USF Libraries
USF Digital Collections

Advanced registered nurse practitioners' judgments of coronary heart disease risk

MISSING IMAGE

Material Information

Title:
Advanced registered nurse practitioners' judgments of coronary heart disease risk
Physical Description:
Book
Language:
English
Creator:
Stamp, Kelly D
Publisher:
University of South Florida
Place of Publication:
Tampa, Fla
Publication Date:

Subjects

Subjects / Keywords:
Decision-making
Social judgment theory
Clinical assessment
Nursing
Patients
Dissertations, Academic -- Nursing -- Doctoral -- USF   ( lcsh )
Genre:
bibliography   ( marcgt )
theses   ( marcgt )
non-fiction   ( marcgt )

Notes

Summary:
ABSTRACT: Coronary heart disease (CHD) is the single largest killer of American males and females in the United States. According to the American Heart Association, (2005) approximately 41% of Americans that experience a coronary attack in a given year will die from it (AHA, 2005). To combat this growing problem, strategies need to be evaluated to assess how the identification of actual and potential CHD risks are made. This study utilized the Social Judgment Theory to gain insight into nurse practitioner's decision-making strategies. Sixty family or adult specialty nurse practitioners affiliated with the University of South Florida (USF) College of Nursing volunteered to take part in a pretest-posttest experimental design. They were randomly assigned to one of three conditions. Condition 1 and 2 received educational interventions and Condition 3 served as the control group, which received no education. This design was used to assess the effects of educational feedback on improving judgment accuracy, achievement, and insight. The findings indicated nurse practitioners agreement with the Framingham prediction model of CHD risk did improve significantly for the two intervention groups from Time 1 to Time 2 (p < .05). the participants also showed a relatively high degree of cognitive control when judging and performing the policy-capturing task (average Rs = .88) as compared to Framingham (Re = .96). Significant amount of unconditional bias (F(2, 57) = 9.85, p < .01) and conditional bias (F(2, 57). 5.42), p < .05) was present in this sample. Nurse practitioners overall performed well when compared with the Framingham Heart Study risk equation, however, nurse practitioners showed little insight into their judgment process. The results of this study may provide the opportunity for nurse practitioners to offer patients more appropriate medicinal and diagnostic treatments. Future cardiac events may be avoided through evidenced-based CHD education for nursepractitioners.
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 Kelly D. Stamp.
General Note:
Title from PDF of title page.
General Note:
Document formatted into pages; contains 136 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 - 001929724
oclc - 212850051
usfldc doi - E14-SFE0001811
usfldc handle - e14.1811
System ID:
SFS0026129:00001


This item is only available as the following downloads:


Full Text

PAGE 1

Advanced Registered Nurse Pr actitioners’ Judgments of Co ronary Heart Disease Risk by Kelly D. Stamp A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Health Sciences Center College of Nursing University of South Florida Major Professor: Mary Webb, Ph.D., RN Major Co-Professor: Mary Evans, Ph.D., RN Jason Beckstead, Ph.D. Carol Bryant, Ph.D., MPH November 1, 2006 Key Words: decision-making, Social Judgmen t Theory, clinical assessment, nursing, patients Copyright 2006, Kelly D. Stamp

PAGE 2

To my husband Tim… for always being there to encour age, support, and believe in me… for telling me to keep my dreams alive…to know that achievement of anything requires faith and belief in yourself, vision, hard work, determination, and dedication. Always remembering…anything is possible.

PAGE 3

Acknowledgements I wish to acknowledge the dissertation committee members with appreciation and gratitude. Dr. Mary Webb, Dissertation Chair, fo r her support, encouragement, and guidance from the development to the completion of the project. It has been a privilege to perform research under her direction. Dr. Mary Evans, Co-Dissertation Ch air, for her research expertise and contribution; her thoughtful review of my ma nuscript and her willingness to participate on the committee in spite of her extremely busy schedule. She is a great role model of the nurse scientist. Dr. Jason Beckstead for his statisti cal expertise, guida nce, and patience throughout the pilot and disserta tion study along with defense pr eparation. He remained available for all of my ques tions and provided encouragemen t during frustrating times. Dr. Carol Bryant for her support, encour agement, and public health perspective; she made herself readily available for committee meetings and manuscript reviews. It is with my great pleasure to rec ognize and thank the members of my committee for their availability and willingness to accomm odate my questions, edits, and meetings. They have provided scholarly expertise and re search mentorship while making the ride to doctoral preparation challenging and enjoyable.

PAGE 4

i Table of Contents List of Tables iii List of Figures iv Abstract v Chapter One: Introduction 1 Purpose of Study 14 Statement of Problem 15 Specific Aims 15 Definition of Terms 16 Significance of This Study 18 Chapter Summary 18 Chapter Two: Review of Literature 20 Theoretical Framework 20 Nurses Decision-Making 26 Physician Decision-Making 28 Insight and Judgment Policies 31 Cognitive Feedback and Judgment 33 Importance of Studying ARNP’s 34 Importance of the Framingham Heart Study 37 Chapter Summary 37 Chapter Three: Research Design 40 Sample Size and Setting 41 Materials 41 Policy Capturing Task 41 Scenario Example 43 Self-Report Cue Importance 43 Demographic Information 44 Institutional Review Board 44 Procedures 44 Session 1 44 Session 2 46 Data Analysis 46 Chapter Summary 48 Chapter Four: Results 49

PAGE 5

ii Sample of Judges 50 Performance of Judges at Time 1 50 Hypothesis Testing 53 Changes in ra, Rs, and G 53 Distribution of Importance Weights Among Eight Cues 57 Chapter Summary 64 Chapter Five: Discussion 66 Discussion 66 Performance of Judges for Time 1 66 Distribution of Importance Weights Among the Eight Cues 69 Implications for Judgment Researchers 73 Implications for Nurse Practitioners and Educators 73 Limitations 74 Foundations for Future Research 76 References 77 Appendix A: Brunswik Lens Model 92 Appendix B: Experimental Manipulations 93 Appendix C: Patient Profiles 95 About the Author End Page

PAGE 6

iii List of Tables Table 1 Sample Composition 51

PAGE 7

iv List of Figures Figure 1. Lens Model 12 Figure 2. Brunswik Lens Model 22 Figure 3. Logic Model 36 Figure 4. Example of Patient Profile 43 Figure 5. Example of Comparative Feedback 45 Figure 6. Amount of Agreement Among NPs and Ecology 54 Figure 7. Cognitive Control Among ARNP Judges 55 Figure 8. Agreement Among NPs Model and Ecology Model 56 Figure 9. Unconditional Bias (Elevation) 57 Figure 10. Conditional Bias (Scatter) 57 Figure 11. Judges Assigned Gender Weight 59 Figure 12. Judges Assigned Age Weight 60 Figure 13. Judges Assigned Systolic Blood Pressure Weight 61 Figure 14. Judges Assigned Left Ventricular Hypertrophy Weight 61 Figure 15. Judges Assigned Total Cholesterol Weight 62 Figure 16. Judges Assigned High Density Lipoprotein Weight 63 Figure 17. Judges Assigned Smoking Status Weight 63 Figure 18. Judges Assigned Diabetes Status Weight 64

PAGE 8

v Advanced Registered Nurse Pr actitioners’ Judgments of Co ronary Heart Disease Risk Kelly D. Stamp ABSTRACT Coronary heart disease (CHD) is the singl e largest killer of American males and females in the United States. According to the American Hear t Association, (2005) approximately 41% of Americans that experien ce a coronary attack in a given year will die from it (AHA, 2005). To combat this gr owing problem, strategies need to be evaluated to assess how the identification of actual and potential CHD risks are made. This study utilized the Social Judgment Theory to gain insight into nurse practitioner’s decision-making strategies. Sixty family or adult specialty nurse practit ioners affiliated with the University of South Florida (USF) College of Nursing volunt eered to take part in a pretest-posttest experimental design. They were randomly assigned to one of three conditions. Condition 1 and 2 received e ducational interventions and Condition 3 served as the control group, which received no education. This design was us ed to assess the effects of educational feedback on improving judgme nt accuracy, achievement, and insight. The findings indicated nurse practiti oners agreement with the Framingham prediction model of CHD risk did improve si gnificantly for the tw o intervention groups from Time 1 to Time 2 (p < .05). The partic ipants also showed a relatively high degree of cognitive control when judging and perfor ming the policy-capturing task (average Rs = .88) as compared to Framingham (Re = .96). Significant amount of unconditional bias

PAGE 9

vi (F(2, 57) = 9.85, p < .01) and conditional bias (F(2, 57) = 5.42), p < .05) was present in this sample. Nurse practitioners overall performed well when compared with the Framingham Heart Study risk equation, however, nurse practitioners showed little insight into their judgme nt process. The results of this study may provide the opportunity for nurse practitioners to offer patients more appropriate medicinal and diagnostic treatments. Future cardiac events may be avoided through evidenced-bas ed CHD education for nurse practitioners.

PAGE 10

1 Chapter One Introduction Coronary Heart Disease Coronary heart disease (CHD) is the singl e largest killer of American males and females residing in the United States. Acco rding to the American Heart Association (AHA), (2005) approximately 41% of Americans that experience a coro nary attack in a given year will die from the event (AHA, 2005). To combat this growing problem, strategies need to be evalua ted to assess how the identific ation of actual and potential CHD risk is made. Many primary care physicians employ advanced registered nurse practitioners (ARNPs). Practitioners are one of the first lines of defense towards the primary prevention of CHD (American Academ y of Nurse Practitioners, 2002). They are at the forefront of assessment, detection, a nd treatment of potential and actual CHD risk factors for their primary care patients. In the early 1980’s studies using the Social Judgment Theory were developed with a focus on understanding the healthcare provider’s decision-making strategi es. At that time, the role of the nurse practitioner was at its early stages rendering unavailable samp le sizes to study. Presently, the role has greatly expanded and is so widely used th at nurse practitioners are now considered primary care providers. This study evaluated the decisionmaking of ARNPs so their ability to accurately detect CHD risks could be validated an d re-validated if necessary. Incidence and Prevalence of Coronary Heart Disease

PAGE 11

2 Coronary heart disease accounts for more th an half of all cardiovascular events in men and women under 75 years of age (AHA, 2005; Hughes & Haymann, 2004). The average age of a person having a first heart attack is 65.8 years for men and 70.4 years for women (Hurst, 2002). Forty-nine percent of men and 32% of women have a lifetime risk of developing CHD after 40 years of age (AHA, 2005). During the year of 2005 it is estimated that 700,000 Americans will have a new coronary attack and approximately 500,000 will have a recurrent at tack. Women lag behind men by 10 years for total CHD and by 20 years for more serious clinical even ts such as heart attack and sudden cardiac death (AHA, 2005). Age-adjusted CHD inci dence rates per 1,000 person years were: white men, 12.5; black men, 10.6; white wo men, 4.0; and black women, 5.1 (Jones, Chambless, Folsom, Heiss, et al., 2002). The prevalence of CHD in 2002 cons isted of 13,000,000 Americans; 7,100,000 were men and 5,900,000 were women. Reported prevalence of myocardial infarctions was of 4,100,000 men and 3,000,000 women. Sim ilarly the prevalence of new and recurrent heart attack an d fatal CHD events consisted of 715,000 men and 485,000 women (AHA, 2005). According to the Nati onal Health and Nutr ition Examination Survey (NHANES) 1999-2002, among Americans ages 40-74, the age adjusted prevalence of self-reported myocardial in farction and verified electrocardiogram myocardial infarction were higher among men than women; however angina prevalence was higher in women than men. It is estimat ed by the American H eart Association that in the year 2005 the direct and indirect cost of CHD will reach 142.1 billion dollars.

PAGE 12

3 Coronary Heart Disease Mortality Rates In 2002, CHD caused one of every five deat hs in the U.S. while total mortality reached 656,000. It is estimated that every 26 seconds an American will suffer a coronary event, and every minute a person will die fr om one (AHA, 2005). In fact, 50% of men and 64% of women who died suddenly of CHD had no previous symptoms of the disease (AHA, 2005). It is estimated that twenty-five percent of me n and thirty-eight percent of women will die within one year after having an initial recognized myocardial infarction (AHA, 2005). The higher death rate in wome n is partially a result of females having heart attacks at an older age than men, whic h renders women more likely to die within a few weeks. CHD Risk Factors and Gender Differences A number of CHD risk f actors have been identified. For the purposes of this study eight CHD risk factors will be discussed. These factors were selected based on the recommendation by the Framingham Heart St udy (Anderson, Wilson, Odell, & Kannel, 1991) and the American Heart Association (2005 ) for assessment of patient risk. This section contains facts, figures, and outcomes concerning the risk factors for coronary heart disease: age, gender, hyperlipidemia (t otal cholesterol and hi gh-density lipoprotein), hypertension, smoking, diabetes, an d left ventricular hypertrophy. Age and Gender Compared to men, pre-menopausal women are more protected from coronary heart disease and a cardiac event (AHA, 2005). However, a woman’s risk and mortality concerning CHD increases with age; in contra st mortality for men is particularly high under the age of 60 (AHA, 2005).

PAGE 13

4 Hyperlipidemia This section will explore both total c holesterol and highdensity lipoprotein (HDL-C). Approximately 70% of all U.S. women have at least one major CHD risk factor such as hyperlipidemia. Beginning at the age of 45 years, women have a higher percentage of total cholesterol (TC) th an men (between 200 and 239 mg/dL) (AHA, 2005). Of women aged 20 years and older 53.6% of white, 46.4% of African Americans, and 44.7% of Hispanics have a total chol esterol level over 20 0 mg/dL (AHA, 2005). Total cholesterol is composed of high-densit y lipoprotein (HDL or "good") cholesterol, low-density lipoprotein (LDL or "bad") cholesterol and very-low density lipoprotein (VLDL), which carry triglycerides. Calories in gested in a meal and not used immediately by tissues are converted to triglycerides and tr ansported to fat cells to be stored (AHA, 2006). The risk of heart attack in both me n and women is highest when their total cholesterol is high and the high-density lipopr otein cholesterol is lower than 40 mg/dL (AHA, 2006). Hypertension One of the major CHD risk factors is hypertension (HTN). Nearly one-in-three adults in the U.S. have high blood pressure (Fields, Burt, Cutler, Hughes, Roccelia, & Sorlie, 2004). Approximately 28% of American adults over the age of 18 years have prehypertension which is defined as bl ood pressures of 120–139/80–89 mm Hg (AHA, 2006; Center for Disease Control (CDC), 2005; National Heart, Lung, and Blood Institute (NHLBI), 2005). Of those with HTN, 30% we re not aware, thirty-four percent were medicated and were controlled, twenty-five percent were medicated and not controlled; and eleven percent were not medicated (JNC 7 2004). A higher percentage of men than

PAGE 14

5 women have HTN through the age of 55; thereaf ter the percentage of HTN in women is higher (AHA, 2005). High blood pressure contributed to appr oximately 261,000 deaths in 2002 and had an estimated direct and indirect cost of 59.7 billion dollars in 2005 (AHA, 2005). Hypertension causes an increased workload on a person’s heart and arteries. If high blood pressure persist organs such as the hear t, kidneys, and brain may be affected. When hypertension is coupled with smoking, increasing age, and hyperlipidemia, the risk for a coronary event is doubled (AHA, 2005). Smoking Smoking is the most preventable cause of premature CHD deaths in the United States (AHA, 2005). It accounts for more than 440,000 deaths per year. It can increase blood pressure, decrease exercise tolerance, and increase the tendency for blood clotting (AHA, 2005). Statistics indicate that 25.2 million men and 20.0 million women smoke tobacco products. Among various ethnic gr oups (Whites, African American, Hispanic, Asians, and American Indian/Alaska Natives), the American Indian/Alaska Natives have the highest incidence of cigarette sm oking in both men and women (AHA, 2005). Cigarette smoking is so widespread in the U.S. and such a signi ficant risk factor that it is now considered the leading preventable cause of disease and deaths in the United States (AHA, 2005). On average, male smokers die 13.2 years earlier than male nonsmokers, and female smokers die 14.5 years earlier th an female nonsmokers (Surgeon General, 2004). Cigarette smoking results in a two-to-t hree fold risk of dying from CHD (AHA, 2005). The estimated annual direct and indirect cost from smoking is 155 billion dollars (AHA, 2005).

