary but all content areas will be covered.. NOTE: Minimum sessions for maintenance in study are part 1 and part 2 (or 6 hours).

PAGE 323

310 Appendix 7: Patient Satisfaction Questionnaire Name ________________________ Date _____________ ___________________ DIABETE CARE INSTITUTE PATIENT SATISFACTION QUESTIONNAIRE Excellent Very Good Good Fair Poor (1) The overall quality of the program was (2) The materials used were (3) The instructors were (4) The physical facilities were Yes No Don’t Know

PAGE 324

Appendix 7: (Continued) 311 (5) I participated in setting my diabetes management goals (6) My instructors showed interest in meeting my needs (7) I understand the information that was presented (8) I feel better able to manage my diabetes after attending this program (9) I have used services at this hospital before (10) I would recommend th ese classes to someone that I know (11) Class time availability was convenient

PAGE 325

Appendix 7: (Continued) 312 Would you like someone to call you about any concerns that you have about your diabetes? Yes No Phone # Please list any topics that you would like to hear more about: ________________________ _____________________ ___________________ Comments: Date:

PAGE 326

313 Appendix 8: Educational Materials Diabetic Topics : Patient handouts on major diabetes self-care topics designed to supplement the pati ent education manual and are cartoon illustrated. Topics include: Benef its of exercise, fat busters, Hypoglycemia, What is Stress? Ea ting out Guidelines, hyperglycemia exercise guidelines, sick day rules, ma king exercise a prio rity, foot care, stress reducers, assertivene ss, sweeteners, and fiber. Education Posters and Flipcharts: Include graphics highlighted with color and bold print. Main points such as: Function of the pancreas, symptoms of diabetes, conversion of food into energy, six exchange groups, component of an exercise program, s ugar and symptoms of stress, oral agent action, insulin action, hyperglycemia/ hypoglycemia, blood glucose goals, stretchy exercises, nutri tional goals, the food you “chews.”

PAGE 327

Appendix 8: (Continued) 314 Audiovisual Materials: The following are Diabete s Treatment Centers of America approved audiovisual materials. Individual center purchase of materials, once approved, must be cl eared through the appropriate RDO. Balance your Act Slides, Prichert and Hall Diabetes Exercise Slides, Prichert and Hall Diabetes Programs in S panish, Milner-Fennrick Diabetes and Complications: Focus on Living, Oracle Film and Video In Balance, In Control, Boehringer Mannheim, 1987 Taking Charge: Living with Dia betes, Oracle Film and Video Pamphlets Meal Planning with Carbohydrat es, American Healthways, Inc Eating Healthy Foods, American Diabetes Association The Exchange List, American Diabetes Association Month of Meals 5, Amer ican Diabetes Association A Menu Planner, American Diabetes Association Dining Out Made Simple, Becton Dickinson Exercise and Its Benefits, Becton Dickinson Are You Ready for New Food Labels?, Nabisco Biscuit Co.

PAGE 328

Appendix 8: (Continued) 315 *The Sodium Story, Amer ican Diabetes Association Living Your Life with Diabetes, Channing L. Bete Co.

PAGE 329

316 Appendix 9: Division of Nursi ng Standards/Policies/Procedures TITLE: Continuous Glucose Monitoring System (CGMS) INDEX NO. APPROVED: No. Pages Signature: PURPOSE: To provide guidelines on the use of the Continuous Glucose Monitoring System (CGMS), to demonstrate blood gluc ose trends, and determine patient’s responses to exercise, medications, food intake, and stressful situations. STANDARD: Tx 3.9 Medication effects on pati ents are continually monitored. LEVEL OF RESPONSIBILITY: RN, CDE to administer procedure of CGMS.

PAGE 330

Appendix 9: (Continued) 317 STANDARD OF PRACTICE: 1. Preplanned outpatient admission for CGMS will be scheduled through Diabetes Care Institute (DCI) wit h prescription from physician. 2. a. Patient will be able to keep a diary of food, insulin, exercise, and hypoglycemic events b. Patient will wear sens or for three days (72 hrs.) c. Patient will return for sensor download 3. The nurse will integrate data from patient diary sheets on to CGMS data sheets. Physician, Nurse Practitioner, and/or RN CDE to analyze and interpret data for appropriate changes in diabetes management. DESCRIPTION OF CGMS SENSOR: Used by diabetes care professionals to record comprehensive glucose profiles and corresponding events, i. e., stress, low blood sugar, food intake, insulin and/or oral diabetes agents. CGMS measures three days (72 hr s.) and records glycemic control of patient within ranges of 40 mg/dl to 400 mg/dl.

PAGE 331

Appendix 9: (Continued) 318 CGMS records sensor signals every five minutes providing 288 glucose readings per day. PROCEDURE: Equipment : Continuous Glucose Monitor Cable Glucose Sensor Com-station for Downloading Sen-serter Belt Clip Transparent Dressing Tape Patient Blood Glucose Meter IV Prep Pads “Shower-pak” Three (3) Patient Diary Sheets MiniMed CGMS Pocket Guidebook

PAGE 332

Appendix 9: (Continued) 319 CONTINUOUS GLUCOSE MONI TORING: INITIATION I. PRE-INITIATION SET-UP STEPS KEY POINTS 1. Gather equipment and supplies. 2. Explain procedure to patient 2.1 CGMS education must be thorough or the data will not be entered correctly as outlined in the instruction manual.3. Evaluate all operational functions of the sensor prior to initiating sensor therapy. 4. Describe and demonstrate the basic operation techniques to the patient. a) Entering meter blood glucose values into the sensor b) Selecting “Cal Yes” and confirming the selection by pressing the “ACT” button c) Entering events and clearing alarms 5. Review with patient the MiniMed CGMS Pocket Guidebook that covers the following topics. 4.1 These are the basic functions of the sensor and must be demonstrated by the patient to assure correct technique for data to be accurate and measurable.

PAGE 333

Appendix 9: (Continued) 320 6. Treat for hypoglycemia if a glucose reading is < 60 mg/dl or if hypoglycemic symptoms occur when < 100 mg/dl. Check blood glucose again in 15 minutes and enter event into the sensor. 6.1 The CGMS monitor will only accept values between 40 mg/dl 400 mg/dl. If the SMBG value is outside of this range, test again after treatment and enter the value once it is within the specified range. II. SITE SELECTION AND PREPARATION STEPS 1. Abdominal ar ea preferred. Keep at least two inches away from pump infusion site or three inches away from insulin injection site. III. GLUCOSE SENSOR INSERTION STEPS 1. Remove sensor from sterile package. KEY POINTS 1.1 Avoid areas around scars, waistline, underwear lines, and within a two inch circle around the navel. KEY POINTS 1.1 Sensor must be stored between + 36 degrees to + 75 degrees Fahrenheit. 1.2 Warm to room temperature for 10 minutes before inserting. 2. Load sensor into Sen-serter by sliding plug until snug. 2.1 Lock Sen-serter to prevent premature firing. 3. Remove clear release paper from adhesive pad (around the needle) Remove needle guard from sensor Unlock Sen-serter

PAGE 334

Appendix 9: (Continued) 321 4. Hold Sen-serter on insertion site. Press top button on Senserter to insert into sensor needle and catheter. 5. To remove sensor from Senserter, slide Sen-serter backwards. 5.1 Grasp handle of sensor. Keep Sen-serter “legs” on skin while sliding backwards. IV. POST SENSOR INSERTION STEPS 1. Remove white backing from sensor KEY POINTS 1.1 Hold sensor against skin to prevent dislodging. 2. Remove introducer needle by gently pulling handle away from needle site. 2.1 Dispose of needle into appropriate sharps receptacle. 3. Check site to verify sensor is inserted flush with skin site. 4 Verify that the sensor is working after insertion by checking the impact signal (ISIG) and the voltage counter (VCTR). Press SEL button to go to Set Up screen, press A CT button. Press SEL button to go to SIGNALS Screen. Press ACT button. The sensor will show ISIG value. Press SEL button to view the VCTR value. Press ACT button to exit. 5. Secure sensor connection with sterile transparent dressing. 4.1a The sensor is working well if ISIG is between 10-100 NA and VCTR is between 0.4 v and 1.6 v. 4.1b If the values are not within the acceptable range, 1) check all connections, 2) connect the cable connector end of the test plug to the cable, find the “Signals menu and the ISIG screen. The value is between 24-29 NA, 3) call MiniMed if ISIG is not within the acceptable range when test plug procedure is done.

PAGE 335

Appendix 9: (Continued) 322 6. Begin initialization. Press SEL button screen to go to Set Up screen, press ACT button. Press SEL button to go to Initialization screen. Press ACT button. The screen will show 60 and it will count down to zero. 6.1a The initialization process takes 60 minutes to complete. 6.1b Do not initialize if the patient’s BG is rapidly fluctuating. 6.1c Do not initialize if the patient’s BG is greater than 400 mg/dl or if it is less than 40 mg/dl. 6.1d Do not Press any button on the monitor during initialization period. It will take you out of initialization.6.1e Do not initialize the same sensor more than once. 6.1f If initialization is interrupted, wait 60 minutes and perform calibration. V. DOWNLOADING CGMS DATA STEPS 1. Download CGMS data per MiniMed manual three days post insertion. 2. The nurse will integrate data from patient diary sheets onto CGMS data sheets. Physician and/or Nurse Practitioner, and/or RN CDE to analyze and interpret data for appropriate changes in diabetes management. KEY POINTS

PAGE 336

Appendix 9: (Continued) 323 VI. DOCUMENTATION STEPS 1. Document per established Hospital guidelines. 2. Complete the CGMS Checklist. 2.1 See attached checklist. VII. DISPOSITION OF DATA 1. One copy of all downloaded CGMS data is sent to: Medical Records Physician Patient.

PAGE 337

324 Appendix 10: Box-plot and Stem-leaf Diagrams of Continuous Variables at Baseline Self-efficacy Score

PAGE 338

Appendix 10: (Continued) 325 Susceptibility Score Benefits Score

PAGE 339

Appendix 10: (Continued) 326 Severity Score

PAGE 340

Appendix 10: (Continued) 327 Cue-to-Action Score Structural Score

PAGE 341

Appendix 10: (Continued) 328 Motivation Score Barriers Score

PAGE 342

Appendix 10: (Continued) 329 Age Hemoglobin

PAGE 343

330 Appendix 11: Chronbach Alpha Resu lts for Individual Questions Time 1 Results Raw Variables (Time 1) Standardized Variables (Time 1) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlati on with Total Alpha 1 Q4 0.502874 0.381125 0.233522 0.514927 0.3811250.233522 0.514927 2 Q14 0.190814 0.449102 0.168631 0.544623 0.4491020.168631 0.544623 3 Q57 Total 0.229640 0.399295 0.265343 0.499928 0.3992950.265343 0.499928 4 Q60 Total 0.363096 0.327118 0.477222 0.392403 0.3271180.477222 0.392403 5 Q63 Total 0.104887 0.485328 0.219408 0.521487 0.4853280.219408 0.521487 Perceived Barriers 6 Q66 Total 0.392258 0.336723 0.350426 0.458374 0.3367230.350426 0.458374 1 Q15 0.0948270.6580940.092487 0.644720 2 Q29 -0.0861460.679209-0.091699 0.708560 3 Q37 0.5527710.4730900.553173 0.451311 4 Q42 0.4651370.4813360.518547 0.467663 5 Q44 0.3877830.5223210.344425 0.545279 Cues-toaction 6 Q45 0.6956080.3881050.710026 0.373263

PAGE 344

Appendix 11 (Continued) 331 Raw Variables (Time 1) Standardized Variables (Time 1) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlati on with Total Alpha 1 Q60.1 0.3727740.873468* 2 Q60.2 0.8063690.847778* 3 Q60.3 0.9714350.834650* 4 Q60.4 -0.0433510.879708* 5 Q60.5 -0.0433510.879708* 6 Q60.6 0.7329050.856292* 7 Q60.7 0.7822550.849788* 8 Q60.8 0.5697090.862583* 9 Q60.9 0.4477760.871113* 10 Q60.10 0.6856100.856023* 11 Q60.11 0.8370610.846007* 12 Q60.12 0.5215160.865344* 13 Q60.13 0.877309* 14 Q60.14 0.3478510.872892* Diet (Perceived Barriers Subscale) 15 Q60.15 0.877309* 1 Q63.1 0.5983780.519273* 2 Q63.2 0.7974930.492750* 3 Q63.3 0.0317220.625799* 4 Q63.4 -0.1685650.675336* 5 Q63.5 0.623738* 6 Q63.6 0.4000780.572950* 7 Q63.7 0.3993810.579241* 8 Q63.8 0.0321410.632143* 9 Q63.9 -0.0800500.661246* 10 Q63.10 0.3215990.590578* 11 Q63.11 0.6722140.497040* 12 Q63.12 -0.1598260.635128* 13 Q63.13 0.623738* 14 Q63.14 0.623738* 15 Q63.15 0.1202410.619412* Exercise (Perceived Barriers Subscale) 16 Q63.16 0.2289360.607263*

PAGE 345

Appendix 11 (Continued) 332 Raw Variables (Time 1) Standardized Variables (Time 1) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlati on with Total Alpha 1 Q57.1 0.7844650.715648* 2 Q57.2 0.7844650.715648* 3 Q57.3 0.0094320.782903* 4 Q57.4 0.7405670.720844* 5 Q57.5 0.777778* 6 Q57.6 0.777778* 7 Q57.7 0.777778* 8 Q57.8 0.777778* 9 Q57.9 0.7637630.711538* 10 Q57.10 0.7844650.715648* 11 Q57.11 0.2175770.779252* 12 Q57.12 0.2996740.768066* 13 Q57.13 0.777778* 14 Q57.14 0.777778* Medication (Perceived Barriers Subscale) 15 Q57.15 0.777778* 1 Q66.1 0.8718510.618211* 2 Q66.2 0.732686* 3 Q66.3 0.732686* 4 Q66.4 0.2027690.726590* 5 Q66.5 0.8760380.624289* 6 Q66.6 0.732686* 7 Q66.7 0.8333330.650667* 8 Q66.8 0.732686* 9 Q66.9 0.732686* 10 Q66.10 0.732686* 11 Q66.11 0.732686* 12 Q66.12 0.732686* 13 Q66.13 0.1435920.734021* 14 Q66.14 0.732686* 15 Q66.15 0.732686* 16 Q66.16 0.7122120.663188* Monitoring (Perceived Barriers Subscale) 17 Q66.17 0.732686*

PAGE 346

Appendix 11 (Continued) 333 Raw Variables (Time 1) Standardized Variables (Time 1) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlati on with Total Alpha 1 Q1 0.212733-0.2992960.116706 -0.055354 2 Q2 -0.0736820.042980-0.152801 0.217014 3 Q27 -0.0995040.0000000.018127 0.051745 4 Q28 0.091502-0.0961540.254751 -0.221097 5 Q30 -0.0972430.116667-0.113094 0.180817 Motivation 6 Q33 -0.0548990.0316460.056152 0.011501 1 Q46 0.6859020.8487330.675021 0.862971 2 Q47 0.8198450.8334120.814822 0.850469 3 Q48 0.6909570.8543030.696928 0.861046 4 Q49 0.7265800.8442260.732856 0.857861 5 Q50 0.7007650.8564010.710679 0.859831 6 Q51 -0.0883080.913772-0.075861 0.921792 7 Q52 0.7938050.8386070.803484 0.851502 8 Q53 0.7974450.8406280.795383 0.852238 SelfEfficacy 9 Q54 0.5957200.8573410.574163 0.871674 1 Q16 0.0653980.1944440.007707 0.369557 2 Q20 0.0963250.1803540.268712 0.228909 3 Q22 0.1304350.141147-0.119968 0.430523 4 Q25 -0.2416620.419932-0.257456 0.490879 5 Q26 0.0963250.1803540.257841 0.235221 6 Q36 0.0297270.2210080.341125 0.185810 7 Q38 0.1869270.1475320.345325 0.183254 Perceived Severity 8 Q41 0.500618-0.1790430.358417 0.175244 1 Q3 0.6864060.1514420.673670 0.095441 2 Q12 0.670635-0.3080360.667063 0.102956 3 Q24 0.2824660.2906250.355782 0.419195 Structural Elements 4 Q34 -0.2479920.763727-0.199238 0.822543 1 Q9 0.3409300.8242670.319445 0.810831 2 Q11 0.8065940.7663740.819926 0.745706 3 Q19 0.5324310.7996130.479457 0.791205 4 Q21 0.0367360.8351650.058666 0.840553 5 Q23 0.7011550.7822350.730432 0.758174 6 Q32 0.6412790.7849500.617046 0.773444 7 Q35 0.4986570.8035640.557348 0.781252 8 Q39 0.5377570.7989090.492314 0.789580 Perceived Susceptibility 9 Q43 0.5184760.8022120.507269 0.787681