PAGE 15

6 Diabetes Diabetes greatly increases the risk for CHD, even when blood glucose levels are well-controlled. More than 80% of people w ith diabetes die of heart or blood vessel disease. Approximately 3 million women and 2.9 million men have undiagnosed diabetes in the U.S (AHA, 2005). Approximately 6 million women and 8.5 million men have been diagnosed with pre-diabetes, (a fasti ng blood glucose level of 110 to 126 mg/dl). Finally, approximately 7 million women and 6.8 million men have a medical diagnosis of diabetes. On average, non-Hispanic bl ack women have the highest incidence of physician-diagnosed diabetes followed closely by white women (AHA, 2005). A person with diabetes is two to four times more like ly to die from heart di sease compared to nondiabetics (AHA, 2005). In 2002, the total direct and indirect cost from diabetes was 132 billion dollars. Left Ventricular Hypertrophy Left ventricular hypertrophy is a conditi on consisting of an enlargement of the left side of the heart (AHA, 2006). A thickening of the heart muscle as a result of an increased workload can cause left vent ricular hypertrophy (L VH); this increased workload could be a result of any one or more of the risk factors listed above. One of the main contributing factors to the develo pment of LVH is hypertension. High blood pressure increases the resistance of the circulatory system and forces the left ventricle to work harder in order to pump the blo od to meet the body’s oxygen demands (AHA, 2006). This increased workload eventually cau ses the ventricles to become enlarged and inefficient leading to chronic heart failure.

PAGE 16

7 Role of Advanced Health Care Providers In general, the role of the health care provider is assessment, diagnosis, treatment, and education of the patient. Advanced practice healthcare providers must recognize potential and actual cardiovascular risk factors in their male and female patients. They design primary or secondary treatment plans and perform education on lifestyle changes and/or medicinal treatments available to lesse n these risk factors for the patient involved. Two types of advanced healthcare provi ders will be discussed: (1) medical doctors and (2) advanced regist ered nurse practitioners or “n urse practitioners”. Both providers have advanced degrees in general medicine or specialty areas of nursing. A medical doctor completes a four-year gra duate degree in medical school along with internships and residencies before moving to private practice. Nurse practitioners have completed a two-year graduate degree in nursi ng, which includes clinical residencies in a variety of clinical specialties, e.g. family practice, adult me dicine, pediatrics or women’s health. The degree to which a nurse practitioner is allowed to practi ce independent of a medical doctor varies among the 50 states. Ho wever, all nurse practitioners are allowed to make medical diagnoses and prescribe pres cription medications. Most graduate nurse practitioners work in a primary care/family practice type of setting. However, a smaller percentage of nurse practitioners work in specialty areas such as cardiovascular medicine. Studies have been conducted in the past to determine whether nurse practitioners can provide comparable care, as do physicians. The results have indicated that nurse practitioners can give equivalent care and th at patients’ perception of their care is much higher. This is likely due to the extra time and health edu cation that nurs e practitioners

PAGE 17

8 tend spend with individual pa tients during consultation (H orrock, Anderson, & Salisbury, 2002; Kinnersley, Andrson, Parry, Clement, Arch ard, et al., 2000). Other studies have been conducted measuring the outcomes of patients belonging to physician groups only compared to physician/nurse practitioner groups. The results indicated that level of care was of similar quality; however, physicia n/nurse practitioner group reported seeing patients more often and providing a cost saving to the healthcare system due to the lower fees of an ARNP visit (Aigner, Drew, & Phipps, 2004). The Framingham Study The Framingham study began in 1948 to ev aluate the circumstances under which CHD occurs, develops and becomes fatal in the general population (AHA, 2005). The intention was to conduct a l ongitudinal study to help unders tand how those that develop CHD differ from those individuals who re main free of disease. Throughout the Framingham Study there have been three cohorts created: the original cohort consisted of 5,029 men and women in 1948; the second coho rt called “the Offs pring Cohort” was developed in 1971 and consisted of 5,124 men and women, and the third cohort created was named the “the Generation III Cohort,” which consists of the offspring of the second cohort and is under current recruitment with a goal sample size of 3,500. As a result of this study the investigators developed coronary heart disease risk e quations. Clinicians use the equation for predicting the development of CHD in those that are free of disease (AHA, 2005; Anderson, Wilson, Odell, & Kannel, 1991). They are based on a nonproportional hazards Weibull accelerated fa ilure time model (Anderson, 1991). The model was applied to eight risk factor s measured on 2,983 women and 2,590 men (age ranged from 30 72 years) from the Fram ingham and Framingham Offspring Cohorts.

PAGE 18

9 The equations and explanation used in this study are provided by Ande rson, et al. (1991). Conceptual Framework Social Judgment Theory and Decision-Making Social Judgment Theory was first used four decades ago to analyze how people make decisions or judgments considering the cues and stimuli in their environment. Since then it has been used in meteorol ogical forecasting, educational decision making, accounting, risk judgments, social welfare, medical and health-related decision making and ethics, risk judgments, and public projec t evaluations (Cooksey, 1996). Therefore, it will be used as the theory guiding the methodol ogy of this study. Social Judgment Theory has also been used to analyze how indi viduals make judgments about ecological situations or probabilit ies of occurrences. Hammond, Stewart, Brehmer, and Steinma nn (1975) conceptualized the SJT to explain how judgments and decisions were fo rmed retrospectively. To understand and model the process of cue u tilization, researchers have developed data collection techniques called “policy capturing.” Polic y capturing derived from SJT as a method used to study representative, samples of alte rnatives between attri butes and the judgment to be made (Cooksey, 1996). Policy capturing he lps define how individuals evaluate and combine evidence from multiple cues to arrive at judgments about different situations (Holzworth et al.,1999). Policy capturing can be thought of as an individualized multiple regression equation. Basicall y, an individual makes a judgm ent regarding each of a series of cue profiles; these judgments are then regressed on the cues in order to obtain a weighted linear composite which character izes the individual’s method for combining cue information into a judgment (Cookse y, 1996). Many of the earlier works using

PAGE 19

10 judgment theory date back to the theo retical and methodologi cal contributions by Brunswik and what he viewed as “Probabilistic Functioning.” This approach attempts to understand the relationship between the person and their environmen t (the ecology) and how these factors may affect human judgment or decisions via per ceptual cues (Cooksey, 1996). He also developed the “Lens Model” of human cognition and information, which will be used to form the basis of analys is for this study. The Lens Model will be described in depth late r in this chapter. Hammond (1996) proposed the integration of the Cognitive Continuum Theory with Social Judgment Theory. His intenti on was for the Cognitive Continuum Theory to be a culmination of an extended history of ideas concerning human cognition originating from Brunswik. The Cognitive Continuum Th eory focuses specifically on the friction and division that exists betw een intuitive and analytical th inking. The continuum is seen as intuitive cognition at one pole and analytical cognition at the other pole with quasirationality in the middle. Intuitive cogn ition is considered to be rapid, covert, nonretraceable, inconsistent, with high confidence in outcome and low in process, and the errors are small and normally distributed (Cooks ey, 1996). At the othe r pole lies analytic cognition, which is slow and sequential in nature, retraceable, consistent, logical, low confidence in outcome and high in process, erro rs tend to be large, and there is a large reliance on quantitative cues (Cooksey, 1996) The middle portion of the continuum contains quasi-rationality, wh ich is thought of as an everyday cognitive process. Hammond (1996) maintains that human cognition constantly moves along the intuitiveanalytical continuum depending on the judgmen t task and the ecological cues present. Applicability of Social Judgment Th eory to the Healthcare Setting

PAGE 20

11 In this study, Social Judgment Theo ry was used to evaluate how nurse practitioners perceive and make judgments about patient’s CHD risk factors and the necessity for change in health promotion be havior. There is a pa ucity of scientific research on how nurse practitioners make deci sions. Social judgment theory can provide understanding of which risks nu rse practitioners perceive, se lect, assemble, and use in conjunction with their environment to reach a judgment about the level of risk a patient has for the development of coronary heart di sease. Although the a pplication of Social Judgment Theory in nursing research began as ea rly as the 1960’s, its use has been rather infrequent throughout the forty years since its original development. Social judgment theory and the concept “judgment analysis” have been used in prior studies to examine clinical nurses’ inference concerning states or physical conditions of patients in acute care facilities (Kelly, 1964; Thompson, Foster, Cole, & Dowding, 2005). This provided an avenue to further development of deci sion-making theory in the health promotion arena and nursing research. Brunswik’s Lens Model Brunswik created the Lens Model as a device to represen t how the various concepts involved in probabilistic func tionalism could be summarized. This model illustrates how one perceives a cue and combines the information with the environment to form a judgment. Figure 1 illustrates how a nurse practitioner perceives CHD risk factors, weighs the risk factors by importance, and collect this information and the stimuli occurring in the environment to arrive at a judgment about a patien t’s individual risk for the development of coronary heart disease. Basically, the left side of the model is the “judgment ecology,” or the true state of

PAGE 21

12 a patient’s actual CHD risk. The lines m oving right towards the cues represent the regression or actual weights represented by each cue. The cues consist of eight CHD risk factors: age, smoking, hyperten sion, total cholesterol, highdensity lipoprotein cholesterol (HDL-C) level, presence or absence of le ft ventricle hypertrophy (LVH), diabetes, and smoking. The right side of the model repres ents the “true ecology state,” or the nurse practitioners’ perception of a patient’s CHD risk. This may also be called the “judged state.” The lines on the right side of the model moving towards the cues represent how important the nurse practitioners viewed each cue. The difference between actual CHD risk (the left side of the m odel) and nurse practitioners perc eived CHD risk (the right side of the model) will indicate the agreement between nurse practitioner’s perceptions and judgments concerning patient’s CHD risk. Analyzing how well the nurse practitioners unite the left and right side of the models in their judgment he lps to predict their accuracy and consistency of perceptions for patients CHD risk.

PAGE 22

13 Figure 1 Lens Model (Cooksey, 1996) YeYsCUES ACTUAL FHS Estimated Risk of CHD in 10 yrs ARNP’sJUDGMENTS of Risk of CHD in 10 yrs es Yss Yee.724 Achievement .800 Agreement FHS & ARNP models .042 Residuals of FHS and ARNPsmodel .967 Predictability .931 Control Sex SBP Age Smoke Diab .186 .189 .224 .234 .136 .246 -.276 .350 .200 -.302 LVH Chol HDL .532 -.302 .172 .260 .094 .481 Relevance of Advanced Registered Nurse Practitioners as a Representative Cohort Advanced registered nurse practitioners account for thirty-t hree percent of all registered nurses in the United States (Ame rican Academy of Nurse Practitioners, 2002). They practice in many different environments such as emergency departments, pediatric units and clinics, critical and acute care f acilities, doctors’ offices and general practice clinics. Nurse practitioners can manage a wide spectrum of patient conditions ranging from acute to chronic while working in tandem with a medical doctor, Doctor of Osteopathy, or Dentist who serve as sponsor. Th eir educational background consists of an undergraduate degree in nursing and a master’s degree in a specialty area such as family, adult, pediatric, mental health, or acute care Nurse practitioners are at the forefront in regard to identifying potential a nd actual CHD risk factors for patients, which gives them

PAGE 23

14 the ability to greatly affect patient outcomes. Many recent studies have cited the increased satisfaction of patient s that have a consu ltation with a nurse practitioner (Meyer & Meirs, 2005; Aigner, Drew, & Phipps, 2004; Horrocks, Anderson, & Salisbury, 2002; Kinnersley, Anderson, Perry et al., 2000). Pati ents that had consu ltations with ARNPs have viewed advantages such as having a l onger consultation visit, which increased the amount of education and understanding of their disease process that in effect increases patient compliance with medica tions, diagnostic treatments and lifestyle modifications (Horrocks, Anderson, & Salisbury, 2002). Advanced registered nurse practitioners we re chosen as the sample for this study due to their nursing education and ability to respond with a high degree of accuracy to concise, scientifically worded questionnaires as demonstrat ed by the Nurses Health Study (1976). Advanced registered nurse practiti oners are a population of highly motivated participants that would likely complete all data request for the study. They play a major role in diagnosis, treatment and education of patients concerning th eir cardiovascular risk factors in clinics and acute care settings. This population of health care providers will provide insight as to how nurse practitioners weigh and subs equently treat different CHD risk factors and how they may cluster a combin ation of risk factors to make judgments on prevention and treatment for women and men. Nurses have been sampled for research in the past (Thompson, Foster, Cole, & Dowding, 2005; Beckstead, 2003; Holzworth & Wills; Kelly, 1964); however, a paucity of research has been conducted on how nurse pr actitioners perceive patient’s actual CHD risk factors. In addition, lit tle research has been conducte d analyzing how much weight they apply to CHD risk factors such as ag e, gender, smoking, chol esterol, hypertension,

PAGE 24

15 LVH, smoking, and diabetes, which may affect when and how nurse practitioners educate about health promotion behaviors. Furthermore, advanced registered nurse practitioners represent a large sample of health care providers from a va riety of ethnicities, and ages. They are one of the main sources of health education, and serve as role models of health for the patients that they work with everyday. Greater understandi ng of the relationship between nurse practitioners judgments of CHD risk and how it affects their trea tment and prevention strategies is warranted. Purpose of the Study The purpose of this study is twofold: (1 ) to describe how advanced registered nurse practitioners combine patient characteri stics (cues) when judging patient's risk of coronary heart disease and (2) to assess the effect that feedback has on improving advanced registered nurse practitio ner’s judgments of patient risk. Statement of Problem There is a paucity of literature conc erning how nurse practitioners analyze and weigh patient risk factors when judging a patien t’s risk for CHD. There is also a lack of literature explaining how nurse practitioners judgments of C HD risk factors compare to the actual Framingham Heart St udy’s estimated risks. A greater understanding of the knowledge and perceptions of nurse practitioners concer ning CHD risk factors is warranted in order to shed light on the accur acy of treatment strategies and educational goals for nurse practitioners and their patients. Specific Aims The specific aims of the study are as follows:

PAGE 25

16 1. To describe how advanced registered nurse practitioners combine patient characteristics (cues) when judging patient's risk of coronary heart disease. Research Questions for Aim 1: 1a How do advanced registered nurse prac titioners distribute importance weights among the various cues as they judge risk? 1b How accurate are their judgments as comp ared with actual Framingham Heart Study’s estimated risks? 1c How well does an additive linear model represent advanced registered nurse practitione rs’ judgment policies? 1d How much insight do advanced registered nurse practitioners have into their judgment processes? 2. To assess the effects that feedba ck has on improving advanced registered nurse practitioners ’ judgments of patient risk. Hypotheses for Aim 2: 2a Participants receiving feedback prior to completing a second policy capturing task will show increased agreement and achievement in their risk assessments as compared to particip ants who do not receive feedback prior to the second policy capturing task. 2b Participants receiving feedback prior to completing second policy-capturing task will show greater insight into their judgments of risk as compared to participants who do not receive feedback. Definition of Terms The following terms are defined and will be used throughout the study. The study

PAGE 26

17 definitions are derived partially from their us e in previous research and are established definitions in judgments and coronary heart disease research. Judgment Judgment is defined as exposure to environmental cues and drawing a cognitive decision to assess situations or circumstances perceptively and to draw sound conclusions (Cooksey, 1996). The Brunswik Lens Model and a multiple regression analysis will be used to measure judgment. Cognitive control Cognitive control refers to the similarity between an individual's judgment policy in a judgment task and the predictions of those responses made by a specific mathematical model, in this case a simple additive linear model. It is expressed as the correlation ( R) between judgments and predictions of those judgments by an individual's policy equation (Cooksey, 1996). Accuracy or agreement Accuracy refers to the degree of correspondence between an individual's responses to cue profiles and the ecological criterion (e.g., ac tual risk of CHD for the profiles according to th e AHA) (Cooksey, 1996). Achievement Achievement refers to the degree of corre lation between an individual's responses to the profiles and the ecological criterion (e .g., actual risk of CHD). It is expressed as the Pearson correlation coefficient (Cooksey, 1996). Insight Insight refers to the correspondence betw een the individual’s self reported cue

PAGE 27

18 importance and the importance weights derived via statistical analysis. Insight may be gauged by substituting self-reported weights in to the regression equation and comparing the predictions from this model to th ose made from the statistical model. Smoking Smoking is defined as the inhalation of tobacco products one or more times a calendar day, week, or month (AHA, 2006). Hypertension Hypertension is defined as maintain ing a systolic blood pressure above 140 mmHg and/or a diastolic blood pressure of 90 mmHg or above with or without current medicinal treatment (AHA, 2005). Healthy Lipid Profile A lipid profile that shows (1) a fasting to tal cholesterol level of 200 mg/dl or less, (2) high density lipoproteins (HDL) level of 50 mg/dl or higher, and (3) a low-density lipoprotein level of 100 mg/dl or less as m easured by a blood serum level (AHA, 2005). Advanced Registered Nurse Practitioner This participant must hold a current advanced register ed nurse practitioner state board of nursing certification, be licensed in the state of Florid a, and be in the specialty area of family or acute care. Significance for Nursing Given that many patients smoke tobacco, ha ve dangerously high lipid levels, and diabetes, and have treated and untreated hype rtension, it is important to study how these variables affect nurse practitioners decisi on-making and judgments towards prevention and treatment of CHD risk. This study w ill attempt to understand the decision-making

PAGE 28

19 and judgments by advanced registered nurse practitioners concerning patient’s risk for developing coronary heart dis ease. The hypothetical cues will reveal how much advanced registered nurse practitioners weigh the impor tance with reference to each of the eight coronary heart disease risk factors: age, gende r, systolic blood pressure, total cholesterol, high-density lipoprotein choles terol, left ventricular hype rtrophy, diabetes, and smoking. Also, advanced registered nurse practiti oners’ agreement and achievement in the estimation of CHD risk factors will be measured. The findings from this study will provide insight into how nurse practitioners judge, diagnosis, and treat the number one killer of Americans, co ronary heart disease. Chapter Summary Coronary heart disease is the leading cause of death for Americans (AHA, 2005). The objectives of this study were first to describe how nurse practitioners combine patient characteristics (cues) when judging a patient's risk of CHD. The second objective was to assess the effects that feedback ha s on improving nurse practitioners' judgments of patient risk. Determining how nurse practit ioners make judgments about patients CHD risk factors will provide the ability to customize continuing education modules and university curriculums directed towards spec ific identified gaps in CHD risk factor knowledge. The current focus in the literature is on recognition of CHD risk factors in women and men; this study will add to the scientific literature that seeks wa ys of identifying how nurse practitioners make judgments concer ning CHD risk. The findings will provide insight into how nurse practiti oners judge CHD risk factor severity and combine patient risk characteristics during the evaluation pro cess. In addition, the effects of feedback

PAGE 29

20 were examined to provide knowledge in areas of agreement and achievement concerning nurse practitioners judgments of CHD risk. Once it is determined that nurse practitioners weigh each CHD risk factor differently and also differ concerning their cognitive control and agreement in judgments, more appropriate education strategies may be used to make curricular changes, continuing education revisi ons, and individual counseling to increase awareness. This provides the opportunity for nurse practitioners to give patients more appropriate medicinal, diagnostic, and educat ion treatments and materials. It also provides an opportunity to increase primary and secondary preventive techniques in the hope to minimize the chances of a future cardi ac event. The next ch apter will present the theoretical framework and the literature supporting the study of nurse practitioners decision-making techniques.