PAGE 347

Appendix 11 (Continued) 334 Raw Variables (Time 1) Standardized Variables (Time 1) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlati on with Total Alpha 1 Q5 0.2209180.7238550.176288 0.785261 2 Q6 0.6967600.6477710.748573 0.709030 3 Q7 -0.1707790.752044-0.140336 0.821383 4 Q8 0.7048650.6658960.728440 0.711969 5 Q10 0.5766970.6795810.598296 0.730495 6 Q13 0.6845800.6886290.685597 0.718156 7 Q17 0.7032550.6318990.717798 0.713514 8 Q18 0.6845800.6886290.685597 0.718156 9 Q31 0.1202990.7701830.108974 0.793283 Perceived Treatment Benefits 10 Q40 0.2132780.7393560.229529 0.778782 Not calculable due to lack of variance on several questions. Time 2 Results Raw Variables (Time 2) Standardized Variables (Time 2) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlation with Total Alpha 1 Q4 0.2342780.5601640.221685 0.695958 2 Q14 0.3046120.5607960.322089 0.664029 3 Q57 Total 0.6216160.4074670.516691 0.597097 4 Q60 Total 0.4057670.4748120.497854 0.603876 5 Q63 Total 0.1457450.6485750.280122 0.677585 Perceived Barriers 6 Q66 Total 0.6786510.3745310.634702 0.553106 1 Q15 0.6095770.6341910.572354 0.656041 2 Q29 0.2511240.7378380.217016 0.757476 3 Q37 0.4463040.6935980.437072 0.696790 4 Q42 0.6533250.6016440.616524 0.642144 Cues-toaction 5 Q44 0.3710020.7009080.419636 0.701847

PAGE 348

Appendix 11 (Continued) 335 Raw Variables (Time 2) Standardized Variables (Time 2) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlation with Total Alpha 6 Q45 0.4690890.6728320.531680 0.668578 1 Q60.1 0.6251390.745955* 2 Q60.2 0.1991220.785435* 3 Q60.3 0.5180340.757560* 4 Q60.4 0.785577* 5 Q60.5 0.785577* 6 Q60.6 0.8673250.709435* 7 Q60.7 0.8618930.711929* 8 Q60.8 0.6248380.745118* 9 Q60.9 -0.0406720.803876* 10 Q60.10 0.6202990.748962* 11 Q60.11 0.6737100.743949* 12 Q60.12 0.785577* 13 Q60.13 0.785577* 14 Q60.14 0.3365300.773852* Diet (Perceived Barriers Subscale) 15 Q60.15 -0.2289850.816722* 1 Q63.1 -0.1427120.832129* 2 Q63.2 0.1256060.812165* 3 Q63.3 0.7572400.779058* 4 Q63.4 0.3702080.803876* 5 Q63.5 -0.2220540.820658* 6 Q63.6 0.0234950.821709* 7 Q63.7 0.5844720.782042* 8 Q63.8 0.5688760.788217* 9 Q63.9 0.2352670.806537* 10 Q63.10 0.810519* 0.788295 11 Q63.11 0.7550380.771706* 12 Q63.12 0.3408110.802124* 13 Q63.13 0.7345950.770422* 14 Q63.14 0.4881410.797534* Exercise (Perceived Barriers Subscale) 15 Q63.15 0.8112520.760170*

PAGE 349

Appendix 11 (Continued) 336 Raw Variables (Time 2) Standardized Variables (Time 2) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlation with Total Alpha 16 Q63.16 0.8629410.749550* 1 Q57.1 0.7352150.608316* 2 Q57.2 0.7001520.627864* 3 Q57.3 0.2148440.669204* 4 Q57.4 0.6356900.594278* 5 Q57.5 0.679543* 6 Q57.6 0.679543* 7 Q57.7 0.679543* 8 Q57.8 0.679543* 9 Q57.9 0.5631250.607026* 10 Q57.10 0.8562840.581445* 11 Q57.11 0.2390770.671541* 12 Q57.12 0.6391370.598643* 13 Q57.13 -0.2055890.785280* 14 Q57.14 0.679543* Medication (Perceived Barriers Subscale) 15 Q57.15 0.679543* 1 Q66.1 0.0456590.587611* 2 Q66.2 0.541570* 3 Q66.3 -0.1132100.557229* 4 Q66.4 0.5358790.491424* 5 Q66.5 0.2954270.499239* 6 Q66.6 0.6304780.373984* 7 Q66.7 0.5224300.427240* 8 Q66.8 0.8730380.424065* 9 Q66.9 0.5358790.491424* 10 Q66.10 0.541570* 11 Q66.11 0.541570* 12 Q66.12 -0.2876930.661352* 13 Q66.13 -0.0103910.547328* 14 Q66.14 0.541570* Monitoring (Perceived Barriers Subscale) 15 Q66.15 0.541570*

PAGE 350

Appendix 11 (Continued) 337 Raw Variables (Time 2) Standardized Variables (Time 2) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlation with Total Alpha 16 Q66.16 0.3726300.471608* 17 Q66.17 0.541570* 1 Q1 0.5949800.2987010.571275 0.488387 2 Q2 0.2815790.5090310.367003 0.574122 3 Q27 0.3448810.5139500.423287 0.551445 4 Q28 0.5811900.3944950.584742 0.482396 5 Q30 -0.0636700.703919-0.137447 0.747960 Motivation 6 Q33 0.3284890.4880140.427466 0.549733 1 Q46 0.6434510.8129010.645958 0.842984 2 Q47 0.6613170.8114400.656759 0.841931 3 Q48 0.6072310.8236980.638925 0.843667 4 Q49 0.8925380.7828750.863070 0.821157 5 Q50 0.2856080.8451380.328559 0.872375 6 Q51 0.0207690.8922000.050979 0.895769 7 Q52 0.7349340.8013390.720377 0.835664 8 Q53 0.6340290.8189980.629826 0.844548 SelfEfficacy 9 Q54 0.8886350.7940980.877142 0.819692 1 Q16 0.4726420.6178690.448771 0.606487 2 Q20 0.1071430.6875000.133929 0.683556 3 Q22 0.6861060.5342110.657716 0.549027 4 Q25 0.2265180.6943190.231649 0.660807 5 Q26 0.2182180.6770830.147106 0.680548 6 Q36 0.6506260.5877530.683508 0.541566 7 Q38 0.4820080.6257670.492177 0.594980 Perceived Severity 8 Q41 0.1960120.6769550.117607 0.687257 1 Q3 0.6875000.3828130.716078 0.328611 2 Q12 0.6369950.1683670.642607 0.391415 3 Q24 0.1478340.6771650.144869 0.742721 Structural Elements 4 Q34 0.2138130.5874130.255744 0.674931 1 Q9 0.7252430.8489240.738991 0.822461 2 Q11 0.8083410.8310670.757140 0.820543 Perceived Susceptibility 3 Q19 0.7784960.8313920.765925 0.819611

PAGE 351

Appendix 11 (Continued) 338 Raw Variables (Time 2) Standardized Variables (Time 2) Scale or Subscale Number of Question on Scale Question from Questionnaire Correlation with Total Alpha Correlation with Total Alpha 4 Q21 0.2277100.8751020.212930 0.873553 5 Q23 0.5638600.8565920.538280 0.842964 6 Q32 0.7619640.8336670.752644 0.821019 7 Q35 -0.0580260.897354-0.069887 0.897641 8 Q39 0.9382260.8143390.972999 0.796896 9 Q43 0.6624330.8447200.661327 0.830547 1 Q5 0.7284780.8546640.789364 0.860131 2 Q6 0.6863510.8458790.789364 0.860131 3 Q7 -0.0854980.886678-0.094467 0.919683 4 Q8 0.6475460.8516850.630623 0.871935 5 Q10 0.7051810.8437500.630663 0.871932 6 Q13 0.7284780.8546640.789364 0.860131 7 Q17 0.8148900.8428180.826798 0.857272 8 Q18 0.7284780.8546640.789364 0.860131 9 Q31 0.5645770.8570540.475783 0.882965 Perceived Treatment Benefits 10 Q40 0.7051810.8437500.630663 0.871932

PAGE 352

339 Appendix 12: SAS Output for Research Questions Rsch Question 1 Anova The GLM Procedure Class Level Information Class Levels Values group2 2 1_Intensive 2_Control Number of Observations Read 159 Number of Observations Used 101 Rsch Question 1 Anova The GLM Procedure Dependent Variable: hgb1 Hemoglobin Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.2807848 0.2807848 0.16 0.6880 Error 99 171.3821855 1.7311332 Corrected Total 100 171.6629703 R-Square Coeff Var Root MSE hgb1 Mean 0.001636 15.43597 1.315725 8.523762 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.28078476 0.28078476 0.16 0.6880 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.28078476 0.28078476 0.16 0.6880 Rsch Question 1 Anova The GLM Procedure

PAGE 353

Appendix 12 (continued) 340 Dependent Variable: hgb2 Hemoglobin Wk 12 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.0014198 0.0014198 0.00 0.9776 Error 99 177.8302634 1.7962653 Corrected Total 100 177.8316832 R-Square Coeff Var Root MSE hgb2 Mean 0.000008 18.78244 1.340248 7.135644 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.00141980 0.00141980 0.00 0.9776 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.00141980 0.00141980 0.00 0.9776 Rsch Question 1 Anova The GLM Procedure Dependent Variable: hgb3 Hemoglobin Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.0437294 0.0437294 0.02 0.8798 Error 99 188.1966667 1.9009764 Corrected Total 100 188.2403960 R-Square Coeff Var Root MSE hgb3 Mean 0.000232 19.75243 1.378759 6.980198 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.04372937 0.04372937 0.02 0.8798 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.04372937 0.04372937 0.02 0.8798 Rsch Question 1 Anova

PAGE 354

Appendix 12 (continued) 341 The GLM Procedure Repeated Measures Analysis of Variance Repeated Measures Level Information Dependent Variable hgb1 hgb2 hgb3 Level of time 1 2 3 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no time Effect H = Type III SSCP Matrix for time E = Error SSCP Matrix S=1 M=0 N=48 Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.47527443 54.10 2 98 <.0001 Pillai's Trace 0.52472557 54.10 2 98 <.0001 Hotelling-Lawley Trace 1.10404757 54.10 2 98 <.0001 Roy's Greatest Root 1.10404757 54.10 2 98 <.0001 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no time*group2 Effect H = Type III SSCP Matrix for time*group2 E = Error SSCP Matrix S=1 M=0 N=48 Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.99713008 0.14 2 98 0.8686 Pillai's Trace 0.00286992 0.14 2 98 0.8686 Hotelling-Lawley Trace 0.00287818 0.14 2 98 0.8686 Roy's Greatest Root 0.00287818 0.14 2 98 0.8686 Rsch Question 1 Anova The GLM Procedure Repeated Measures Analysis of Variance Tests of Hypotheses for Between Subjects Effects Source DF Type III SS Mean Square F Value Pr > F group2 1 0.0428302 0.0428302 0.01 0.9150 Error 99 370.6043645 3.7434784 Rsch Question 1 Anova The GLM Procedure Repeated Measures Analysis of Variance Univariate Tests of Hypotheses for Within Subject Effects Adj Pr > F Source DF Type III SS Mean Square F Value Pr > F G G H F time 2 146.1588793 73.0794397 86.75 <.0001 <.0001 <.0001 time*group2 2 0.2831037 0.1415519 0.17 0.8455 0.7693 0.7741

PAGE 355

Appendix 12 (continued) 342 Error(time) 198 166.8047510 0.8424482 Greenhouse-Geisser Epsilon 0.7100 Huynh-Feldt Epsilon 0.7247 Rsch Question 1 Anova The GLM Procedure Tukey's Studentized Range (HSD) Test for hgb1 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 99 Error Mean Square 1.731133 Critical Value of Studentized Range 2.80611 Minimum Significant Difference 0.5202 Harmonic Mean of Cell Sizes 50.37624 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 8.5792 48 2_Control A A 8.4736 53 1_Intensive Rsch Question 1 Anova The GLM Procedure Bonferroni (Dunn) t Tests for hgb1 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 99 Error Mean Square 1.731133 Critical Value of t 1.98422 Minimum Significant Difference 0.5202 Harmonic Mean of Cell Sizes 50.37624 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2

PAGE 356

Appendix 12 (continued) 343 A 8.5792 48 2_Control A A 8.4736 53 1_Intensive Rsch Question 1 Anova The GLM Procedure Tukey's Studentized Range (HSD) Test for hgb2 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 99 Error Mean Square 1.796265 Critical Value of Studentized Range 2.80611 Minimum Significant Difference 0.5299 Harmonic Mean of Cell Sizes 50.37624 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 7.1396 48 2_Control A A 7.1321 53 1_Intensive Rsch Question 1 Anova The GLM Procedure Bonferroni (Dunn) t Tests for hgb2 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 99 Error Mean Square 1.796265 Critical Value of t 1.98422 Minimum Significant Difference 0.5299 Harmonic Mean of Cell Sizes 50.37624 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2

PAGE 357

Appendix 12 (continued) 344 A 7.1396 48 2_Control A A 7.1321 53 1_Intensive Rsch Question 1 Anova The GLM Procedure Tukey's Studentized Range (HSD) Test for hgb3 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 99 Error Mean Square 1.900976 Critical Value of Studentized Range 2.80611 Minimum Significant Difference 0.5451 Harmonic Mean of Cell Sizes 50.37624 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 7.0000 53 1_Intensive A A 6.9583 48 2_Control Rsch Question 1 Anova The GLM Procedure Bonferroni (Dunn) t Tests for hgb3 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 99 Error Mean Square 1.900976 Critical Value of t 1.98422 Minimum Significant Difference 0.5451 Harmonic Mean of Cell Sizes 50.37624 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 7.0000 53 1_Intensive

PAGE 358

Appendix 12 (continued) 345 A A 6.9583 48 2_Control Rsch Question 2 Pretests The GLM Procedure Class Level Information Class Levels Values group2 2 1_Intensive 2_Control Number of Observations Read 159 Number of Observations Used 153 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: preself Self Efficacy Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.459638 1.459638 0.06 0.8037 Error 151 3553.481538 23.532990 Corrected Total 152 3554.941176 R-Square Coeff Var Root MSE preself Mean 0.000411 16.00292 4.851081 30.31373 Source DF Type I SS Mean Square F Value Pr > F group2 1 1.45963801 1.45963801 0.06 0.8037 Source DF Type III SS Mean Square F Value Pr > F group2 1 1.45963801 1.45963801 0.06 0.8037 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: presus Susceptibility Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3.488396 3.488396 0.11 0.7407 Error 151 4793.348205 31.744028

PAGE 359

Appendix 12 (continued) 346 Corrected Total 152 4796.836601 R-Square Coeff Var Root MSE presus Mean 0.000727 20.14560 5.634184 27.96732 Source DF Type I SS Mean Square F Value Pr > F group2 1 3.48839618 3.48839618 0.11 0.7407 Source DF Type III SS Mean Square F Value Pr > F group2 1 3.48839618 3.48839618 0.11 0.7407 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: pretb Benefits Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 9.422061 9.422061 0.58 0.4480 Error 151 2457.885128 16.277385 Corrected Total 152 2467.307190 R-Square Coeff Var Root MSE pretb Mean 0.003819 11.10819 4.034524 36.32026 Source DF Type I SS Mean Square F Value Pr > F group2 1 9.42206134 9.42206134 0.58 0.4480 Source DF Type III SS Mean Square F Value Pr > F group2 1 9.42206134 9.42206134 0.58 0.4480 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: presev Severity Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.875063 0.875063 0.05 0.8177

PAGE 360

Appendix 12 (continued) 347 Error 151 2479.294872 16.419171 Corrected Total 152 2480.169935 R-Square Coeff Var Root MSE presev Mean 0.000353 15.90469 4.052058 25.47712 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.87506285 0.87506285 0.05 0.8177 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.87506285 0.87506285 0.05 0.8177 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: precue Cue-to-Action Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 20.047702 20.047702 1.27 0.2620 Error 151 2388.592821 15.818496 Corrected Total 152 2408.640523 R-Square Coeff Var Root MSE precue Mean 0.008323 22.24932 3.977247 17.87582 Source DF Type I SS Mean Square F Value Pr > F group2 1 20.04770236 20.04770236 1.27 0.2620 Source DF Type III SS Mean Square F Value Pr > F group2 1 20.04770236 20.04770236 1.27 0.2620 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: prestr Structural Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.738472 0.738472 0.06 0.8102

PAGE 361

Appendix 12 (continued) 348 Error 151 1925.902051 12.754318 Corrected Total 152 1926.640523 R-Square Coeff Var Root MSE prestr Mean 0.000383 32.83725 3.571319 10.87582 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.73847159 0.73847159 0.06 0.8102 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.73847159 0.73847159 0.06 0.8102 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: premot Motivation Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3.500251 3.500251 0.36 0.5509 Error 151 1479.179487 9.795891 Corrected Total 152 1482.679739 R-Square Coeff Var Root MSE premot Mean 0.002361 17.34391 3.129839 18.04575 Source DF Type I SS Mean Square F Value Pr > F group2 1 3.50025138 3.50025138 0.36 0.5509 Source DF Type III SS Mean Square F Value Pr > F group2 1 3.50025138 3.50025138 0.36 0.5509 Rsch Question 2 Pretests The GLM Procedure Dependent Variable: prebar Barriers Wk 1 Sum of Source DF Squares Mean Square F Value Pr > F