PAGE 30

21 Chapter Two Review of Literature This chapter presents the theoretical framew orks for this study as well as a review of empirical literature pertin ent to advanced registered nurse practitioners’ judgments about men and women’s risk for the development of CHD. Introduction The understanding of how clinicians asse ss patient risk for disease and make decisions to refer patients to specialists when appropriate is important for optimizing professional training and practice and, for en suring that patients receive the highest quality of care possible. Several studies on c linical inferences made by nurses have been conducted (Kelly & Hammond, 1964). Some of the topics examined included the types of processes nurses utilized during pract ice (Hammond, Kelly, Schneider, & Vancini, 1966), information-seeking strategies nurses used when assessing the state of their patients (Hammond, Kelly, Schneider, &, Vanc ini, 1966), and how nurses revised their judgments of the patient when presente d with new information (Hammond, Kelly, Castellan, Schneider, & Vancini, 1967). Theoretical Framework Social Judgment Theory In 1975, Hammond, Stewart, Brehmer, a nd Steinmann proposed the Social Judgment Theory to examine how judgments and decisions are formed retrospectively.

PAGE 31

22 Researchers have applied this theory as a tool for conducting st udies evaluating how individuals make judgments a bout ecological (environmental) situations or probabilities of occurrences. It has been utilized in educational decision-making (Cooksey, 1988; Heald, 1991; Snow, 1968; Shulman & Elstei n, 1975), medical and health-related decision-making and ethics (Flynn, 1994; Slovic, Rorer, & Hoffman, 1971; Smith & Wigton, 1988; Wigton, 1988), accounting and au diting (Libby, 1981; Waller, 1988), risk judgments and social welfare (Cooksey, 1996). It has also been used successfully to study decision making across a wide array of clin ical settings includ ing diagnostic and treatment decisions among physicians (Fis ch, Hammond, Joyce & O’Reilly, 1981; Gillis, Lipkin, & Moran, 1981; Smith, Gilhooly, & Wa lker, 2003), physicians’ and patients’ compliance with treatment regimens (Rothert 1982), and nurses’ decisions to seclude and restrain psychiatric patients (Holzw orth & Wills, 1999). The Social Judgment Theory approach has also proven useful fo r assessing the effectiveness of physician (Wigton, Poses, Collins, & Cebul, 1990) and nurse (Thompson, Foster, Cole & Dowding, 2005) education programs. Despite the broad use of the Social Judg ment Theory a paucity of scientific research has been conducted concerni ng the decision-making process of nurse practitioners and their judgme nts about coronary heart disease risk factors. Social Judgment Theory was utilized in this study to provide a framework and understanding of nurse practitioner’s decision-ma king process and how they an alyze the cues to form a diagnosis of CHD risk. When studying how nurse practitioners judge the 10-year coronary heart disease risk for a patient, the patient typically will present with one or more risk factors or

PAGE 32

23 symptoms. The nurse practitioner must have a systematic way of collectively weighing cues derived from the patient and come to a decision or judgment about their appreciable risk. Figure 2 Brunswik Lens Model (Cooksey, 1996) Ye YsCUES ECOLOGY (CRITERION) SUBJECT'S JUDGMENTS Achievement X1 X2 X3 Xk. . Ecological Validities Cue Utilization Validities rij Brunswik Lens Model Social Judgment Theory was developed from Brunswick’s (1956) concepts of probabilistic functionalism a nd representative design. Hamm ond et al., (1975) provided the first complete description of how these concepts could be applied to the study of human judgment. The theory is concerned with the correspondence between a person’s judgments and the environment. These relati onships are illustrated using the Lens Model above. Balzer, Doherty and O’Conner (1989) pres ented a conceptual organization of the Lens Model. Briefly, Xk denotes the attributes of some multi-attribute obj ect of judgment that in the context of the judgm ent task are called cues. The Y refers to either criteria or

PAGE 33

24 responses. The subscripts e and s designate the environment and the subject, respectively. The hats (^) represent pred icted values, usually from least squares regression models of the respec tive “slides” of the lens. Cues are related to the criteria via some weight (wek) and correspondingly to th e subject’s responses (wsk). These weights may be correlation coefficients, or alternatively regression coefficients. The model allows for correlations (r edundancies) among the cues. The term ra is the correlation between the person’s judgments and the environment, and is referred to as achievement For example, in the present study achievement pertains to ARNP s’ precision in assessing the re lative degree of risk among a series of patient profiles as defined by the Framingham Heart St udy’s standards. Two multivariate correlations in the Lens Model, Re, the extent to which the criterion is predictable in the environment and Rs, the extent to which the person applies his or her judgment policy in a systema tic manner, given the name cognitive control are estimated from the regression models. The amount of knowledge that the person has about the relationships of cues to the environment is expressed using two other bi-variate correlations G and C referring to linear and nonlin ear relationships respectively. G is calculated as the correlation between the pr edicted values from each regression equation ( e and s) and C is the correlation between the residuals from each regression (Ye – e and Ys – s) The relationships among these variou s correlations in the Lens Model are summarized in the Lens Model Equati on developed by Hammond, Hursch, and Todd (1964) and simplified by Tucker (1964) as: ra = ReRsG + C[1 – R2 e)(1 – R2 s)]1/2. Hence a person’s precision in predicting the environment (ra) is a function of the extent to which the environment is predictable (Re) their knowledge of the environment

PAGE 34

25 (G and C) and the extent to which they sy stematically apply their knowledge (Rs) In order to understand and model the process of cue ut ilization, researchers have developed techniques called “ policy capturing.” Policy captur ing (PC) is the process of applying multiple regression methods to obtai n a representation of a judge's policy or judgment process. The typical policy-capturing task presen ts a judge (e.g., an ARNP) with a series of profiles (e.g., patients) to be judged on some relevant dimension (risk of CHD). The profiles are constructed of 8 ( k ) representative cues that can take on different values. Analysis proceeds on an individual basis where the fundamental data set for policy capturing comprises 70 ( m ) profiles of 8 ( k ) suitably quantified cue values and the set of 70 ( m ) judgments made by the judge. The relative importance (weights) of the cues is then determined using the standardized regression coefficients. In this study we will use a policy capturing approach to unders tand how the nurse practitioners form judgments of patient risk for CHD across 70 profiles constructed from eight cues. In policy capturing designs that include replicated profiles makes it possible to estimate cognitive consistency (Rc) or the extent to which th e judge performs similarly when responding to identical profiles on diffe rent occasions. Both cognitive control and cognitive consistency represent how orderly the individual is at making sense of the environment. Furthermore, in policy capturi ng designs where there is an environmental criterion, it is possible to estim ate achievement. These types of designs also allow for other indices of accuracy. Two such indices are precision the amount on average that a judge’s response differs from the criterion value, and elevation the amount by which the judge’s overall mean rating of risk is too high or too low when compared to the mean of the criterion values.

PAGE 35

26 Brunswik created the Lens Model as a device for representing how the various concepts involved in probabilistic func tionalism could be summarized. This model illustrates how one perceives a cue and combines the information with the environment to form a judgment. Figure 1 illustrates how a nurse practitioner may perceive coronary heart disease risk factors, weigh the ri sk factors by importance, and recall this information and the stimulus occurring in the environment to arrive at a judgment about a person’s individual risk for the development of CHD. Basically, the left side of the model is the “judgment ecology,” or the true state of actual CHD risk. The lines, moving right towards the cues, represent the regression or ac tual weights represented by each cue. The cues consist of eight coronary heart disease risk factors: age, gender, smoking, systolic blood pressure, total cholesterol, HDL-C level, diabetes, and L VH. The right side of the model represents the “true ecology state,” or the nurse practiti oner’s judgment of patients’ CHD risk. This may also be calle d the “judged state.” The lines on the right side of the model moving towards the cues will represents how important the nurse practitioners viewed each cue. The difference between actual coronary heart disease risk (the left side of the model) and a nurse practitione rs’ judged coronary heart disease risk (the right side of the model) will indicate how precise the nurse pr actitioners’ judgments are concerning patients’ CHD ri sk. Analyzing how well the nur se practitioners unite the left and right side of the m odels in their judgment helps to predict their accuracy and consistency of judgments for patient s’ coronary heart disease risk. Review of Literature Many studies have been conducted eval uating how physicians make diagnostic decisions concerning their patient s; however, there is a paucity of research concerning the

PAGE 36

27 decision-making process of nurse practitione rs. The following are studies conducted using the Social Judgment Theory to anal yze the decision making of physicians and nurses concerning patient dia gnosis and interventions. Nurses Decision-Making Throughout the last 40 years researchers have evaluated how nurses not only make decisions about the state of their patie nts but also the accur acy and consistency by which they make the decisions (Kell y, 1964; 1966; Holzworth & Willis, 1999; Watson, 1994; Thompson, Foster, Cole, and Dowding, 2005). Nurses’ judgments have been compared with other nurses employed in th e same setting and patient population to examine if nurses would make similar judgm ents or decisions about the same patient scenario. Interestingly, nurse s throughout the duration of 40 years of research have demonstrated inconsistency with their decisi on-making strategies and tend to demonstrate little agreement with their peers (Kelly, 1964; Holz worth & Willis, 1999; Watson, 1994; Thompson, Foster, Cole, and Dowding, 2005). Nurses are legally responsible for ev aluating signs and symptoms a patient presents in order to plan and implement a ppropriate nursing interven tions (Kelly, 1964). Watson (1994) inferred that nurs es’ decision-making skills need to be evaluated since the nurse is held responsible and accountable for their outcomes. No longer are nurses directly dependent on physicians to make ev ery decision during patient care. Nurses are expected to base their decisi ons on scientific evid ence and to demonstrate the ability to give their reasons for interventions when challenged (Watson, 1994). Examples of essential nursing decisions include recogniti on of symptoms leading to the patients’ declining state of condition, determ ining when a medicinal/ physician intervention is

PAGE 37

28 necessary, and appropriateness of seclusion or restraint of a psychiat ric or hallucinating patient for the safety of themselves and others (Kelly, 1964; 1966; Hammond 1964; Watson, 1994; Holzworth & Wills, 1999; Thompson, Foster, Cole, & Dowding, 2005). In 1999, Holzworth & Wills studied the deci sion making of nine registered nurses who were employed in a psychiatric care faci lity. Those nurses we re evaluated for the systematic process, accuracy, insight, and consistency regarding the need to closely observe, physically restrain, or seclude their short-term ps ychiatric patients. Social Judgment Theory was used as the primary theoretical framework for the study. Interestingly, nurses favored observation of pa tients over seclusion or restraint. Nurses generally had good insight into their own judgments’. However, individual differences in cue utilization and inconsistency in stra tegy usage led to disagreement among nurses about specific interven tions utilized. Furthermore, nurse s agreed with others’ judgments only one-third of the time (Holzworth & Wills, 1999). Similarly, Watson (1994) completed an e xploratory study to evaluate decisionmaking by nurses in a medical-su rgical hospital clinical area. Nurses were followed for two hours during one shift. A secondary objec tive of this study was to evaluate why some nurses make irrational decisions concerni ng patient care. The judges (nurses) were asked to make decisions on which dressing wa s most appropriate fo r a particular wound, the probability of the wound healing within th ree to four weeks, the current amount of patient comfort, and frequency with whic h the dressing should be changed. Watson concluded that 83% of the time nurses based their decisions on prior experience (p = 0.04) versus protocol. The nurse judges were not consistent the majority of the time with the exception of three judges. Very rarely was the protocol of the nursing unit a reason

PAGE 38

29 given by the nurses in thei r decision-making process. Social Judgment Theory has been used in studies that evaluate the educational and learning needs in different specializati ons of nursing. Thompson, Foster, Cole & Dowding, (2005) used the SJT to evaluate th e how nurses use clinical information when diagnosing hypovolemic shock in a series of simulated patient cases showing signs and symptoms of shock. The researchers’ ma in purpose was to examine how nurses combined evidence-based research knowledge with knowledge of available resources, clinical expertise and the patients’ preferences to make a decision. The participants were given a pre-test via computer then asked to sit in on a 30-minute lecture concerning signs and symptoms of hypovolemic shock (blood pr essure status, pulse, respiratory rate, oxygen saturation, and urine output). Next the st udents or nurses were asked to return to their computers and take the post-test measur ing how much data they learning during the lecture concerning the diagnosis of hypovolemic shock. The results indicated that there was little agreement among the nurses concerning their judgments of whether the patients were in shock or not. The authors found that there was a consiste nt 10% disagreement between the nurses regarding the patient’s he modynamic status. The study showed that clinicians use information in different ways to form their judgment policies. Physician Decision-Making Social Judgment Theory has been used to investigate a variety of physician practice patterns in medical settings (Wigt on, 1988; 1996; Engel et al., 1990). Examples include evaluating how physicians make deci sions concerning their prescriptive practices (Gillis, Lipkin, & Moran, 1981;Smith, Gil hooly, & Walker, 2003), diagnosing criteria (Wigton, Poses, Collins, & Cebul 1990), referr al practices (Rothert Roverner, Elstein,

PAGE 39

30 Holzman, Holmes, & Ravitch, 1984 ), treatment strategies, and frequency of ordering laboratory test (Holmes, Rover, Rothert, Schmitt, Given, & Ialongo, 1989), and attitudes about patient regimen compliance (Rothert, 1982 ). The use of research guided by the SJT has also proven useful for assessing the effectiveness of physician education programs (Wigton, Poses, Collins, & Cebul 1990). Investigators have determined that phys icians improve their diagnosing accuracy when receiving multiple scenarios and period ic education concerning their diagnostic decisions (Wigton, Poses, Collins, & Cebul 1990). Many SJT studies have found that there is little agreement betw een physician judges when comparing prescriptive practices, treatment strategies, and refe rral practices. The literature has indicated that physicians did agree concerning their e xpectations of patients rema ining compliance to their treatment regimen. This could lead to a decrease in patient compliance due to unconscious self-full-filling prophecy by th e physician (Rothert, 1982). Moreover, differences among physicians have demonstr ated an inconsistent use of symptom information and weighing of the cues (symptom s). Physicians’ expect ations of patients’ compliance with prescriptive regimens along with the amount of referral to specialist differed greatly (Gillis, Lipkin, & Mo ran, 1981;Smith, Gilhooly, & Walker, 2003; Rothert, Roverner, Elstein, Holzman, Holmes, & Ravitch, 1984). Based on past studies it has been determin ed that feedback to physicians about how they use information in making judgments can improve the quality of their judgments (Tape, Kripal, & Wigton, 1992). Tape, Kripal, & Wigton (1992) completed a study measuring different types of feedb ack and first year medical students’ successfulness with recognizing C HD risk factors in patients. The researchers’ had two