PAGE 362

Appendix 12 (continued) 349 Model 1 330.83656 330.83656 0.57 0.4511 Error 151 87524.10462 579.62983 Corrected Total 152 87854.94118 R-Square Coeff Var Root MSE prebar Mean 0.003766 25.43714 24.07550 94.64706 Source DF Type I SS Mean Square F Value Pr > F group2 1 330.8365611 330.8365611 0.57 0.4511 Source DF Type III SS Mean Square F Value Pr > F group2 1 330.8365611 330.8365611 0.57 0.4511 Rsch Question 2 Pretests The GLM Procedure Repeated Measures Analysis of Variance Repeated Measures Level Information Dependent Variable preself presus pretb presev precue prestr premot prebar Level of subscale 1 2 3 4 5 6 7 8 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no subscale Effect H = Type III SSCP Matrix for subscale E = Error SSCP Matrix S=1 M=2.5 N=71.5 Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.01650840 1234.06 7 145 <.0001 Pillai's Trace 0.98349160 1234.06 7 145 <.0001 Hotelling-Lawley Trace 59.57523541 1234.06 7 145 <.0001 Roy's Greatest Root 59.57523541 1234.06 7 145 <.0001 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no subscale*group2 Effect H = Type III SSCP Matrix for subscale*group2 E = Error SSCP Matrix S=1 M=2.5 N=71.5 Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.96608372 0.73 7 145 0.6491 Pillai's Trace 0.03391628 0.73 7 145 0.6491 Hotelling-Lawley Trace 0.03510698 0.73 7 145 0.6491 Roy's Greatest Root 0.03510698 0.73 7 145 0.6491

PAGE 363

Appendix 12 (continued) 350 Rsch Question 2 Pretests The GLM Procedure Repeated Measures Analysis of Variance Tests of Hypotheses for Between Subjects Effects Source DF Type III SS Mean Square F Value Pr > F group2 1 40.12381 40.12381 0.44 0.5094 Error 151 13852.64744 91.73939 Rsch Question 2 Pretests The GLM Procedure Repeated Measures Analysis of Variance Univariate Tests of Hypotheses for Within Subject Effects Adj Pr > F Source DF Type III SS Mean Square F Value Pr > F G G H F subscale 7 741048.7019 105864.1003 1206.46 <.0001 <.0001 <.0001 subscale*group2 7 330.2443 47.1778 0.54 0.8063 0.5081 0.5101 Error(subscale) 1057 92749.1413 87.7475 Greenhouse-Geisser Epsilon 0.1843 Huynh-Feldt Epsilon 0.1864 Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for preself NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 23.53299 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 1.5501 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 30.4133 75 1_Intensive A A 30.2179 78 2_Control Rsch Question 2 Pretests

PAGE 364

Appendix 12 (continued) 351 The GLM Procedure Bonferroni (Dunn) t Tests for preself NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 23.53299 Critical Value of t 1.97580 Minimum Significant Difference 1.5501 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 30.4133 75 1_Intensive A A 30.2179 78 2_Control Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for presus NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 31.74403 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 1.8003 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 28.1154 78 2_Control A A 27.8133 75 1_Intensive Rsch Question 2 Pretests

PAGE 365

Appendix 12 (continued) 352 The GLM Procedure Bonferroni (Dunn) t Tests for presus NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 31.74403 Critical Value of t 1.97580 Minimum Significant Difference 1.8003 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 28.1154 78 2_Control A A 27.8133 75 1_Intensive Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for pretb NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 16.27738 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 1.2891 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 36.5733 75 1_Intensive A A 36.0769 78 2_Control Rsch Question 2 Pretests The GLM Procedure

PAGE 366

Appendix 12 (continued) 353 Bonferroni (Dunn) t Tests for pretb NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 16.27738 Critical Value of t 1.97580 Minimum Significant Difference 1.2891 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 36.5733 75 1_Intensive A A 36.0769 78 2_Control Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for presev NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 16.41917 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 1.2948 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 25.5513 78 2_Control A A 25.4000 75 1_Intensive Rsch Question 2 Pretests The GLM Procedure

PAGE 367

Appendix 12 (continued) 354 Bonferroni (Dunn) t Tests for presev NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 16.41917 Critical Value of t 1.97580 Minimum Significant Difference 1.2948 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 25.5513 78 2_Control A A 25.4000 75 1_Intensive Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for precue NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 15.8185 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 1.2708 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 18.2308 78 2_Control A A 17.5067 75 1_Intensive Rsch Question 2 Pretests The GLM Procedure Bonferroni (Dunn) t Tests for precue

PAGE 368

Appendix 12 (continued) 355 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 15.8185 Critical Value of t 1.97580 Minimum Significant Difference 1.2708 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 18.2308 78 2_Control A A 17.5067 75 1_Intensive Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for prestr NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 12.75432 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 1.1411 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 10.9467 75 1_Intensive A A 10.8077 78 2_Control Rsch Question 2 Pretests The GLM Procedure Bonferroni (Dunn) t Tests for prestr

PAGE 369

Appendix 12 (continued) 356 NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 12.75432 Critical Value of t 1.97580 Minimum Significant Difference 1.1411 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 10.9467 75 1_Intensive A A 10.8077 78 2_Control Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for premot NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 9.795891 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 1.0001 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 18.2000 75 1_Intensive A A 17.8974 78 2_Control Rsch Question 2 Pretests The GLM Procedure Bonferroni (Dunn) t Tests for premot NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher

PAGE 370

Appendix 12 (continued) 357 Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 9.795891 Critical Value of t 1.97580 Minimum Significant Difference 1.0001 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 18.2000 75 1_Intensive A A 17.8974 78 2_Control Rsch Question 2 Pretests The GLM Procedure Tukey's Studentized Range (HSD) Test for prebar NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 579.6298 Critical Value of Studentized Range 2.79420 Minimum Significant Difference 7.6928 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 96.147 75 1_Intensive A A 93.205 78 2_Control Rsch Question 2 Pretests The GLM Procedure Bonferroni (Dunn) t Tests for prebar NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ.

PAGE 371

Appendix 12 (continued) 358 Alpha 0.05 Error Degrees of Freedom 151 Error Mean Square 579.6298 Critical Value of t 1.97580 Minimum Significant Difference 7.6928 Harmonic Mean of Cell Sizes 76.47059 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 96.147 75 1_Intensive A A 93.205 78 2_Control Rsch Question 2 Posttests The GLM Procedure Class Level Information Class Levels Values group2 2 1_Intensive 2_Control Number of Observations Read 159 Number of Observations Used 87 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: postself Self Efficacy Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3.594690 3.594690 0.17 0.6769 Error 85 1747.991516 20.564606 Corrected Total 86 1751.586207 R-Square Coeff Var Root MSE postself Mean 0.002052 14.66107 4.534822 30.93103 Source DF Type I SS Mean Square F Value Pr > F group2 1 3.59469046 3.59469046 0.17 0.6769

PAGE 372

Appendix 12 (continued) 359 Source DF Type III SS Mean Square F Value Pr > F group2 1 3.59469046 3.59469046 0.17 0.6769 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: postsus Susceptibility Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 30.743872 30.743872 0.95 0.3330 Error 85 2756.865323 32.433710 Corrected Total 86 2787.609195 R-Square Coeff Var Root MSE postsus Mean 0.011029 20.24807 5.695060 28.12644 Source DF Type I SS Mean Square F Value Pr > F group2 1 30.74387197 30.74387197 0.95 0.3330 Source DF Type III SS Mean Square F Value Pr > F group2 1 30.74387197 30.74387197 0.95 0.3330 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: posttb Benefits Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.1143879 0.1143879 0.01 0.9103 Error 85 761.2879109 8.9563284 Corrected Total 86 761.4022989 R-Square Coeff Var Root MSE posttb Mean 0.000150 8.038468 2.992713 37.22989 Source DF Type I SS Mean Square F Value Pr > F

PAGE 373

Appendix 12 (continued) 360 group2 1 0.11438793 0.11438793 0.01 0.9103 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.11438793 0.11438793 0.01 0.9103 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: postsev Severity Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 90.994113 90.994113 4.65 0.0339 Error 85 1663.902439 19.575323 Corrected Total 86 1754.896552 R-Square Coeff Var Root MSE postsev Mean 0.051852 17.03953 4.424401 25.96552 Source DF Type I SS Mean Square F Value Pr > F group2 1 90.99411270 90.99411270 4.65 0.0339 Source DF Type III SS Mean Square F Value Pr > F group2 1 90.99411270 90.99411270 4.65 0.0339 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: postcue Cue-to-Action Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 24.134884 24.134884 2.02 0.1592 Error 85 1016.922587 11.963795 Corrected Total 86 1041.057471 R-Square Coeff Var Root MSE postcue Mean 0.023183 19.05775 3.458872 18.14943 Source DF Type I SS Mean Square F Value Pr > F