PAGE 40

31 treatment groups and two control groups. One treatment group consisted of probabilistic feedback, where the students received the co rrect probabilities of the patient developing CHD after each scenario was judged. Group two received cognitive feedback detailing the accuracy and consistency of their judgmen t after each scenario. The two control groups consisted of one group that received educ ation materials after the pre-test, before post-test; the second control group received no information before or after the preposttest. The student physicians had higher di scrimination ability fo llowing the probabilistic feedback intervention and improved achieveme nt with the cognitive feedback group. The control groups did not show signi ficant achievement scores afte r the post-test (Tape, et al, 1992). This outcome leads to the conclusion th at the participants receiving the correct probabilities of the patient developing CHD exhibited greater discrimination in recognizing which CHD risk fact ors were most heavily weighe d in patients versus the cognitive feedback group that received the information of how accurate and consistent they were with their judgments demonstrated a higher achievement score, which means a greater correlation with the actual CHD risk factors (ecology si de of the model). Tape, Heckerling, Ornato, & Wigton (1991) used the Lens Model to compare physicians’ likelihood estimates of pneumonia w ith the actual relations hips of patients’ clinical findings and their ra diographic diagnoses. Three sites were used for the study: Nebraska, Illinois, and Virginia. The st udy indicated that Nebraska and Virginia physicians were more accurate than the physicians in Illinois with regard to predicitions of pneumonia. Furthermore, the researchers found that the physicia ns’ in Nebraska and Virginia had strategies that were close to th e optimal strategies as calculated from the patient data at all sites. It was estimated that the differences in predictability were due to

PAGE 41

32 the differences in the study popul ations rather than differences in physicians’ ability. The Illinois site had a larger indigent patient population than th e Nebraska or Virginia sites. However, these types of differences have been observed in other regional variation studies (Lewis, 1969; Ep stein & McNeil (1985). Insight and Judgment Policies Insight refers to the correspondence betw een the individual’s self-reported cue importance and the importance weights derived via statistical analysis. Insight may be gauged by substituting self-reported weights in to the regression equation and comparing the predictions from this model to those ma de from the statistical model. Typically, subjects are asked to divide 100 points among the cues analyzed and these results are compared to the actual values given with in the scenarios (Rei lly & Doherty, 1989). Researchers have examined the amount of insi ght professionals have into their decisionmaking outcomes and found that i ndividuals typically lack in sight when evaluating selfreported cues (Slovic & Lichtenstein, 1971; Nisbett & Wilson, 1977; Schmitt & Levine, 1977; Anderson & Zalinskii, 1988; Reilly & Doherty, 1989; Reilly & Doherty, 1992). Reilly and Doherty (1989) completed a study evaluating how forty college students majoring in accounting made holistic judgm ents concerning 160 hypothetical scenarios about job offers. A similar study conducted by Roose and Doherty (1978) evaluated how 42 faculty members provided holistic judgments of what they believed to be fair employment salaries. The sample frame was thought to be a highly intelligent with the possibility of a highly insightfu l group of subjects to measure. In the previous study with the sample comprising of accounting students, the results indicated a great degree of insight when the judges were asked to physically identify their judgment out of a stack of

PAGE 42

33 different judgments on the same topic. However, when the subject’s data were statistically analyzed with direct comparis ons of the weights, l ittle insight was found among the accounting students. This allowed th e researchers to concluded that subjects were able to visually identif y their self-reported cue weights (p = .025); but when insight was expressed via a standard method of pr oducing weights the accountant students were found to have moderate self-insight at best (Reilly & Doherty, 1989). The same results occurred in the study evaluating PhD faculty’s insight into employm ent salaries (Roose & Doherty, 1978). Reilly and Doherty (1992) continued studi es on self-insight using a sample of female college students that were housed in a college dormitory to evaluate if the students had self-insight into their judgments concerni ng the desirability of a potential roommate. They also wanted to evaluate if the subject s who could identify thei r own captured policy and identify their own subjective weights. Th e results revealed that subjects were more likely to select their own policie s when there were twelve cues versus six (p > .05). Only a small proportion of the students’ statistical weights were highly correlated with their subjective weights (Reilly & Doherty, 1992). Many researchers studying self -insight agree that from the variability of policies clinicians are not operating under one set of diagnostic principles (Reilly & Doherty, (1989; Ullman & Doherty, 1984). Furthermore, it has been reported that there was a great amount of disagreement in the decision-making process of physicians, psychologists, and others in the diagnostic fields ((Reilly & Doherty, (1989; Ullman & Doherty, 1984; Bohn, 1984). The li terature indicates that subj ects in many of the fields studied have an insight of 50% (Reilly & Doherty, 1992).

PAGE 43

34 Cognitive Feedback and Judgment Cognitive feedback refers to the process of presenting the subject information about the relationship in the environment such as task information, cognitive information, and the relationship between the environmen t and the subject’s perceptions of the environment (Blazer, Doherty, & O’Conner, 1989). Cognitive feedback refers to information about relations instead of outco mes. Cognitive feedback has been found to improve the accuracy of judges in many circ umstances (Blazer, Doherty, & O’Conner, 1989). The cognitive feedback concept was derived from the framework of the lens model. Task information refers to the relation ship between the cues and the criterion (eight CHD risk factors and th e risk of development of CHD in the next 10 years), information about the criterion or the cu es themselves, or both. The cognitive information component greatly mimics task information, with the difference being the correspondence between the cues and the subject side of the model. Functional validity indexes include the achievement correlation (ra) the correlation between the predictions of the linear model between the environment and the linear model of the subject (G) and the correlation between the residuals from th e prediction of the environmental and subject models (C) (Blazer, Doherty, & O’Conner, 1989). E ssentially, feedback is the process by which information from the ecology side of the model is compared to the subjects’ judgment with the ecology and the results are revealed to the participant before taking another policy capturing tas k. This enables the indivi dual to understand the “gold standard” of information and impr ove their judgments if necessary. Balzer, Sulsky, Hammer, and Sumner (1992) completed a study evaluating

PAGE 44

35 whether different types of cognitive feedback lead to different levels of performance. Undergraduate college students (n = 133) were assigned to fi ve different groups: (1) task information only; (2) cognitive information only; (3) task and cognitive information; (4) task, cognitive, and functional validity inform ation; (5) and control group (no feedback). The subjects were asked to predict the number of baseball wins from a multiple cue probability learning task. The results i ndicated that subjects who received task information in any of the groups showed a si gnificantly better performance than subjects in the control group. The cognitive info rmation group performed no better than the control group. In 1994, Balzer, Hammer, Sumner, Birchenough, Martens, and Raymark performed a follow-up study to replicate the effect s of cognitive feedback. Again the investigators used undergraduate college stude nts as their subjects and found that task information was a necessary component to improve performance. The Importance of Studying ARNPs Advanced Registered Nurs e Practitioners account for thirty-three percent of all registered nurses in the United States (Ame rican Academy of Nurse Practitioners). They practice in many different envi ronments such as emergency departments, pediatric units and clinics, critical and acute care facilities, doctors’ offices, and genera l practice clinics. Advanced registered nurse practitioners work concurrently w ith a medical doctor (MD), Doctor of Osteopathy (DO), or De ntist (DDS) who serve as sponsor. Their educational background consists of an underg raduate degree in nursing and a master’s degree in a specialty area such as family, adu lt, pediatric, mental health, or acute care.

PAGE 45

36 Nurse practitioners are at the forefront in re gard to identifying potential and actual CHD risk factors for patients, which gives them the ability to greatly affect patient outcomes. Many recent studies have cited the increas ed satisfaction of pa tients that have a consultation with an ARNP (Aigner, Drew, Phipps, 2004; Meyers & Meir, 2005; Horrocks, Anderson, Salisbury, 2002; Kinnersley, Anderson Parry, Clement, Archard, et al., 2000). Patients that had consultations w ith ARNPs cited perceived advantages as having a longer consultation visit which increased the amount of education and understanding of their disease process that in turn increases pa tient compliance with medications, diagnostic treatments and lifes tyle modifications (Horrocks, Anderson, Salisbury, 2002). In the United States today many pr imary care physicians employ nurse practitioners to share in the car e of their patients (LeClaire, 2005). In 2004 there were an estimated 141,209 licensed ARNPs (U.S. Depart ment of Health and Human Services, 2005). According to the American Academ y of Nurse Practitioners (2005), 66% of ARNPs work in at least one primary care site Research on the increasing role of nurse practitioners in the health care industry ha s demonstrated that patient care under the MD/ARNP model is comparable to the MD alone model (Ai gner, Drew & Phipps, 2004; Horrocks, Anderson & Salisbury, 2002; Kinne rsley, Anderson Parry, Clement, Archard, et al., 2000). Patients report greater satisfaction with the healthcare they receive under the MD/ARNP model (Horrocks et al. 2002; Kinnersley, et al., 2000) Comparing nurse practitioners to MDs, Kinnersley, et al. (2000) found that nurse practitioners made referrals to care specialists at rates greater than or equal to general practice MDs. Horrocks, et al, (2002) report that ARNPs spe nd more time with patients per clinic visit

PAGE 46

37 than do physicians. Given their role in prim ary care, nurse practitioners represent an important resource in early detec tion of numerous diseases. Understanding how nurse practitioners use patient symptoms or risk factors to form judgments of patient risk for diseas e and how such risk assessments influence referral decisions may improve general diseas e prevention efforts. Below is a figure detailing a logic model of th is study; it begins by indica ting the population of adult or family nurse practitioners with a varied amount in years of experien ce and that they will exhibit increased critical thinking skills wh en making assessments concerning patient risk factors by distributing weight among risk factors. Next a judgment will be made concerning a patient’s risk of having a cardiac ev ent in the next 10 years. This decisionmaking process may be affected if they receive feedback about their performance concerning assigning importance weight to the eight CHD ri sk factors. Information concerning the educational need for nurse prac titioners will also be evaluated depending on the results obtained. Input Intervention Output Figure 3 Logic Model -ARNPs with adult or family track education -Years of experience as an ARNP -Increased critical thinking of judgments -Clinical Practice (adult or family) -Distribution of weight among cues -Assessment & judgment of patient’s risk for heart disease -Cognitive feedback -Increased accuracy & consistency of risk assessment by ARNPs -Knowledge of need for curricular or continuing education improvements

PAGE 47

38 The Importance of the Framingham Heart Study The original Framingham Heart Stu dy (FHS) was conducted during 1948-1971 to measure Americans risk for the development of heart disease. In 1971, a new phase of the study began called the Framingham Offspr ing Study. This study has evaluated the development of CHD in the offspring of the or iginal cohort of participants. In 2002, the study entered a third new phase by enrolling th e grandchildren of the original cohort, which has allowed researchers to have access to three generations of individuals to determine not only environmental or exposures that contribute to h eart disease risk but genetic components as well. Many studies were conducted to eval uate the type and amount of risk factors were causal for C HD. The FHS provided gender specific CHD prediction functions for assessing risk of developing CHD in a ethnically diverse middleclass population. Many prospective studies ha ve evolved using the Framingham data and were reviewed in the above literature: th e Atherosclerosis Risk in Communities Study (1987-1988); Physician’s Health Study ( 1982); Honolulu Heart Program (1980-1982); Puerto Rico Heart Health Program ( 1965-1968); Strong Heart Study (1989-1991); and Cardiovascular Health Study (1989-1990). Summary Coronary heart disease is the single largest killer of American males and females (AHA, 2005). According to the AHA approximately 41% of Americans that experience a coronary attack in a given year will die from it (AHA, 2005). To combat this growing problem, strategies need to be evaluated to assess how the identif ication of actual and potential CHD risk decisions is conducte d. Many primary care physicians employ nurse practitioners making them one of the first lines of defense towards primary prevention of

PAGE 48

39 CHD (American Academy of Nurse Practioners 2002). They are at the forefront of assessment, detection, and treatment of poten tial and actual CHD risk factors for their primary care patients. In the early 1980’s studies using Social Judgment Theory were developed with a focus on understanding the healthcare provider’s decision-making strategies. At that time, the role of the nurse practitioner was at its early stages rendering unavailable sample sizes to study. Presently, the role has greatly developed and is so widely used that nurse practitioners ar e now considered primary care providers. The current focus in the literature is on recognition of CHD risk factors in women and men; this study will add to the scientific literature that seeks ways of identifying educational needs of nurse practitioners. Th e findings provide insight into how nurse practitioners judge CHD risk factor severity and combine patient risk characteristics during the evaluation process. In additi on, the effects of cognitive feedback were examined to provide knowledge in areas of accuracy and achievement concerning nurse practitioners judgments of CHD risk. If it is determined that nurse practitioners weigh each CHD risk factor differently and also differ in terms of their accuracy and consistency in judgments, more appropriate e ducation strategies may be used to make curricular changes, continuing education revisi ons, and individual counseling to increase awareness. This will provide an opportunity fo r nurse practitioners to give patients more appropriate medicinal, diagnosti c, and education treatments a nd materials. It also will provide an opportunity to increase primar y and secondary preventive techniques in the hope of minimizing the chances of a future cardiac event. Foundation for Future Research

PAGE 49

40 Future research concerning how healthcare providers make decisions regarding preventive patient care can be explored. Judge s’ strategies and agreement levels can be determined and presented to the clinician to evaluate how thei r weights or decision processes compare with those of their colle agues and why the differ. Such comparisons can, in turn, serve as the foundation for disc ussing diagnostic stra tegies and reducing inter-judge discrepancies. This technique will also allow a way of evaluating the success of educational tools for clinicians by re testing using the lens model and cognitive feedback method. Findings of this study will also be useful for informing policy and curricular decisions. The next chapter will disc uss the study sample, desi gn, research questions, and hypothesis for the evaluation and analysis of nurse practitioners decision-making techniques concerning cor onary heart disease risk.

PAGE 50

41 Chapter Three This chapter describes the methods that we re used to explore advanced registered nurse practitioner judgments of coronary heart disease risk factors in patients. The sample selection, data collection, and instrume ntation are described. This is followed by a discussion of the research procedures and data analysis methods used for the study. Research Design This study used a three-group pretest-postte st experimental design. Participants were randomly assigned to one of three groups : (1) comparative feedback condition; (2) Framingham Heart Study CHD risk predicti on worksheet condition; and (3) control group. The comparative feedback conditi on participants were given information concerning how Framingham wei ghted the eight CHD risk f actors compared to how the subjects assigned weight to each CHD risk factors. The Framingham Heart Study CHD risk prediction worksheet feedb ack consisted of the actual re gression weights of the eight CHD risk factors according to the upda ted Framingham Heart Study. For each policycapturing (PC) experimental group the interven tion was given after the completion of the first PC task. (The PC task, comparativ e feedback, and Framingham worksheet are described in forthcoming text). The control group simply completed the PC task twice, one week apart. Participants in both experimental gro ups completed the same policy-capturing (PC) task twice, separated by approximately one week; the difference between the two

PAGE 51

42 conditions centered on whether having this information made a difference and if so did the comparative feedback or Framingham worksheet group enhance their agreement, achievement and insight more than the contro l group. This design al lowed the researcher to describe ARNPs' risk judgment processes as stated in Aim 1, and to assess the effects that comparative feedback had on improvi ng judgment agreement, achievement, and insight as stated in Aim 2. Sample Sixty family or adult specialty nurse pract itioners affiliated with the USF College of Nursing, representing five counties located within the sout hwest region of the Florida were recruited to participate in the study. Policy capturing is an ide ographic technique that requires many judgments from one participant. More crucial than the numbe r of participants is the number of judgments made by each participant. To test the hypot hesis that exposure to different forms of information regarding clinical assessment of CHD risk would a lter judgments and cue weights power was set at .80 with an alpha of .05. Twenty participants per condition were considered adequate. Materials Policy Capturing Task The materials for use in this study were presented to partic ipants in a short booklet. Booklets contained a cover page describing th e purpose of the study and instructions for the judgment task, a section asking participants to indicate how they assigned importance to the cues during the judg ment task and whether or not they would refer each patient to a cardiologist, a nd a section requesting basic demographic

PAGE 52

43 information. (Pilot testing i ndicated that the booklet w ould take between 20 and 60 minutes to complete). To capture the judgment policy used by each participant to assess risk of CHD, a series of 70 unique patient profiles were cons tructed by the primary investigator using the eight variables identified by the Framingham Heart Study: gender, age, smoking status, total cholesterol level, HDL level, systolic blood pressure, the diagnosis of diabetes and a diagnosis of left ventricular hypertrophy. Thes e variables (cues) were selected because they form the basis of the “gold standard ” to assess patient risk for CHD (Anderson, Wilson, Odell & Kannel, 1991). Anderson et al. used data from 5,573 patients that were followed over a 12-year period to construc t a complex nonlinear model for estimating risk from several known and suspected risk fa ctors. Anderson et al. then used these equations to produce prediction rules or “works heets” for use by practit ioners in clinical settings. The worksheets provided clinicians with algorithms or prediction rules for estimating patient risk of CHD using a simple tallying method to estimate patient risk. Male and female patients are assigned various points based on their age and presence of other risk factors (smoking, left ventricular heart failu re, diabetes status, cholesterol level, etc). Worksheets formed one of our experimental interventions. Seventy descriptive profiles made up of ei ght cues (Figure 3) were given to each participant. Given the number of profiles each nurse practitioner had to review this brief format seemed appropriate to mini mize respondent fatigue and boredom.