PAGE 374

Appendix 12 (continued) 361 group2 1 24.13488378 24.13488378 2.02 0.1592 Source DF Type III SS Mean Square F Value Pr > F group2 1 24.13488378 24.13488378 2.02 0.1592 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: poststr Structural Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 8.1302337 8.1302337 0.77 0.3819 Error 85 894.4904560 10.5234171 Corrected Total 86 902.6206897 R-Square Coeff Var Root MSE poststr Mean 0.009007 26.35167 3.243982 12.31034 Source DF Type I SS Mean Square F Value Pr > F group2 1 8.13023366 8.13023366 0.77 0.3819 Source DF Type III SS Mean Square F Value Pr > F group2 1 8.13023366 8.13023366 0.77 0.3819 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: postmot Motivation Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 3.238874 3.238874 0.21 0.6463 Error 85 1298.278367 15.273863 Corrected Total 86 1301.517241 R-Square Coeff Var Root MSE postmot Mean 0.002489 21.06639 3.908179 18.55172

PAGE 375

Appendix 12 (continued) 362 Source DF Type I SS Mean Square F Value Pr > F group2 1 3.23887447 3.23887447 0.21 0.6463 Source DF Type III SS Mean Square F Value Pr > F group2 1 3.23887447 3.23887447 0.21 0.6463 Rsch Question 2 Posttests The GLM Procedure Dependent Variable: postbar Barriers Wk 24 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 198.56696 198.56696 0.55 0.4623 Error 85 30959.38706 364.22808 Corrected Total 86 31157.95402 R-Square Coeff Var Root MSE postbar Mean 0.006373 20.66427 19.08476 92.35632 Source DF Type I SS Mean Square F Value Pr > F group2 1 198.5669604 198.5669604 0.55 0.4623 Source DF Type III SS Mean Square F Value Pr > F group2 1 198.5669604 198.5669604 0.55 0.4623 Rsch Question 2 Posttests The GLM Procedure Repeated Measures Analysis of Variance Repeated Measures Level Information Dependent Variable postself postsus posttb postsev postcue poststr postmot postbar Level of subscale 1 2 3 4 5 6 7 8 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no subscale Effect H = Type III SSCP Matrix for subscale E = Error SSCP Matrix S=1 M=2.5 N=38.5 Statistic Value F Value Num DF Den DF Pr > F

PAGE 376

Appendix 12 (continued) 363 Wilks' Lambda 0.01091692 1022.50 7 79 <.0001 Pillai's Trace 0.98908308 1022.50 7 79 <.0001 Hotelling-Lawley Trace 90.60094620 1022.50 7 79 <.0001 Roy's Greatest Root 90.60094620 1022.50 7 79 <.0001 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no subscale*group2 Effect H = Type III SSCP Matrix for subscale*group2 E = Error SSCP Matrix S=1 M=2.5 N=38.5 Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.91975818 0.98 7 79 0.4485 Pillai's Trace 0.08024182 0.98 7 79 0.4485 Hotelling-Lawley Trace 0.08724230 0.98 7 79 0.4485 Roy's Greatest Root 0.08724230 0.98 7 79 0.4485 Rsch Question 2 Posttests The GLM Procedure Repeated Measures Analysis of Variance Tests of Hypotheses for Between Subjects Effects Source DF Type III SS Mean Square F Value Pr > F group2 1 20.449197 20.449197 0.29 0.5896 Error 85 5928.361148 69.745425 Rsch Question 2 Posttests The GLM Procedure Repeated Measures Analysis of Variance Univariate Tests of Hypotheses for Within Subject Effects Adj Pr > F Source DF Type III SS Mean Square F Value Pr > F G G H F subscale 7 386952.6550 55278.9507 935.18 <.0001 <.0001 <.0001 subscale*group2 7 339.0688 48.4384 0.82 0.5713 0.4066 0.4092 Error(subscale) 595 35170.7645 59.1105 Greenhouse-Geisser Epsilon 0.2036 Huynh-Feldt Epsilon 0.2086 Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for postself NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85

PAGE 377

Appendix 12 (continued) 364 Error Mean Square 20.56461 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.9365 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 31.1463 41 2_Control A A 30.7391 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for postself NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 20.56461 Critical Value of t 1.98827 Minimum Significant Difference 1.9365 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 31.1463 41 2_Control A A 30.7391 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for postsus NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 32.43371

PAGE 378

Appendix 12 (continued) 365 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 2.432 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 28.756 41 2_Control A A 27.565 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for postsus NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 32.43371 Critical Value of t 1.98827 Minimum Significant Difference 2.432 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 28.756 41 2_Control A A 27.565 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for posttb NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 8.956328 Critical Value of Studentized Range 2.81184

PAGE 379

Appendix 12 (continued) 366 Minimum Significant Difference 1.278 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 37.2683 41 2_Control A A 37.1957 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for posttb NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 8.956328 Critical Value of t 1.98827 Minimum Significant Difference 1.278 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 37.2683 41 2_Control A A 37.1957 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for postsev NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 19.57532 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.8894

PAGE 380

Appendix 12 (continued) 367 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 27.0488 41 2_Control B 25.0000 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for postsev NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 19.57532 Critical Value of t 1.98827 Minimum Significant Difference 1.8894 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 27.0488 41 2_Control B 25.0000 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for postcue NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 11.9638 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.4771 Harmonic Mean of Cell Sizes 43.35632

PAGE 381

Appendix 12 (continued) 368 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 18.7073 41 2_Control A A 17.6522 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for postcue NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 11.9638 Critical Value of t 1.98827 Minimum Significant Difference 1.4771 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 18.7073 41 2_Control A A 17.6522 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for poststr NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 10.52342 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.3853 Harmonic Mean of Cell Sizes 43.35632

PAGE 382

Appendix 12 (continued) 369 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 12.6341 41 2_Control A A 12.0217 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for poststr NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 10.52342 Critical Value of t 1.98827 Minimum Significant Difference 1.3853 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 12.6341 41 2_Control A A 12.0217 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for postmot NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 15.27386 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.6689 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal.

PAGE 383

Appendix 12 (continued) 370 Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 18.7561 41 2_Control A A 18.3696 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for postmot NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 15.27386 Critical Value of t 1.98827 Minimum Significant Difference 1.6689 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 18.7561 41 2_Control A A 18.3696 46 1_Intensive Rsch Question 2 Posttests The GLM Procedure Tukey's Studentized Range (HSD) Test for postbar NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 364.2281 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 8.1499 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal.

PAGE 384

Appendix 12 (continued) 371 Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 93.783 46 1_Intensive A A 90.756 41 2_Control Rsch Question 2 Posttests The GLM Procedure Bonferroni (Dunn) t Tests for postbar NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 364.2281 Critical Value of t 1.98827 Minimum Significant Difference 8.1499 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 93.783 46 1_Intensive A A 90.756 41 2_Control Rsch Question 2 Differences The GLM Procedure Class Level Information Class Levels Values group2 2 1_Intensive 2_Control Number of Observations Read 159 Number of Observations Used 87 Rsch Question 2 Differences The GLM Procedure Dependent Variable: selfd Difference in Self Efficacy

PAGE 385

Appendix 12 (continued) 372 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 41.803251 41.803251 1.74 0.1903 Error 85 2038.127784 23.977974 Corrected Total 86 2079.931034 R-Square Coeff Var Root MSE selfd Mean 0.020098 2028.646 4.896731 0.241379 Source DF Type I SS Mean Square F Value Pr > F group2 1 41.80325081 41.80325081 1.74 0.1903 Source DF Type III SS Mean Square F Value Pr > F group2 1 41.80325081 41.80325081 1.74 0.1903 Rsch Question 2 Differences The GLM Procedure Dependent Variable: susd Difference in Susceptibility Sum of Source DF Squares Mean Square F Value Pr > F Model 1 21.403176 21.403176 1.04 0.3113 Error 85 1753.654295 20.631227 Corrected Total 86 1775.057471 R-Square Coeff Var Root MSE susd Mean 0.012058 2469.800 4.542161 0.183908 Source DF Type I SS Mean Square F Value Pr > F group2 1 21.40317646 21.40317646 1.04 0.3113 Source DF Type III SS Mean Square F Value Pr > F group2 1 21.40317646 21.40317646 1.04 0.3113 Rsch Question 2 Differences The GLM Procedure Dependent Variable: tbd Difference in Benefits

PAGE 386

Appendix 12 (continued) 373 Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.0177716 0.0177716 0.00 0.9633 Error 85 707.4994698 8.3235232 Corrected Total 86 707.5172414 R-Square Coeff Var Root MSE tbd Mean 0.000025 643.5885 2.885052 0.448276 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.01777160 0.01777160 0.00 0.9633 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.01777160 0.01777160 0.00 0.9633 Rsch Question 2 Differences The GLM Procedure Dependent Variable: sevd Difference in Severity Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.008343 0.008343 0.00 0.9825 Error 85 1471.393955 17.310517 Corrected Total 86 1471.402299 R-Square Coeff Var Root MSE sevd Mean 0.000006 1809.857 4.160591 0.229885 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.00834339 0.00834339 0.00 0.9825 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.00834339 0.00834339 0.00 0.9825 Rsch Question 2 Differences The GLM Procedure

PAGE 387

Appendix 12 (continued) 374 Dependent Variable: cued Difference in Cue-to-Action Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.0013713 0.0013713 0.00 0.9911 Error 85 939.5848356 11.0539392 Corrected Total 86 939.5862069 R-Square Coeff Var Root MSE cued Mean 0.000001 -4820.882 3.324746 -0.068966 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.00137127 0.00137127 0.00 0.9911 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.00137127 0.00137127 0.00 0.9911 Rsch Question 2 Differences The GLM Procedure Dependent Variable: strd Difference in Structural Sum of Source DF Squares Mean Square F Value Pr > F Model 1 1.9249156 1.9249156 0.25 0.6191 Error 85 657.1325557 7.7309712 Corrected Total 86 659.0574713 R-Square Coeff Var Root MSE strd Mean 0.002921 234.8546 2.780462 1.183908 Source DF Type I SS Mean Square F Value Pr > F group2 1 1.92491559 1.92491559 0.25 0.6191 Source DF Type III SS Mean Square F Value Pr > F group2 1 1.92491559 1.92491559 0.25 0.6191 Rsch Question 2 Differences The GLM Procedure

PAGE 388

Appendix 12 (continued) 375 Dependent Variable: motd Difference in Motivation Sum of Source DF Squares Mean Square F Value Pr > F Model 1 0.0123414 0.0123414 0.00 0.9685 Error 85 668.2635207 7.8619238 Corrected Total 86 668.2758621 R-Square Coeff Var Root MSE motd Mean 0.000018 1355.224 2.803912 0.206897 Source DF Type I SS Mean Square F Value Pr > F group2 1 0.01234139 0.01234139 0.00 0.9685 Source DF Type III SS Mean Square F Value Pr > F group2 1 0.01234139 0.01234139 0.00 0.9685 Rsch Question 2 Differences The GLM Procedure Dependent Variable: bard Difference in Barriers Sum of Source DF Squares Mean Square F Value Pr > F Model 1 59.63106 59.63106 0.19 0.6619 Error 85 26327.21951 309.73199 Corrected Total 86 26386.85057 R-Square Coeff Var Root MSE bard Mean 0.002260 -1444.463 17.59920 -1.218391 Source DF Type I SS Mean Square F Value Pr > F group2 1 59.63106252 59.63106252 0.19 0.6619 Source DF Type III SS Mean Square F Value Pr > F group2 1 59.63106252 59.63106252 0.19 0.6619 Rsch Question 2 Differences

PAGE 389

Appendix 12 (continued) 376 The GLM Procedure Repeated Measures Analysis of Variance Repeated Measures Level Information Dependent Variable selfd susd tbd sevd cued strd motd bard Level of subscale 1 2 3 4 5 6 7 8 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no subscale Effect H = Type III SSCP Matrix for subscale E = Error SSCP Matrix S=1 M=2.5 N=38.5 Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.88546610 1.46 7 79 0.1939 Pillai's Trace 0.11453390 1.46 7 79 0.1939 Hotelling-Lawley Trace 0.12934872 1.46 7 79 0.1939 Roy's Greatest Root 0.12934872 1.46 7 79 0.1939 MANOVA Test Criteria and Exact F Statistics for the Hypothesis of no subscale*group2 Effect H = Type III SSCP Matrix for subscale*group2 E = Error SSCP Matrix S=1 M=2.5 N=38.5 Statistic Value F Value Num DF Den DF Pr > F Wilks' Lambda 0.95165321 0.57 7 79 0.7755 Pillai's Trace 0.04834679 0.57 7 79 0.7755 Hotelling-Lawley Trace 0.05080295 0.57 7 79 0.7755 Roy's Greatest Root 0.05080295 0.57 7 79 0.7755 Rsch Question 2 Differences The GLM Procedure Repeated Measures Analysis of Variance Tests of Hypotheses for Between Subjects Effects Source DF Type III SS Mean Square F Value Pr > F group2 1 15.303571 15.303571 0.28 0.5950 Error 85 4567.730912 53.738011 Rsch Question 2 Differences The GLM Procedure Repeated Measures Analysis of Variance Univariate Tests of Hypotheses for Within Subject Effects Adj Pr > F Source DF Type III SS Mean Square F Value Pr > F G G H F subscale 7 259.68257 37.09751 0.74 0.6416 0.4426 0.4458 subscale*group2 7 109.49866 15.64267 0.31 0.9494 0.6666 0.6722

PAGE 390

Appendix 12 (continued) 377 Error(subscale) 595 29995.14502 50.41201 Greenhouse-Geisser Epsilon 0.2110 Huynh-Feldt Epsilon 0.2164 Rsch Question 2 Differences The GLM Procedure Tukey's Studentized Range (HSD) Test for selfd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 23.97797 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 2.0911 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 0.9756 41 2_Control A A -0.4130 46 1_Intensive Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for selfd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 23.97797 Critical Value of t 1.98827 Minimum Significant Difference 2.0911 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2

PAGE 391

Appendix 12 (continued) 378 A 0.9756 41 2_Control A A -0.4130 46 1_Intensive Rsch Question 2 Differences The GLM Procedure Tukey's Studentized Range (HSD) Test for susd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 20.63123 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.9397 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 0.6522 46 1_Intensive A A -0.3415 41 2_Control Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for susd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 20.63123 Critical Value of t 1.98827 Minimum Significant Difference 1.9397 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2

PAGE 392

Appendix 12 (continued) 379 A 0.6522 46 1_Intensive A A -0.3415 41 2_Control Rsch Question 2 Differences The GLM Procedure Tukey's Studentized Range (HSD) Test for tbd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 8.323523 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.232 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 0.4634 41 2_Control A A 0.4348 46 1_Intensive Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for tbd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 8.323523 Critical Value of t 1.98827 Minimum Significant Difference 1.232 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 0.4634 41 2_Control

PAGE 393

Appendix 12 (continued) 380 A A 0.4348 46 1_Intensive Rsch Question 2 Differences The GLM Procedure Tukey's Studentized Range (HSD) Test for sevd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 17.31052 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.7767 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 0.2391 46 1_Intensive A A 0.2195 41 2_Control Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for sevd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 17.31052 Critical Value of t 1.98827 Minimum Significant Difference 1.7767 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 0.2391 46 1_Intensive A

PAGE 394

Appendix 12 (continued) 381 A 0.2195 41 2_Control Rsch Question 2 Differences The GLM Procedure Tukey's Studentized Range (HSD) Test for cued NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 11.05394 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.4198 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A -0.06522 46 1_Intensive A A -0.07317 41 2_Control Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for cued NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 11.05394 Critical Value of t 1.98827 Minimum Significant Difference 1.4198 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A -0.06522 46 1_Intensive A A -0.07317 41 2_Control

PAGE 395

Appendix 12 (continued) 382 Rsch Question 2 Differences The GLM Procedure Tukey's Studentized Range (HSD) Test for strd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 7.730971 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.1874 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 1.3415 41 2_Control A A 1.0435 46 1_Intensive Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for strd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 7.730971 Critical Value of t 1.98827 Minimum Significant Difference 1.1874 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 1.3415 41 2_Control A A 1.0435 46 1_Intensive

PAGE 396

Appendix 12 (continued) 383 Rsch Question 2 Differences The GLM Procedure Tukey's Studentized Range (HSD) Test for motd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 7.861924 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 1.1974 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A 0.2195 41 2_Control A A 0.1957 46 1_Intensive Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for motd NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 7.861924 Critical Value of t 1.98827 Minimum Significant Difference 1.1974 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A 0.2195 41 2_Control A A 0.1957 46 1_Intensive Rsch Question 2 Differences

PAGE 397

Appendix 12 (continued) 384 The GLM Procedure Tukey's Studentized Range (HSD) Test for bard NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 309.732 Critical Value of Studentized Range 2.81184 Minimum Significant Difference 7.5155 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Tukey Grouping Mean N group2 A -0.341 41 2_Control A A -2.000 46 1_Intensive Rsch Question 2 Differences The GLM Procedure Bonferroni (Dunn) t Tests for bard NOTE: This test controls the Type I experimentwise error rate, but it generally has a higher Type II error rate than REGWQ. Alpha 0.05 Error Degrees of Freedom 85 Error Mean Square 309.732 Critical Value of t 1.98827 Minimum Significant Difference 7.5155 Harmonic Mean of Cell Sizes 43.35632 NOTE: Cell sizes are not equal. Means with the same letter are not significantly different. Bon Grouping Mean N group2 A -0.341 41 2_Control A A -2.000 46 1_Intensive Success at Week 12 by Group