PAGE 53

44 Figure 4 Example of Patient Profile Patient Profile #1 Age: 42 Gender: Female Total Cholesterol: 188 Smoker: No HDL: 45 Systolic Blood Pressure: 110 Diabetes: Yes Left Ventricular Hypertrophy: No On a scale form 0% to 100%, how would you ra te the likelihood that this patient will have a heart attack within the next 10 years? ______ Would you refer this patient to a cardiologist? YES or NO Self-report of Cue Importance Following completion of the policy-capturing task, participants were asked to specify how much importance they placed on each cue type during the task by assigning 100 points among the eight cues (age, gender, sy stolic blood pressure, total cholesterol, high-density lipoprotein, smoking status, diab etes, and left ventri cular hypertrophy). Essentially the participant stated how much we ight or values they assigned to each cue. Values had to total to 100 points. These da ta were compared to the subjective weights resulting from the policy-capturing task to determine the degree of insight each participant had into his or her judgment process.

PAGE 54

45 Demographic Information Basic information was collected on particip ants that included their age, gender, years of practice as an ARNP, and area of sp ecialization. This information was used to evaluate regular patterns in cu e utilization, judgmen t agreement, and participant's insight into their judgment pr ocesses as functions of these variables. Institutional Review Board The study proposal was reviewed and approved by the University of South Florida Institutional Review Board. The r ecruitment material consisted of a written announcement that was sent via electronic ma il to all USF affiliated ARNP preceptors and their colleagues. The announcements contained the name and contact information of the primary investigator and encouraged potenti al participants to call with questions or comments. The participants did not write their names on any materials except the informed consent. Procedure Session 1 The investigator contacted participants and scheduled appointments to discuss the study. After explaining the st udy and obtaining informed consent, participants were presented with the judgment booklet. The partic ipants were asked not to place their name on the booklet to maintain confidently. Partic ipants were randomly assigned to the comparative feedback, Framingham worksheet, or control group; th e random assignment was determined by a computer randomizati on equaling 20 participants in each group. Participants in the comparat ive feedback group completed th e policy capturing (PC) task and then were given the comparative feedbac k. They were seen again in one week to

PAGE 55

46 complete the PC task a second time. Part icipants in the Framingham worksheet group were provided the Framingham Heart Study Ri sk Prediction worksheet after completion of the first PC task and were seen again in one week to complete the PC task. Participants in the control group simply comple ted the PC task twice, one week apart. All participants were debriefed and thanked for th eir participation, and their name was added to a raffle to win a gift cert ificate to a local restaurant. Figure 5 Example of Comparative Feedback How important was each cue as you formed your estimates of CHD risk? Divide 100 points among the cues below. Assign the most points to the cue(s) you relied on the most. ____ Gender ____ Age ____ Systolic Blood Pressure ____ Left Ventricular Hypertrophy ____ Total Cholesterol Level ____ High Density Lipoprotein Level ____ Smoker ____ Diabetes ____ TOTAL Actual importance of each cue 8.6 Gender 22.2 Age 10.4 Systolic Blood Pressure 20.1 Left Ventricular Hypertrophy 6.1 Total Cholesterol Level 11.4 High Density Lipoprotein Level 9.5 Smoker 11.6 Diabetes 100 TOTAL

PAGE 56

47 The investigator was available to answ er any questions. After the participant completed the booklet an appointment was sc heduled approximately one week later. Session 2 Approximately one week following session 1, the investigator met with each participant to repeat the PC task. To mi nimize attrition of the study participants the primary investigator traveled to their place of employment, designate d work site, or other location the participant chose for completion of the informed consent and questionnaires. Analysis Plan Aim 1: Describing Judgment Policies Social judgment theory provided the framework for conducting analyses and constructing feedback. Each par ticipant's judgment of risk wa s analyzed separately using a SPSS regression procedure. Sta ndardized regression coeffici ents were interpreted as estimates of importance weights. To obtai n the actual CHD risk weights, the updated Framingham Heart Study CHD risk prediction equation for the 70 prof iles were regressed onto the eight cues and transformed onto the 100-point scale described above. Research Questions 1a How do ARNPs distribute importan ce weights among the various cues as they judge risk? Examining the semi-partial correlation coefficients obtained from each participant and comparing the group mean s was used to answer this question. 1b – What is the agreement between ARNPs' judgments as compared with actual Framingham estimated risks? This question was answered by expressing agreement as the degree of discrepancy between judged risk an d actual risk. 1c How well did a linear model represen t ARNPs' judgment policies? To answer

PAGE 57

48 this question the degree of cognitive control exhibited by each participant was assessed. The multiple correlation Rs was inspected from each policy equation. 1d How much insight did ARNPs have into their judgment processes? This question was addressed by comparing the s ubjective cue weights obtained from each participant's policy equation with his or he r self-reported cue im portance weights. The degree of consistency between these two sets of weights was expressed by using a skill score. The closer the value was to one, th e greater the degree of insight. These values were also aggregated to provi de a single index for the sample of ARNPs examined here. Aim 2: Assessing the Effects of Cogniti ve Feedback on ARNPs' Judgment of Risk The three-group pretest-posttest design pe rmitted testing the hypotheses regarding the influence of the comparative feedback provided. An analysis of variance (ANOVA) was used to determine if there were significan t differences from pre-to-posttest across the three conditions. Hypothesis testing used an alpha value of .05 to evaluate differences among the participants in each experimental manipulation group compared to the control group. Again, the hypothesis of this study stat ed that there would be a significant difference between the two intervention groups as compared to the control group. Hypothesis 2a Participants assigned to a ma nipulation group (comparative feedback or Framingham worksheet) prior to completing the second PC task would show increased agreement and achievement in their risk assessments compared to participants who did not receive feedback prior to the second PC ta sk. An ANOVA was conducted to compare groups on the agreement of each par ticipant's judgment polic y relative to the Framingham-estimated actual risk. A second ANOVA was conducted to compar e group differences on participants'

PAGE 58

49 achievement index. This index was calculated as the Pearson correlation coefficient, but was transformed using Fisher's r to Z adjustment prior to analysis. These achievement indices from the first PC task were treated as a covariate and the indices from the second PC task served as the dependent variable. Hypothesis 2b Participants that were assigned to a manipulation group (comparative feedback or Framingham worksh eet) prior to completing the second policycapturing task will show greater insight into their judgments of risk as compared to participants who do not r eceive feedback. An ANOVA wa s conducted to compare group differences between each participant's subjectiv e cue weights and his or her self-reported cue importance weights. The next chapter will discuss the results of the study concerning how accurate nurse practitioners were in their importance weights of each cue, the amount of cognitive control exhibited, the amount of insight exhibited by NPs, how well they agreed with the Framingham Heart Study risk prediction equa tion, and how well the NPs judgment model agreed with the Framingham Hear t Study’s risk prediction model.

PAGE 59

50 Chapter Four Results Chapter four includes the analyzed data to address the research questions and hypothesis described in chapter three. Th e following sections include a demographic description of the sample, evidence detai ling how well an additive linear model captured nurse practitioners’ judgment policies, in formation on the agreement between nurse practitioners’ judgments compared to the Framingham estimated CHD risks (ecology), the distribution of nurse practitioners’ impor tance weights among the eight cues, and the participants’ insight into their judgment processes. Sample of Judges A total of 99 nurse practitioners were contacted for recruitment, sixty nurse practitioners agreed to participate and comp leted the study; 39 refused participation or did not respond to recruitment calls. The fi nal sample consisted of 58 females and two males. Nurse practitioner speci alties included two acute criti cal care, 18 adult, 33 family, three geriatrics, and three oncol ogy nurse practitioners. The m ean age of the subjects was 49 years (SD = 6.65). The range of age was 33 to 65 years (Table 1). Years of work experience ranged from one to twenty-nine year s, with the largest percentage of years worked being six years (16.7%). The second la rgest percentage of years worked for this sample was eight years (11.7%). The mean years worked were 8.44 years, (SD = 6.12). The majority of respondents held a nurse practitioner certifica tion in the following

PAGE 60

51 categories: family (53%), adult (30%), geri atrics (5%), oncology (5%), acute critical care (3.3%), and women’s health (1.7%). Performance of All Judges at Time 1 A multiple linear regression analysis on the 70 cases was conducted to estimate the predictability of the judgment task. The Re was .96, confirming that the Framingham CHD risk prediction equations (Anderson’s et al. 1991) could be adequately represented by an additive linear regression model. A relatively high value of Re indicates that if nurse practitioners used an addi tive linear model to form judg ments of patient risk and if they assigned cue weights proportional to t hose used in the regression model, then it would be possible for them to obt ain high degrees of achievement (ra) and agreement (G) on the judgment task. The ra (achievement) index reveals the amount of agreement between the judges and the Framingham CHD risk prediction equation. Achievement ranged from .44 to .87 (average ra = .70). The Rs index measures the amount of cognitive control a participan t had concerning their judgments and how well an additive linear model captures the judgment policies of the subject. The participants showed Rs values that ranged from .72 to .95 (average Rs = .88) indicating an additive linear model adequately captured the pract itioners judgment policies. The G index indicates the amount of agreem ent between the a dditive linear model of the Framingham CHD risk prediction equation and the ad ditive linear model of the judge. The G index ranged from .57 to .97 (average G = .85).

PAGE 61

52 Table 1 Sample Composition ComparativeFeedback Worksheet Feedback Control Total Characteristics (%) n (%) n (%) n (%) n Gender Male (0) 0 (0) 0 (10) 2 (3.3) 2 Female (100) 20 (100) 20 (90) 18 (96.6) 58 NP Specialty Acute Critical Care (5) 1 (5) 1 (0) 0 (3) 2 Adult (35) 7 (25) 5 (30) 6 (32) 19 Family Practice (45) 9 (60) 12 (55) 11 (53) 32 Geriatrics (10) 2 (5) 1 (0) 0 (5) 3 Oncology (5) 1 (0) 0 (10) 2 (5) 3 OBGYN (0) 0 (5) 1 (0) 0 (2) 1 Years Worked 1-5 years (30) 6 (30) 6 (35) 7 (32) 19 6-10 years (30) 6 (40) 8 (40) 8 (42) 25 11-15 years (20) 4 (5) 1 (20) 4 (15) 9 16-20 years (10) 2 (10) 2 (5) 1 (8) 5 21-26 years (0) 0 (5) 1 (0) 0 (2) 1 27-29 years (5) 1 (0) 0 (0) 0 (2) 1

PAGE 62

53 The C index indicates the stre ngth of correlation among re siduals from judgment models and the additive linear model of th e Framingham CHD risk prediction equation. A large value suggests that some aspects of judges’ use of cu es were not captured by the additive linear model. Results for participan ts revealed a range from -.38 to .55 (average C = .01). This suggests that some clinicia ns may be using more complex judgment policies than an addi tive linear model. Another useful index of agreement is Mu rphy’s (1988) skill score. A skill score provides insight into properties of the environmental/information system and the cognitive system within nurse practitioners judgments; it shows how those properties interact to influence judgment skill. Th e two components of sk ill score are conditional and unconditional bias. Conditional bias refers to the standard de viation of the judgments compared to the standard deviation of the Framingham ri sks (Murphy, 1988). For example: if a practitioner is sensitive to extreme cue values (such as a systolic blood pressure of 200 mmHg) her judgments of risk will be more varied than those predicted by the Framingham risk equations. This type of bias can cause an error in prediction if the variable being judged does not have a strong relationship with risk. The second type of bias is unconditional bias, which reflects any mis-match between the mean of the judgments and the me an of the Framingham risks. For example, a nurse practitioner’s ranking of patient risk may match those from Framingham but she overestimated risk for all patients by 10. T hus her skill score would show unconditional bias. Insight was measured by asking participants to subjectively weigh each CHD risk

PAGE 63

54 factor used in the lens model compared to the Rs value = .88 at Time 1. For every subject, the R = .37 for the subjective e quation was significantly lower than the Rs, which suggest that the nurse practitioners’ insight was not modest. The mean skill scores at Time 1 were 34.8 (SD = 15.1) and the mean for the Framingham equation was 14.4 (SD = 12.6). To determine whether unconditional bias or conditional bias was the bigger source of error the within subject average ra ting and within subject standard deviation were correlated with skill scores. (The correlation of skill scores and average ratings was .99 the correlation of skill score and standa rd deviation = -.62). These correlations suggest that elevation (ove restimation of risk) was th e bigger source of error. Hypothesis Testing It was hypothesized that nurse practitioners’ agreement and achievement would increase if they were exposed to one of the experimental interventions as compared to the control group. This hypothesis was tested using a 3 x 2 (condition by time) ANOVA. The key test of significance was the interaction that addr essed the general question, Was the change in the dependent variable the same for all three conditions? Also of interest was the main effect of change over conditions. This test addressed the question: Did the dependent variable change over time for the en tire sample? It was also hypothesized that nurse practitioners exposed to the experimental interventions would have greater insight into their judgment policy for CHD risk in the primary care patient versus the control group. Changes in ra, Rs, and G The agreement between the nurse pract itioners’ assessment of risk and the Framingham CHD risk prediction equati on may be expressed as achievement (ra) the

PAGE 64

55 correlation between a judge’s ratings and the criterion values. Achievement for Time 1 ranged from .44 to .87 (average ra = .71). Achievement for Time 2 ranged from .21 to .96 (average ra = .75) (Figure 5). The main eff ect for time was F(1, 57) = 4.22, p < .05, indicating that for the entire sample, ag reement improved. There was no significant interaction indicating that the intervention groups did not differ significantly than the control group. 0.68 0.69 0.70 0.71 0.72 0.73 0.74 0.75 0.76 0.77 0.78 TIME1TIME2Ra C.Wts Wk Sht Control Figure 6 Amount of agreement among ARNPs judgments & FHS Risk Prediction Equation C.Wts = Comparative Feedback; Wk Sht = Framingham Heart Study CHD Risk Prediction Worksheet; Control = Control Group The average Rs value (index of cognitive control) across judges was .88 and ranged from .72 to .95, suggesti ng that the judgment policies were adequately captured by the additive linear model. Figure 6 indi cates that the main effect for time was significant F(1, 57) = .75, p > .05; however no significant interact ion among conditions was present.

PAGE 65

56 0.86 0.87 0.87 0.88 0.88 0.89 0.89 0.90 0.90 TIME1TIME2R C.Wts Wk Sht Control Figure 7 Cognitive Control among ARNP Judges The degree of agreement be tween the additive linear model of the Framingham equations and the nurse pract itioners’ judgment model, ( G index), averaged .85 and ranged from .50 to .99 suggesting that the two a dditive linear models agree. As indicated in Figure 7, there was a significant main effect for time F(1, 57) = 9.29, p < .05 and interaction by condition F(2, 57) = 3.30, p < .05. Follow up tests re vealed that both intervention groups showed significant increases in G [F(1,57) = 5.37, p < .05] for both the comparative weight condition and F(1, 57) = 10.45, p < .05 the Framingham worksheet condition. The control gr oup did not change significantly.

PAGE 66

57 0.82 0.83 0.84 0.85 0.86 0.87 0.88 0.89 0.90 0.91 0.92 TIME1TIME2G C.Wts Wk Sht Control Figure 8 Agreement among the Judges' Model and the Simple Additive Linear Model of the Framingham Equation Comparing the components of the skill scores from Times 1 and 2 showed how nurse practitioners’ agreement with Fram ingham equations improved overtime. The main source of disagreement with th e Framingham equations at Time 1 was unconditional or elevation bias. On the averag e, nurse practitioners overestimated risk. Figure 8 shows that the entire sample showed less unconditional bias at Time 2 (F(2, 57) = 9.85, p < .01) and that the reduction fo r the Framingham worksheet condition was greater than the other tw o conditions (F(1, 57) = 3,43, p < .05). Figure 8 shows unconditional bias (elevation) or the mean estimate of risk. Figure 9 indicates the standard deviation for risk estimates or condi tional bias (scatter). There was a significant main effect F(1, 57) = 9.85, p < .05 a nd interaction F(2, 57) = 3.43, p < .05 for unconditional bias (elevation). There also was a significant main effect F(1, 57) = 36.22, p < .01 and interaction F(2, 57) = 5.42), p < .05 for conditional bias (scatter).

PAGE 67

58 Figure 9 Unconditional bias (elevation) from Time 1 to Time 2 20 22 24 26 28 30 32 34 36 38 40 42 44 TIME1TIME2Average Est of Ris k Wts Wk Sht Control Figure 10 Conditional bias (s catter) from Time 1 to Time 2 10 11 12 13 14 15 16 17 18 19 20 TIME1TIME2St Dev of Risk Ests Wts Wk Sht Control Distribution of Importance Weights Among the Eight Cues The nurse practitioners varied in their distribution of cue importance weights as

PAGE 68

59 they estimated patient risk. The cue we ights were estimated from semi-partial correlations ( rsp). Generally, sample-based correla tions are not normally distributed except when the population parameter value th ey are estimating is zero (Cooksey, 1996). Therefore the appropriate correction for non-nor mality of correlation coeffecients is the Fisher r to zr transformation (Cooksey, 1996). To assess the systematic differences in the eight cue we ights among judges, a multi-level regression analysis was performed. At level one, risk assessments were regressed onto the eight cues. Demographic characteristics of the judges (gender, age, years of work experience, and area of speci alization) along with their ratings of the realism of the patient profiles were treated as level two variables. The multi-level analysis was used to address whether dem ographic differences among judges could be explained by the observed variation in each cue’s weight. Each cue was individually evaluated a nd compared from Time 1 to Time 2 for statistically significant improvement in thei r judgment of risk as compared to the Framingham CHD risk prediction equation. Ea ch analysis was also compared by group (comparative feedback, Framingham worksheet, or control). The results are as follows: As indicated in Figure 10 subjects did not show a significant main effect F(1, 57) = 3.41, p > .05 or interaction in how they weighed the cue gender.

PAGE 69

60 0.00 0.01 0.02 0.03 0.04 0.05 0.06 0.07 0.08 TIME1TIME2Gender Weight C. Wts Wk Sht Control Figure 11 Judges assign ed weight to Gender from Time 1 to Time 2 In Figure 11, the experiment al manipulation groups had a significant main effect F(1, 57) = 17.20, p < .001 and interaction F(2, 57) = 3.59, p < .05 for age. The participants significantly increased the weight they assigned to the cue, age, from Time 1 to Time 2 respectively. The control group did not significantly change their assigned weight for the cue age from Time 1 to Time 2.

PAGE 70

61 0.10 0.12 0.14 0.16 0.18 0.20 0.22 0.24 0.26 0.28 0.30 TIME1TIME2Age Cue Weight C.Wts Wk Sht Control Figure 12 Judges assigned weight for AGE from Time 1 to Time 2 For the systolic blood pressure cue ther e was a significant main effect F(1,57) = 12.57, p < .05, but no interaction among the groups was identified as shown in Figure 12.

PAGE 71

62 Figure 13 Judges Assigned We ights for Systolic Blood P ressure from Time 1 to Time 2 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 TIME1TIME2SBP Cue Weight C.Wts Wk Sht Control There was not a significant interaction between groups fo r the cue LVH as shown in Figure 13, however the main effect was significant F(1, 57) = 5.54, p < .05. 0.19 0.20 0.21 0.22 0.23 0.24 0.25 0.26 0.27 0.28 0.29 TIME1TIME2LVH Cue Weight C.Wts Wk Sht Control Figure 14 Judges assigned weight s for LVH from Time 1 to Time 2

PAGE 72

63 The cholesterol cue did not show a significan t main effect F(1, 57) = 2.62, p > .05 or interaction F(2, 57) = 3.12, p > .05 among groups as indicated in Figure 14. The sample mean for Time 1 was 0.15 and Time 2 was 0.10 with a difference of 0.13; the effect size was moderate at .10. 0.10 0.11 0.12 0.13 0.14 0.15 0.16 0.17 0.18 0.19 0.20 TIME1TIME2Cholesterol Cue Weight C.Wts Wk Sht Control Figure 15 Judges assigned weight for cholesterol from Time 1 to Time 2 The HDL cue results indicated there was not a significant main effect F(1, 57) = .38, p > .05 or interaction (see Figure 15). No significant main effect F(1, 57) = 1.55, p > .05 or interaction was noted for the cue smoki ng status as indicate d in Figure 16. The results for the last cue of diabetes status indicated a significant main effect F(1, 57) = 5.98, p < .05 and interaction F(2, 57) = 6.38, p < .01 among groups as displayed in Figure 17. The participants exposed to the Fram ingham worksheet mani pulation significantly increased the weight they assigned to th e cue diabetes F(1, 57) = 15.18, p < .01; the comparative feedback and control groups di d not significantly change their weight

PAGE 73

64 assignment from Time 1 to Time 2 respect ively F(1, 57) = 2.21, p > .05; F(1, 57) = 1.31, p > .05. -0.11 -0.10 -0.09 -0.08 -0.07 -0.06 -0.05 TIME1TIME2HDL Cue Weight C.Wts Wk Sht Control Figure 16 Judges assigned weight for HDL from Time 1 to Time 2 0.24 0.26 0.28 0.30 0.32 0.34 TIME1TIME2Smoking Cue Weight C.Wts Wk Sht Control Figure 17 Judges assigned weight for smoking status from Time 1 to Time 2

PAGE 74

65 0.24 0.26 0.28 0.30 0.32 0.34 0.36 0.38 0.40 TIME1TIME2Diabetes Cue Weight C.Wts Wk Sht Control Figure 18 Judges assigned weight for diabetes status from Time 1 to Time 2 Summary The data was analyzed to test the hypothe ses that nurse practitioners receiving an experimental intervention (Comparative f eedback and Framingham education) would show better agreement and insight in judgmen t when compared to a control group. The investigator found that nurse practitioners in the Framingham worksheet condition did show an increased amount of agreemen t between their judgment model and the Framingham CHD equations, but insight into their own judgment processes did not change from Time 1 to Time 2. It was also hypothesized that providing nur se practitioners with information about the weight of cues as determined by th e Framingham CHD risk prediction equation would modify their cue weights to be more in line with the Framingham equation model. There was some evidence to support this hypot hesis. The cues that showed significant

PAGE 75

66 interactions were age and diab etes. The cues that showed significant main effects were age, systolic blood pressure, left ventricular hypertrophy, and diabetes. Chapter five discusses the analyzed resu lts of nurse practitioners decision-making process and why certain cues were significant while others were not. Furthermore, it will discuss the limitations, implications for nursi ng and future research for decision-making studies concerning nurse practitioners.

PAGE 76

67 Chapter Five Discussion The nurse practitioners participating were affiliated with the local university. Recruitment was performed in person, via tele phone, or electronic mail. Ninety-nine nurse practitioners were approached for participation and 60 agreed to participate in the study. Typically nurse practitioners that refu sed participation verbalized a fear of not performing well on the task or not feeling co mfortable with properly assessing coronary heart disease risk factors. Though many of the NPs that refused participation did work in primary care clinics, they verbalized that th ey did not routinely tr eat cardiac disease and felt uncomfortable taking part in a heart di sease assessment study. A small percentage of refusals were due to lack of time. On th e average the nurse practitioners were middleaged to older adults who had been practici ng an average of 29 years. This was a relatively small convenience sample from five counties within a geographic region of the Southeastern United States and may not be representative of the entire country. Performance of Judges for Time One A multiple linear regression analysis on the 70 cases was conducted to estimate the predictability of the judgment task. The Framingham Heart Study risk prediction equation was compared to the nurse practitione rs judgment model to assess the amount of cognitive control each judge had in their decision-making along with assessing how well an additive linear model captured NPs j udgments. The nurse practitioners’ judgment

PAGE 77

68 model indicated that their policy capturing methods were successfully captured for assessment of patient risk for coronary hear t disease and they had adequate control in their decision-making. Nurse practitioners ’ functioning on the task was of reasonable quality when compared to the ecology. Th ey could have functioned well on the tasks due, in part, to the fact that th e cues were related to the criter ion in straight forward linear ways and that the cues were select ed a priori, based on relevance. Achievement ( ra) was analyzed to determine how well the nurse practitioners agreed with the Framingham CHD risk pred iction equation. This wa s important because Social Judgment Theory is used to evalua te the connection between the individual’s judgments and ecological criterion values, na mely, achievement. Nurse practitioners’ judgments of patient risk for CHD show ed considerable agreement with the Framingham’s gold standard for making such estimates. The results did not show a significant increase in the amount of agreement by time or condition; however, there was an increasing trend from Time 1 to Time 2, es pecially with the intervention groups. This is an expected finding considering the contro l did not receive edu cation between Time 1 and Time 2. The achievement values repor ted for Time 1 and Time 2 supported nurse practitioners’ success at judging patient risk for CHD based on the eight highly relevant cues. The G index measures how well the judges’ model (nurse practitioner) agreed with the ecology model (Framingham Heart St udy risk prediction model). During the evaluation of G index, it was determined that nurse practitioner’s judgment policies were reasonably high from Time 1 to Time 2. Th ere was a significant increase in agreement between the judges’ model and the Frami ngham Heart Study risk prediction model by

PAGE 78

69 time and condition. As expected the two inte rvention groups that received education did significantly increase their agreement, while the control group remained the same over time. However, the change was not c onsistent among the three groups. Both intervention groups had a significant increase in their agreement with the Framingham CHD risk prediction model, but the Fram ingham worksheet group showed a greater degree of agreement than the comparative feedb ack group. This could be the result of the greater amount of detailed CHD risk inform ation that was covered on the Framingham worksheet than simply giving the judge compara tive weights for each risk factor alone. An overall conclusion is that nurse practitioners may perform better with more detailed and complex education information than a synopsis type format. A skill score is indicative of how much insight one may have into their decisionmaking judgments. By measuring the particip ant’s skill score it was revealed that on average most of the nurse practitioners overes timated a patient’s risk for the development of CHD within the next 10 years. The ove restimation was explai ned by unconditional bias in their judgment. The nurse practi tioners exposed to the Framingham worksheet were the only group to significantly decrea se their amount of unconditional bias and conditional bias from Time 1 to Time 2. For example, anecdotal evidence obtained during data collection reveal ed that some judges would admit to assigning a higher percentage of risk for CHD to patients with a diagnosis of LVH due to having an “already damaged” heart. The decrease in both unconditional and conditional bias for the one intervention group (Framingham worksheet group) could be explained by the amount of detailed information given within the worksheet. Speci fic gender and age related information was obtained from the Framingha m worksheet along with the percentage of

PAGE 79

70 probable risk within the next 5 and 10 year s. The highest probability of having a CHD event in the next 10 years was 42% on the risk prediction worksheet; however, some of the judges gave risk rating up to 100%. The comparative feedback only gave the judge’s subjective weighted risk for each of the eight risk factors compared to the Framingham weights for each risk factor, and the control group did not significantly make a change between Time 1 and Time 2. After completing the 70 patient profiles the judges were asked to subjectively weigh each cue for importance of risk. Th e purpose of this task was to compare the judges’ subjective weight importance value to the cognitive control value, which would indicate if the judge rated risk within the scenario booklet as they rated the importance of each cue at the end of the task. The resu lts indicated that the subjective weight importance was much lower than their cognitive control suggesting that nurse practitioners have only modest insight into their judgment polic ies. This finding has been supported in other research concerning insight of professional deci sion-making (Reilly & Doherty, 1992; Reilly & Doherty, 1989; Stew art & Lusk, 1994). Nurse practitioners did no better or worse in their amount of insight th an professions studied previously; the type of judges studied included but was not limited to physicians, accountants, weather forecasters, college students and college professors. Distribution of Importance Weights Among the Eight Cues The eight cues evaluated were gender, age, systolic blood pressure, left ventricular hypertrophy, choleste rol, HDL, smoking status, and di abetes status. Each cue was individually evaluated and compared fr om Time 1 to Time 2 for a statistically significant improvement in their judgment of risk as compared to the Framingham CHD

PAGE 80

71 risk prediction equation. Gende r was the least weighted cue out of all risk factors across all groups by the nurse practitioners (Frami ngham weighted gender seventh out of eight in risk factor importance). The comparat ive feedback condition changed the importance rating for gender from Time 1 to Time 2 in th e correct direction; how ever this was a nonsignificant change. This could be explained by the fact that the judges in this group were able to compare Framingham’s estimation of ge nder risk with their own and gave the cue more weight at Time 2, but did not change their real view of importance for this cue. The Framingham worksheet condition broke the ri sk factors down by gender when predicting percentage of 5or 10year risk. It was the expectation that this group would significantly increase their risk rating for gende r from Time 1 to Time 2 due to the more complex education given. For example, th e Framingham CHD risk prediction worksheet broke the gender cue down in the followi ng way, a woman with diabetes was given 6 points whereas a man with a diagnosis of di abetes was given 3 points. This can be interpreted that women have double the risk of having a CHD event if they are positive for the risk factor, diabetes. However, the Framingham condition did not judge the gender cue any differently than the control group. This indicates a need for further education on the importance of gender differences concerning coronary heart disease. An emphasis has been placed in the last 5 to 10 years that women experience different coronary heart disease sympto ms at different times in life than men. This study only confirms the need for further continuing education and possible curricular changes concerning this risk factor. Age was the sixth most weighted risk fact or at Time 1 and increased in weight at Time 2, the amount of importance increas ed over time and by condition. The

PAGE 81

72 intervention groups significantly increased their weight of risk; however the control group stayed the same over time. There wa s a stronger relationship with change over time for the Framingham worksheet group. This finding could be explained by the emphasis the Framingham CHD prediction worksheet placed on age by gender, the older a person’s age, the more risk points they will receive for risk of a CHD event occurring. It is not surprising that th e control group which received no education between tasks did not change their risk assessment by time. Framingham weighted age as the most important coronary heart disease risk factor out of the eight. Systolic blood pressure was the fourth most weighted risk factor at Time 1; however it decreased in weight at Time 2 fo r all groups. The comparative feedback and control group had the strongest change in ri sk rating over time. The change with the comparative feedback group may be explained by the format in which the information was given. Referring to the comparativ e feedback (Appendix A) instrument demonstrates that the judges were shown how they weighed each cue as compared to how the Framingham equation weighed each cue. This group of judges was able to visually compare how much they over-weighed systol ic blood pressure at Time 1 and thereby decrease the amount of weight assigned dur ing the Time 2 task. The change in the control group was significant but not as great as the compar ative feedback group. This could have been due to learni ng the task over time or gain ing knowledge about systolic blood pressure between tasks. The judges ra ted left ventricular hypertrophy as the third most important cue for both Time 1 and Time 2. However, there was not a significant interaction among groups. This could be e xplained by a judge having knowledge that LVH was a significant risk factor for the de velopment of a CHD event and it remained

PAGE 82

73 stable overtime and condition. Framingham weighted SBP as fi fth most important out of the eight risk factors. The cues cholesterol, HDL, and smoking status were ranked as fifth, seventh, and second weighted cues with no significant ch ange over time or c ondition respectively. Framingham weighted cholesterol, HDL, and sm oking status as the ei ghth, fourth, and six most important risk factors out of the ei ght respectively. The cholesterol cue was overestimated in risk of importance by the nurse practitioners but decreased over time with the exception of the control group. Th e HDL cue had no change for the intervention groups; however, the control group did increase their weight importance from Time 1 to Time 2. The test for this c ondition may indicate a slight lack of power and the need to increase sample size to signi ficantly detect the differences among the groups. The nurse practitioners’ lack of original knowledge about the normal range for HDL levels at Time 1 could explain why there was a lack of cha nge between the Time 1 and Time 2 tasks. Smoking status at Time 1 was weighted quite high for all groups; however the intervention groups did decrease their wei ghing of risk over time. The control group actually increased their weight of risk from Time 1 to Time 2. This drop in risk rating can be explained by the education given con cerning smoking status for both intervention groups. It is not unlikely that the control group did not change their weight estimate for smoking status due to the fact that no educati on was given; it also may be considered the most modifiable risk factors for CHD, theref ore, viewed as less tolerable by clinicians. Diabetes was considered the highest rate d risk factor among all groups from Time 1 to Time 2 as compared to Framingham that ranked diabetes as the third most important risk factor. There was a change over time and by condition for this cue; however, the