PAGE 398

Appendix 12 (continued) 385 The FREQ Procedure Table of group2 by suc12 group2(Group) suc12(Success at Wk 12) Frequency ‚ Percent ‚ Row Pct ‚ Col Pct ‚1_Succes‚2_Unsucc‚ Total ‚sful ‚essful ‚ ƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆ 1_Intensive ‚ 33 ‚ 27 ‚ 60 ‚ 27.05 ‚ 22.13 ‚ 49.18 ‚ 55.00 ‚ 45.00 ‚ ‚ 55.00 ‚ 43.55 ‚ ƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆ 2_Control ‚ 27 ‚ 35 ‚ 62 ‚ 22.13 ‚ 28.69 ‚ 50.82 ‚ 43.55 ‚ 56.45 ‚ ‚ 45.00 ‚ 56.45 ‚ ƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆ Total 60 62 122 49.18 50.82 100.00 Statistics for Table of group2 by suc12 Statistic DF Value Prob ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ Chi-Square 1 1.5999 0.2059 Likelihood Ratio Chi-Square 1 1.6034 0.2054 Continuity Adj. Chi-Square 1 1.1745 0.2785 Mantel-Haenszel Chi-Square 1 1.5868 0.2078 Phi Coefficient 0.1145 Contingency Coefficient 0.1138 Cramer's V 0.1145 Fisher's Exact Test ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ Cell (1,1) Frequency (F) 33 Left-sided Pr <= F 0.9261 Right-sided Pr >= F 0.1392 Table Probability (P) 0.0653 Two-sided Pr <= P 0.2771 Success at Week 12 by Group The FREQ Procedure Statistics for Table of group2 by suc12 Estimates of the Relative Risk (Row1/Row2) Type of Study Value 95% Confidence Limits ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ Case-Control (Odds Ratio) 1.5844 0.7753 3.2376

PAGE 399

Appendix 12 (continued) 386 Cohort (Col1 Risk) 1.2630 0.8774 1.8180 Cohort (Col2 Risk) 0.7971 0.5589 1.1369 Sample Size = 122 Success at Week 24 by Group The FREQ Procedure Table of group2 by suc24 group2(Group) suc24(Success at Wk 24) Frequency ‚ Percent ‚ Row Pct ‚ Col Pct ‚1_Succes‚2_Unsucc‚ Total ‚sful ‚essful ‚ ƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆ 1_Intensive ‚ 29 ‚ 24 ‚ 53 ‚ 28.16 ‚ 23.30 ‚ 51.46 ‚ 54.72 ‚ 45.28 ‚ ‚ 50.00 ‚ 53.33 ‚ ƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆ 2_Control ‚ 29 ‚ 21 ‚ 50 ‚ 28.16 ‚ 20.39 ‚ 48.54 ‚ 58.00 ‚ 42.00 ‚ ‚ 50.00 ‚ 46.67 ‚ ƒƒƒƒƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆƒƒƒƒƒƒƒƒˆ Total 58 45 103 56.31 43.69 100.00 Statistics for Table of group2 by suc24 Statistic DF Value Prob ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ Chi-Square 1 0.1127 0.7371 Likelihood Ratio Chi-Square 1 0.1128 0.7370 Continuity Adj. Chi-Square 1 0.0188 0.8910 Mantel-Haenszel Chi-Square 1 0.1116 0.7383 Phi Coefficient -0.0331 Contingency Coefficient 0.0331 Cramer's V -0.0331 Fisher's Exact Test ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ Cell (1,1) Frequency (F) 29 Left-sided Pr <= F 0.4457 Right-sided Pr >= F 0.7033 Table Probability (P) 0.1490 Two-sided Pr <= P 0.8429 Success at Week 24 by Group The FREQ Procedure

PAGE 400

Appendix 12 (continued) 387 Statistics for Table of group2 by suc24 Estimates of the Relative Risk (Row1/Row2) Type of Study Value 95% Confidence Limits ƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒƒ Case-Control (Odds Ratio) 0.8750 0.4012 1.9082 Cohort (Col1 Risk) 0.9434 0.6715 1.3255 Cohort (Col2 Risk) 1.0782 0.6943 1.6742 Sample Size = 103

PAGE 401

About the Author Jodee Meisenhelder-Smith received her Bachelors Degree in Food Science (minor in nutrition in 1978 and a Ma sters of Nutritional Science in 1980 from Clemson University. She was on t he Dean’s list at Clemson. She became a Registered Dietitian in 1980 and has work ed in the profession for 26 years. She became a Florida Licensed Dietitian in 1980. She became a certified Nutrition Specialist in 1994.and a Certified Diabete s Educator in 1995. She entered the Ph.D. program at the Univer sity of South Florida in 1994. She has worked as a Certified Diabetes Educator Dietitian at University Community Hospital for 11 years. While at USF she was on the Dean’s List, and was awarded a Public Health Scholarship and a Maternal and Ch ild Health training grant. She has been active in several professional organi zations including American Diabetes Association, American Association of Diabetes Educator s, and the American Dietetic Association.


xml version 1.0 encoding UTF-8 standalone no
record xmlns http:www.loc.govMARC21slim xmlns:xsi http:www.w3.org2001XMLSchema-instance xsi:schemaLocation http:www.loc.govstandardsmarcxmlschemaMARC21slim.xsd
leader nam Ka
controlfield tag 001 001920107
003 fts
005 20080104115121.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080104s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001837
035
(OCoLC)187304443
040
FHM
c FHM
049
FHMM
090
RA425 (ONLINE)
1 100
Meisenhelder-Smith, Jodee.
4 245
The effects of American Diabetes Association (ADA) diabetes self-management education and continuous glucose monitoring on diabetes health beliefs, behaviors and metabolic control
h [electronic resource] /
by Jodee Meisenhelder-Smith.
260
[Tampa, Fla] :
b University of South Florida,
2006.
3 520
ABSTRACT: The purpose of this study was to determine whether adults with type 2 diabetes participating in American Diabetes Association (ADA) diabetes self-management education (DSME) randomly assigned to an intensive follow-up group (IFG), utilizing continuous glucose monitoring system (CGMS), or a standard follow-up group (SFG) have any significant differences in mean HgbA1c values and health belief scores over time. Baseline HgbA1c values and health beliefs were measured using the revised Expanded Health Belief Model (HBM) questionnaire. The questionnaire measured the 8 HBM domains: perceived susceptibility; severity; treatment benefit; cues to action; motivation; barriers; self-efficacy and structural elements. Twelve weeks after DSME, patients returned for follow-up based on random assignment. The SFG received routine follow-up care: HgbA1c measurements; behavioral goals and education assessments. The IFG received routine follow-up and CGMS.^ Patients wore the CSMS for 72 hours and recorded their daily food, blood glucose values, medications and physical activities. Results were analyzed and reviewed with patients. Both groups returned in 24 weeks for HgbA1c measurements and to complete the HBM questionnaire. A repeated measure ANOVA analysis showed a statistically significant reduction in mean HgbA1c at each time period (F=86.75. p>.0001 ) from week 1 to week 12 (SFG 8.6-7.1; IFG 8.5 --7.1,) and from week 12 to week 24 ( SFG 7.1 to 6.9; IFG 7.1 --^ 7.0). There were no significant differences found between the groups. (F = 0.17 p > 0.87). Following DMSE and follow-up intervention some health belief scores improved but no significant differences were found between groups except for severity scores. (SFG 27.05, IFG 25.00, p=0.03). The power of the study to detect small differences between the groups was affected by the higher than anticipated attrition and the significant lowering of HgbA1c in the education arm of the study. Both groups achieved a high success rate (58% IFG; 55% SFG) to lower the HgbA1c to the ADA goal of less than 7. DSME and follow-up care (both standard follow-up and more intensive follow-up) achieved a significant lowering of HgbA1c (1.6%), which has been shown to reduce diabetes related morbidity and health costs.
502
Dissertation (Ph.D.)--University of South Florida, 2006.
504
Includes bibliographical references.
516
Text (Electronic dissertation) in PDF format.
538
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
500
Title from PDF of title page.
Document formatted into pages; contains 387 pages.
Includes vita.
590
Adviser: Robert J. McDermott, Ph.D.
653
Health belief model.
Glycosylated hemoglobin.
Diabetes patient care assessment.
Diabetes outcomes.
Outpatient program.
0 690
Dissertations, Academic
z USF
x Public Health
Doctoral.
773
t USF Electronic Theses and Dissertations.
856
u http://digital.lib.usf.edu/?e14.1837