PAGE 83

74 only group that significantly decreased their im portance weight for risk over time was the Framingham worksheet condition. All subjects overestimated the weight for diabetes; the control group continued to overestimate ri sk at Time 2 and the intervention groups decreased their weight importance at Time 2. This result can be explained by the educational information supplied to the inte rvention groups. The c ontrol group received no information between tasks, and therefore, did not change thei r beliefs over time. Implications for J udgment Researchers The study demonstrates that an additive linear model did an adequate job of capturing the systematic way that nurse pr actitioner judges make decisions about CHD risk. The findings gives judgment researcher s a representation of how 60 professionals produced judgments and were able to analy ze the applicatio n of their judgments through the representation of the environmental mode l. The Social Judgment Theory provides another way of evaluating educational inte rventions for professi onals. Many current evaluations simply focus on changes in knowledg e rather than practi ce. The results of this study may provide not only an estimate of impact concerning correct assessment of risk, but also a source of explanation of when the need for interventions is appropriate. Finally, SJT allows for the development of pr edictive models, validation and reference to real ecologies as a mean of adding incr eased value to the analysis of risk. Limitations Although the validity of penc il and paper patient profiles has been demonstrated in several studies examining clinical decision making (Fisch, Hammond, Joyce & O’Reilly, 1981; Gillis, Lipkin, & Moran, 1981; Holzworth & Wills, 1999; Rothert, 1982; Smith, Gilhooly, & Walker, 2003), it is possible that the nurse practit ioners studied here

PAGE 84

75 could weigh patient characteristics differently in practice than they did on our policy capturing task. Only eight CHD cues were used in this study which may not have captured all complex aspects that nurse prac titioners may consider when diagnosing and treating a patient for CHD risk. However, this study demonstrates how well the nurse practitioners whom participated in this study performed next to the gold standard of CHD risk prevention and treatment. The patient profiles used here were quite brief and presented in a form not usually seen by nurse practitioners. Although efforts were made to produce a set of patient profiles representative of those used to develop the Framingham CHD risk prediction rule, diffe rent case mixes will presumably produce different cue weights. A second limitation of this study is that ot her factors such as insurance status and family history were not provided within the scenarios. This info rmation may moderate judgment for profiled patients. Documenta tion of the thinking/j udgment processes of nurses and the critical cues to actions used during actual patient situations could be evaluated. The sample size was a limitation; it was cl ear during the analysis of the data that some of the cues may have shown a signifi cant change over time and by condition if the study had more power. The sample consiste d of only two men; however, this was not seen as a limitation due to fact that men st atistically comprise of approximately four percent of nurses in the United States. The two male nurse practitioners in this study equaled 3.3 % of the sampled 60 nurse practitioners. Education programs and practice regulations (licensure, prescriptive privileges) for nurse practitioners vary considerably from state to state. The results reported in this

PAGE 85

76 study are descriptive of judgme nt processes and demonstrate that there is considerable variation among nurse practitioners. The final limitation was that nurse practitioners received the minimal manipulation in this study. Future studies may need to include the option of giving multiple choices that more closely mimic the decisions that nurse practitioners make in real practice. For example, a study by Re yna and Lloyd (2006) l ooked at how physician and student physicians made decisions regarding the deviation from the guidelines concerning the treatment of cardiac risk. The authors gave each judge the choice to treat a patient presenting to the emergency depart ment (ED) with chest pain by sending them to a medical-surgical unit without telemetry, admission to a telemetry unit in the hospital, admission to a cardiovascular intensive care unit, sending them home with a follow-up appointment with their local physician, or othe r (specified by the participant). Future studies concerning how nurse pr actitioners make decisions re garding cardiac risk could display such choices for detection and treatment. Foundations for Future Research In this study, the analysis was conducted us ing an additive linear model to capture the judgments of nurse practitioners concerni ng CHD risk of patients. Other nonlinear or configural judgment models may be used in future studies. Medical prediction rules may include different scoring algorithms for male and female patients reflecting the inherent nonlinear ecological relationships among patient cues. In th e context of judging patient risk for disease, it is unknown how intuitiv e such inter-cue relationships are to nurse practitioners despite dissemin ation of prediction rules in the professional literature. Future studies could explore the extent to wh ich nurse practitioners perceive such inter-

PAGE 86

77 cue relationships and use configural and multiplicative judgment models when forming patient risk assessments. The Framingham worksheet could be of continued use to evaluate how the worksheet presented under different conditions may change nurse practitioners accuracy in CHD risk assessment. Furthermore, other risk factors such as family history may be added as a cue to mimi c the more realistic information that nurse practitioners are exposed to when meeting with a patient and collecting risk factor information. Also, the inclusion of insurance status could be an im portant factor when the clinician is making a decision to refer or not refer to a specialist for follow-up. Implications for Nurse Practitioners and Educators The Social Judgment approach to pl anning and evaluating nurse educational interventions allows an objective evaluation of how professional clinicians make complex decisions on a daily basis. Contemporary ed ucational strategies ar e learner centered and target interventions as mean s of changing knowledge and prac tice. These results suggest that it is possible to construc t and study the information used by nurse practitioners and evaluate the complex judgment and decisionmaking techniques associated with practice for future development of appropriate educat ional opportunities. The lack of agreement with the Framingham risk equation concerni ng the cue, gender, indicate a need for education about the gender differences in CHD risk factors to increase awareness. This has implications for the development of educational opportun ities and continuing education modules. Also, this study indi cates how well evidence-based practice increases nurse practitioners knowledge of C HD risk factors in a primary care population. As healthcare becomes more complex and nurse practitioners are given increasing responsibility for assessment, prevention, a nd treatment of CHD mo re evidence-based

PAGE 87

78 practice needs to be implemented. Nurse prac titioners like many clinicians need to be able to cite why they made the decision to treat, not treat, or refer for follow-up. These types of evidence based decision-making tech niques will improve patient outcomes as well as decrease a nurse practiti oners risk for liability.

PAGE 88

79 References Abbott, R., Donahue, R., Kannel, W., Wilson, P. (1988). The impact of diabetes on survival following myocardial infarc tion in men vs women. The Framingham study. Journal of American Medical Association, 260 3456-3460. Aigner M, Drew S, Phipps J. (2004). A comparative study of nursing home resident outcomes between care provided by nurse practitioners/physicians versus physician only. Journal of American Medi cal Directory Association, 5, 16-23. American Heart Association. Heart Disease and Stroke Statistics --2005 Update. (2004) Dallas: American Heart Association. American Academy of Nurse Practitioners, Revised 2002. http://www.nurse.org/acnp/facts/whatis.shtml Anderson, J., and Kennenich, C. (2001). Women and coronary heart disease. Nurse Practitioner, 26(8), 12-31. Anderson K, Wilson P, Odell P, Kannel W. (1991). An updated coronary risk profile: a statement for health professionals. Circulation 83(1), 356-362. Assman, G., Gotto, A., DPhil. (2004). HDL cholesterol and protective factors in atherosclerosis. Circulation, 109, SIII 8SIII 14. Barker, W., Mullooly, J., and Getchell, W. (2006). Changing incidence and survival for heart failure in a well-defined older population, 1970-1974 and 1990-1994. Circulation, 113, 799-805.

PAGE 89

80 Berlowitz D., Ash, A., Hickey, E. (1998). Inadeq uate management of blood pressure in a hypertension population. New England Journal of Medicine, 339 1957-63. Brown, D., Giles, W., and Croft, J. (2000). Le ft ventricular hypertr ophy as a predictor of coronary heart disease mortalit y and the effect of hypertension. American Heart Journal, 140, 848-56. Chobanian, A., Bakris, G., Black, H., Cush man, W., Green, L., Izzo, J., Jones, D., Materson, B., Oparil, S., Wright, J., and Roccella, E. (2003). The national heart, lung, and blood institute joint national committee on prevention, detection, evaluation, and treatment of high blood pressure; national high blood pressure education program coordinating committee. The seventh report of the joint national committee on prevention, dete ction, evaluation, an d treatment of high blood pressure: the JNC 7 report. Hypertension, 42 1206-1252. Colhoun, H. (2006). Coronary heart disease in women: why the disp roportionate risk? Current Diabetes Reports 6, 22-28. Colhoun, H., Betteridge, D. (2006). Treatment of lipid disorders in patients with Diabetes. Current Treatment Options in Cardiovascular Medicine, 8, 37-45. Collins, R., Armitage, J., Parish, S., Slei ght, P., Peto, R. (2002). MDC/BFH Heart Protection study of cholesterol loweri ng with simvastatin in 20,536 high-risk individuals: a randomized placebo-controlled trial. Lancet, 360, 7-22. Cook N., Cohen, J., Hevert, P., Taylor, J., a nd Hennekens, C. (1995). Implications of small reductions in diastolic blood pressure for primary prevention. Archieves of Internal Medicine, 155, 701-709. Cooksey R. (1996). Judgment Analysis: Theory, Methods, and Applications San Diego:

PAGE 90

81 Academic Press. Cooksey, R.W. (1988). Social judgment theory in educ ation: current and potential applications. In B. Brehmer & C.R.B. Joyce (Eds.), Human judgment: The SJT view Amsterdam: North Holland Elsevier, 273-316. Curb, J., Abbott, R., Rodriguez, B., Masaki K., Popper, J., Chen, R., Petrovitch, H., Blanchette, P., Schatz, I., and Yano, K. (2004). Prospective association between low and high total and low-density li poprotein cholesterol and coronary heart disease in elderly men. Journal of American Geriatric Society, 52, 1975-1980. Cushman, W., Ford, C., Cutler, J., Margolis K., Davis, B., Grimm, R., Black, H., Hamilton, B., Holland, J., Nwachuku, C., Papdemetriou, V., Robstfieldd, J., Wright, J., Alderman, M., Weiss, R., P iller, L., Bettencourt, J., and Walsh, S. (2000). Success and predictors of blo od pressure control in diverse North American settings: the anti hypertensive and lipid-lowe ring treatment to prevent heart attack trial (ALLHAT). The Journal of Clinical Hypertension, 283, 19731975. De Simone, G., Devereux, R., Chinali, M., Roman, M., Best, L., Welty, T., Lee, E., and Howard, B. (2006). Risk factors fo r arterial hypertension in adults with initial optimal blood pressu re. The Strong Heart Study. Hypertension, 47, 162167. Devereux, R., Roman, M., Parani cas, M., O’Grady, M., Lee, E., Welty, T., Fabsitz, R., Robbins, D., Rhoades, E., and Howard, B. (2000). Impact of diabetes on cardiac structure and function: the strong heart study. Circulation, 101 2271-2276. Eaton, C., Galliher, J., McBride, P., Bonham, A., kappus, J., and Hickner, J. (2006).

PAGE 91

82 Family physician’s knowledge, beliefs, and self-reported practice patterns regarding hyperlipidemia: a nationa l research network (NRN) survey. Journal of American Board of Family Medicine, 19 46-53. Epstein A, McNeil B. (1985). Physician ch aracteristics and orga nizational factors influencing use of ambulatory tests. Medical Decision Making, 5, 401-15. Fields L, Burt V, Cutler J, Hughes J, Ro ccella E, & Sorlie P. (2004). The burden of adult hypertension in the United States 1999 to 2000: A rising tide. Hypertension, 44 ,398-404. Flather, M., Ysuf, S., Kober, L., Pfeffer, M., Hall, A., Murray, G., Torp-Pedersen, C., Ball, S., Pogue, J., Moye, L., and Braunwald, E. (2000). Long-term ACE-inhibitor therapy in patients with heart failure or left-ventricular dysfunction: a systematic overview of data from individual patients. Lancet, 355, 1575-81. Flynn, M. (1994). Judgment theories and residential placement decision making within aged care Unpublished Honors thesis, Univer sity of New England, Armidale, NSW, Australia. Folsom, A., Chambless, L., Duncan, B., Gilbert, A., Pankow, J. (2003). The atherosclerosis risk in communities study investigators: prediction of coronary heart disease in middle-aged adults with diabetes. Diabetes Care, 26, 2777-2784. Foody, J., Shah, R., Galusha, D., Masoudi, F., Havranek, E., and Krumholz, H. (2006). Statins and mortality among elderly pa tients hospitalized wi th heart failure. Circulation, 113, 1086-1092. Franco, O., Peeters, A., Bonneux, L., and L aet, C. (2005). Blood pressure in adulthood

PAGE 92

83 and life expectancy with cardiovascular disease in men and women. A life course analysis. Hypertension, 46, 280-286. Frijling, B., Lobo, C., Keus, I., Jenks, K., A kkermans, R., Hulscher, M., Prins, A., van der Wouden, and Grol, R. (2004). Perceptions of cardi ovascular risk among patients with hypertension or diabetes. Patient Education and Counseling, 52, 47-53. Frijling, B., Spies, T., Lobo, C. (2001). Blood pressure control in treated hypertensive Patients: clinical performan ce of general practitioners. British Journal of General Practitioners, 51(462), 9-14. Gillis J, Lipkin J, & Moran T. (1981). Dr ug therapy decisions: a social judgment analysis. Journal of Nervous and Mental Disease, 169(7), 439-447. Gomberg-Maitland, M. (2001). Tr eatment of congestive heart failure. Guidelines for the primary care physician and the heart failure specialist. Archieves of Internal Medicine, 161, 342-350. Goff, D., Bertoni, A., Kramer, H., Bonds, D ., Blumenthal, R., Tasi, M., and Psaty, B. (2006). Dyslipidemia prevalence, treat ment, and control in the multi-ethnic study of atherosclerosis (MESA). Gende r, ethnicity, and coronary artery calcium. Circulation, 113 647-656. Grimshaw, J., Shirran, L., Thomas, R., Mowatt, G., Fraser, C., Bero, L., Grilli, R., Harvey, E., Oxman, A., and O’Brien, M. (2001). Changing behavior: an Overview of systematic reviews of interventions. Medical Care, 39(8), 2-45. Grundy, S., Cleeman, J., Merz, C., Brewer, H ., Clark, L., Hunninghake, D., Pasternak, R., Smith, S., and Stone, N. (2004). The na tional heart, lung, and blood institute,

PAGE 93

84 the American college of cardiology founda tion, the American heart association: implications of recent clinical trials fo r the national choleste rol education program adult treatment panel III guidelines. Circulation, 110, 227-239. Gu, K., Cowie, C., and Harris, M. (1998). Mort ality in adults with and without diabetes In a national cohort of the U.S. population, 1971-1993. Diabetes Care, 21(7), 1138-1145. Gu, Q., Paulose-Ram, R., Dillon, C., and Bu rt, V. (2006). Antihypertensive medication use among US adults with hypertension. Circulation, 113 213-221. Haffner, S., Lehto, S., Ronnemaa, T., Pyorala, K., Laakso, M. (1998). Mortality from coronary heart disease in subjects w ith type 2 diabetes and in nondiabetic subjects with and without prior myocardial infarction. New England Journal of Medicine, 339, 229-234. Hammond KR, Stewart TR, Brehme r B, Steinmann DO. (1975). So cial judgment theory. In M. Kaplan & S. Schwartz (Eds.), Human judgment and decision processes New York: Academic Press. Hammond, K., Mumpower, J., and Smith, T. (1977). Linking environmental models with models of human judgme nt: a symmetrical decision aid. IEEE Transactions on Systems, Man, and Cybernetics, 7(5), 358-367. Harvey, L. (1992). The critical operating ch aracteristic and the evaluation of expert judgment. Organizational Behavior and Huma n Decision Processes, 53, 229251. Harvey, L., Hammond, K., Lusk, C., and Mross, E. (1992). The application of signal detection theory to weather forecasting behavior. Monthly Weather Review, 120,

PAGE 94

85 863-883. Heald, J.E. (1991). Social judgment theory: a pplications to educational decision making. Educational Administration Quarterly, 27 343-357. Hochman, J., Tamis, J., Thompson, T., Weav er, D., White, H., Van De Werf, F., Alyward, P., Topol, E., and Califf, R. (1999). Sex, clinical presentation, and outcome in patients with acute coronary syndromes. The New England Journal of Medicine, 341, 226-32. Holmes M, Rovner D, Rothert M, Schmitt N, Given C, & Ialongo N. (1989). Methods of analyzing physician practice patterns in hypertension. Medical Care, 27(1), 59-68. Horrocks S, Anderson E, Salisbury C. (2002). Systematic review of whether nurse practitioners working in primary car e can provide equivalent care to doctors. British Medical Journal 324, 819-823. Howard, B., Best, L., Galloway, J., Howard, W., Jones, K., Lee, E., Ratner, R., Resnick, H., and Devereux, R. (2006). Coronary hear t disease risk equiva lence in diabetes depends on concomitant risk factors. Diabetes Care, 29, 391-397. Hughes S, Hayman L. (2004). Improving cardiov ascular health in women: an opportunity for nursing. Journal of Cardiovascular Nurs ing; 19(2),145-147. Hulley, S., Grady, D., Bush, T., et al. (1998). For the heart and estrogen/progestin replacement study (HERS) research group randomized trial of estrogen puls progestin for secondary prevention of cor onary heart disease in postmenopausal women. Journal of American Medical Association, 280(7), 605-13. Hurst W. The heart, arteries, and veins. 10th ed. New York, NY: McGraw-

PAGE 95

86 Hill; 2002. Joint National Committee. (1997). The sixth report of the joint national committee on prevention, detection, evaluation, an d treatment of high blood pressure. Archieves of Internal Medicine, 157 2413-2446. Jones, D., Chambless, L., Folsom, A., Heiss, G., Hutchinson, R., Sharrett, R., Szklo, M., and Taylor, H. (2002). Ri sk factors for coronary h eart disease in African Americans. The atherosclerosis risk in communities study, 1987-1997. Archieves of Internal Medicine, 162, 2565-2571. Kahn H, Sempos C. Statistical Methods in Epidmeiology (1989). 12th ed. New York, NY: Oxford University Press. Kannel, W. (2000). Incidence and epidemiology of heart failure. Heart Failure Reviews, 5, 167-173. Kannel, W., and Wilson, P. (1995). Risk fact ors that attenuate the female coronary disease advantage. Archieves of Internal Medicine, 155 57-61. Kawachi, I., Coldwitz, G., Stampfer, M., Wille tt, W., Manson, J., Rosner, B., Speizer, F., and Hennekens, C. (1994). Smoking ce ssation and time course of decreased risks of coronary heart disease in middle-aged women. Archieves of Internal Medicine, 154, 169-175. Kawachi, I., Colditz, G., Stampfer, M., Wille tt, W., Manson, J., Rosner, B., Hunter, D., Hennekens, C., and Speizer, F. (1993). Smoking cessation in relation to total mortality rates in women. A prospective cohort. Annuals of Internal Medicine, 119, 992-1000. Kinnersley P, Anderson E, Parry K, et al (2000). Randomised c ontrolled trial of

PAGE 96

87 nurse practitioner versus gene ral practitioner care fo r patients requesting “same day” consultations in primary care. British Medical Journal 320, 10431048. Kuller, L. (2006). Prevention of coronary heart disease and the national cholesterol education program. Circulation, 113 598-600. Kuusisto, J., Lempiainen, P., Mykkanen, L., and Laasko, M. (2001). Insulin resistance syndrome predicts coronary heart disease events in elderly type 2 diabetic men. Diabetes Care, 24 1629-1633. LaRosa, J., Grundy, S., Waters, D. (2005). In tensive lipid-lowering with atorvastatin in patients with stable coronary disease. New England Journal of Medicine, 3350, 1425-1435. LaRosa, J., He, J., Vupputuri, S. (1999). Effect of statins on risk of coronary disease. A meta-analysis of randomized controlled trials. Journal of American Medical Association, 282(24) 2340-2346. Lewis C. (1969). Variations in the incidence of surgery. New England Journal of Medicine, 281, 880-4. Libby, R. (1981). Accounting and human informati on processing: Theory and applications Englewood Cliffs, NJ: Prentice-Hall. Lloyd-Jones, D., Evans, J., Levy, D. (2005). Hypertension in adults across the age spectrum. Current outcomes and control in the community. Journal of American Medical Association, 294, 466-472. Massie, B. (1999). Heart. C onnecticut: Appleton & Lange. Mazza, A., Tikhonoff, V., Schiavon, L., and Casi glia, E. (2005). Triglycerides + high-

PAGE 97

88 density-lipoprotein-cholester ol dyslipidaemia, a coronary risk factor in elderly women: the cardiovascular study in the elderly. Internal Medicine Journal, 35, 604-610. Meyer S, Miers L. Cardiovascular surg eon and acute care nurse practitioner: collaboration on postoperative outcomes. AACN Clinical Issues 2005; 16(2): 149-158. Nam, B., Kannel, W., D’Agostino, R. (2006). Search for optimal at herogenic lipid risk profile: from the Framingham study. American Journal of Cardiology, 97 372-375. Nash, I., Mosca, L., Blumenthal, R., Davids on, M., Smith, S., and Pasternak, R. (2003). Contemporary awareness and understanding of cholesterol as a risk factor. Archieves of Internal Medicine, 163, 1597-1600. Natarajan, S., Liao, Y., Guichan, C., Lipstiz, S., and McGee, D. (2003). Sex differences in risk for coronary heart disease mortality associated with diabetes and established coronary heart disease. Archieves of Internal Medicine, 163 17351740. National Health and Nutrition Examinati on Survey III (NHANES III, 1988-94). (2005). http://www.cdc.gov/ nchs/products/ elec_prods/subject/ nhanes 3.htm Centers for Disease Control and Prev ention/National Center for Health statistics (CDC/NCHS). http://www.cdc.gov/nchs/ National Heart, Lung, and Blood Institute ( NHLBI). (2004). h ttp://www.nhlbi.nih.gov/ Joint National Committee on Prevention, De tection, Evaluation and Treatment of High Blood Pressure (JNC 7). http://www.nhlbi.nih.gov/gui delines/hypertension/

PAGE 98

89 Nissen, S., Tuzcu, Schoenhagen, Brown, B., Ganz, P., Vogel, R., Crowe,T., Howard, G., Cooper, C., Brodie, B., Grines, C., DeMaria, A. (2004). The REVERSAL investigators: effect of intensive compar ed with moderate li pid-lowering therapy on progression of coronary atherosclerosi s: a randomized controlled trial. Journal of American Medical Association, 291 1071-1080. Prescott, E., Hippe, M., Schnohr, P., Hein, H., Vestbo, J. (1998). Smoking and risk of myocardial infarction in women a nd men: longitudinal population study. British Medical Journal, 316, 1043-7. Psaty, B., Anderson, M., Kronmal, R., Trac y, R., Orchard, T., Fried, L. (2004). The association between lipid levels and the risk of incident myocardial infarction, stroke and total mortality. The cardiovascular health study. Journal of American Geriatrics Society, 52, 1639-47. Reilly, B., and Doherty, M. (1989). A not e on the assessment of self-insight in judgment research. Organizational Behavior and Human Decision Processes, 44 123-131. Reilly, B. and Doherty, M. (1992). The assess ment of self-insight in judgment policies. Organizational Behavior and Huma n Decision Processes, 53 285-309. Reyna, V., and Lloyd, F. (2006). Physicia n decision making and cardiac risk: effects of knowledge, risk perception, risk tolerance, and fuzzy processing. Journal of Experimental Psychology, 12(3), 179-195. Rosen, B., Saad, M., Shea, S., Nasir, K., E dvardsen, T., Burke, G., Jerosch-Herold, M., Arnett, D., Lai, S., Bluemke, D., and Lima, J. (2006). Hypertension and smoking are associated with reduced regional le ft ventricular function in asymptomatic

PAGE 99

90 individuals. Journal of American College of Cardiology, 47 1150-58. Rothert M. (1982). Physicians’ and patie nts’ judgments of compliance with a hypertensive regimen. Medical Decision Making, 2(2) 179-195. Rothert M, Rovner D, Elstein A, Holzma n G, Holmes M, & Ravitch M. (1994). Differences in medical referral decisions for obesity among family practitioners, general internists, and gynecologist. Medical Care, 22(1), 42-53. Shai, I., Rimm, E., Hankinson, S., Curhan, G., Manson, J., Rifai, N., Stampfer, M., Ma, J. (2004). Multivariate assessment of lipid parameters as predictors of coronary heart disease among postmenopa usal women. Potential implications for clinical guidelines. Circulation, 110 2824-2830. Simons, L., and Simmons, J. (1998). Diab etes and coronary heart disease. New England Journal of Medicine, 339, 1714-1715. Snow, R.E. (1968). Brunswikian appr oaches to research on teaching. American Educational Research Journal, 5, 475-489. Shulman, L.S., & Elstein, A.S. (1975). St udies of problem solving, judgment, and decision making: Implications for educatio nal research. In F.N. Kerlinger (Ed.), Review of research in education (Vol 3). Itasca: Peacock. Slovic, P., Rorer, L.G. & Hoffman, P.J. (1971). Analyzing the use of diagnostic drugs. Investigative Radiology, 6, 18-26. Smith, D.G., & Wigton, R.S. (1988). Research in medical ethics: The role of social judgment theory. In B. Brehmer & C.R.B. Joyce (Eds.), Human judgment: The SJT view Amsterdam: North Holland Elsevier, 427-442. Stewart, T. (1990). A decomposition of th e correlation coefficient and its use in

PAGE 100

91 analyzing forecasting skill. Notes and Correspondence, 5, 661-666. Stewart, T. and Lusk, C. (1994). Seven components of judgmental forecasting skill: implications for research and the improvement of forecasts. Journal of Forcasting, 13, 579-599. Surgeon General’s Health Consequences of Smoking (2004). http://www.americanheart.or g/downloadable/ heart/1135358648580551026_HS_Stats06book.pdf Tape, T., Kripal, J., Wigton, R. (1992). Co mparing methods of learning clinical prediction from case simulations. Medical Decision Making 12, 231-221. Tape, T., Heckerling, P., Ornato, J., Wigton, R. (1991). Use of clinical judgment analysis to explain regional variations in physicians’ accuracies in diagnosing pneumonia. Medical Decision Making, 11, 189-197. Tarolli, K. (2003). Left ventricula r systolic dysfunction and nonischemic cardiomyopathy. Critical Care Nurse, 26(1), 3-15. Thomspon, C., Foster, A., Cole, I., and Dowd ing, D. (2005). Using social judgment theory to model nurses’ use of clinical information in critical care education. Nurse Education Today, 25, 68-77. The Long-Term Intervention with Pravastati n in Ischaemic diseas e (LIPID) study group. (1998). Prevention of cardiovascular events and death with pravas tatin in patients with coronary heart disease and a broad range of initial ch olesterol levels. New England Journal of Medicine, 339, 1349-1357. Waller, W.S. (1988). Brunswikian research in accounting and auditing. In B. Brehmer & C.R.B. Joyce (Eds.), Human judgment: The SJT view Amsterdam: North

PAGE 101

92 Holland Elsevier, 247-272. Warren, C., Jones, N., Eriksen, M., Asma, S. (2006). Patterns of gl obal tobacco use in young people and implications for future chronic disease bu rden in adults. Lancet, 367 749-53. Wassenberg, M., Willemsen, J., and Braam, C. (2004). Hypertension management in primary care: standard care and attitu de towards a disease management model. The Netherlands Journal of Medicine, 62(10), 375-382. Whelton, P., He, J., Appel, L., Cutler, J., Ha vas, S., Kotchen, T., Roccella, E., Stout, R., Vallbona, C., Winston, M. and Karimbak as, J. (2002). Primary prevention of hypertension: clinical and pubic health advisory from the national high blood pressure education program. Journal of American Medical Association, 288(15), 1882-1888. Wigton, R.S. (1988). Applications of judgme nt analysis and cognitive feedback to medicine. In B. Brehmer & C.R.B. Joyce (Eds.), Human judgment: The SJT view Amsterdam: North Holland Elsevier, 227-246. Wigton R, Poses R, Collins M, & Cebul R. (1990). Teaching old dogs new tricks: using cognitive feedback to improve physicians’ diagnostic judgments on simulated cases. Academic Medicine, 65(9), S5-S6. Wigton, R. (1996). Social judgment theory and medical judgment. Thinking and Reasoning, 2(2) 175-190. Willet, W., Green, A., Stampfer, M., Speizer F., Colditz, G., Rosner, B., Monson, R., Stason, W., and Hennekens, C. (1987). Re lative and absolute excess risks of coronary heart disease among women who smoke cigarettes. New England

PAGE 102

93 Journal of Medicine, 317, 1303-1309. Wilson, K., Gibson, N., Willan, A., and Cook, D. (2000). Effect of smoking cessation mortality after myocardial infarction. Archieves of Inter nal Medicine, 160 939944.

PAGE 103

94 Appendix A: Brunswik Lens Model Lens Model with CHD Cues (Cooksey, 1996) YeYsCUES ACTUAL FHS Estimated Risk of CHD in 10 yrs ARNP’sJUDGMENTS of Risk of CHD in 10 yrs es Yss Yee.724 Achievement .800 Agreement FHS & ARNP models .042 Residuals of FHS and ARNPsmodel .967 Predictability .931 Control Sex SBP Age Smoke Diab .186 .189 .224 .234 .136 .246 -.276 .350 .200 -.302 LVH Chol HDL .532 -.302 .172 .260 .094 .481

PAGE 104

95 Appendix B: Experimental Manipulations Condition 1: Comparative Feedback Example: How important was each cue as you fo rmed your estimates of CHD risk? Divide 100 points among the cues belo w. Assign the most points to the cue(s) you relied on the most. ____ Gender ____ Age ____ Systolic Blood Pressure ____ Left Ventricular Hypertrophy ____ Total Cholesterol Level ____ High Density Lipoprotein Level ____ Smoker ____ Diabetes ____ TOTAL Actual importance of each cue 8.6 Gender 22.2 Age 10.4 Systolic Blood Pressure 20.1 Left Ventricular Hypertrophy 6.1 Total Cholesterol Level 11.4 High Density Lipoprotein Level 9.5 Smoker 11.6 Diabetes 100 TOTAL

PAGE 105

96 C ondition 2: Framingham Heart Study CHD Risk Prediction Worksheet Appendix C: Patient Profiles

PAGE 106

97 Appendix C: Patient Profiles Patient Number: 1 Gender: FEMALE Age: 57 Systolic Blood Pressure: 144 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 179 High Density Lipoprotein Level: 64 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 2 Gender: MALE Age: 33 Systolic Blood Pressure: 124 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 169 High Density Lipoprotein Level: 34 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 107

98 Patient Number: 3 Gender: FEMALE Age: 63 Systolic Blood Pressure: 170 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 233 High Density Lipoprotein Level: 69 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 4 Gender: FEMALE Age: 53 Systolic Blood Pressure: 149 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 219 High Density Lipoprotein Level: 56 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 108

99 Patient Number: 5 Gender: MALE Age: 67 Systolic Blood Pressure: 176 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 230 High Density Lipoprotein Level: 34 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 6 Gender: MALE Age: 55 Systolic Blood Pressure: 129 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 225 High Density Lipoprotein Level: 42 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 109

100 Patient Number: 7 Gender: FEMALE Age: 65 Systolic Blood Pressure: 135 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 291 High Density Lipoprotein Level: 46 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 8 Gender: FEMALE Age: 57 Systolic Blood Pressure: 137 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 247 High Density Lipoprotein Level: 57 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 110

101 Patient Number: 9 Gender: MALE Age: 44 Systolic Blood Pressure: 107 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 176 High Density Lipoprotein Level: 34 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 10 Gender: FEMALE Age: 71 Systolic Blood Pressure: 165 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 317 High Density Lipoprotein Level: 50 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 111

102 Patient Number: 11 Gender: MALE Age: 72 Systolic Blood Pressure: 128 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 218 High Density Lipoprotein Level: 35 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 12 Gender: FEMALE Age: 72 Systolic Blood Pressure: 140 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 226 High Density Lipoprotein Level: 45 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 112

103 Patient Number: 13 Gender: FEMALE Age: 55 Systolic Blood Pressure: 100 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 256 High Density Lipoprotein Level: 41 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 14 Gender: FEMALE Age: 56 Systolic Blood Pressure: 119 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 262 High Density Lipoprotein Level: 46 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 113

104 Patient Number: 15 Gender: MALE Age: 65 Systolic Blood Pressure: 173 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 191 High Density Lipoprotein Level: 29 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 16 Gender: FEMALE Age: 52 Systolic Blood Pressure: 130 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 182 High Density Lipoprotein Level: 50 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 114

105 Patient Number: 17 Gender: MALE Age: 35 Systolic Blood Pressure: 144 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 187 High Density Lipoprotein Level: 53 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 18 Gender: FEMALE Age: 65 Systolic Blood Pressure: 151 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 257 High Density Lipoprotein Level: 61 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 115

106 Patient Number: 19 Gender: MALE Age: 55 Systolic Blood Pressure: 132 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 221 High Density Lipoprotein Level: 36 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 20 Gender: MALE Age: 65 Systolic Blood Pressure: 153 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 230 High Density Lipoprotein Level: 47 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 116

107 Patient Number: 21 Gender: MALE Age: 34 Systolic Blood Pressure: 118 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 224 High Density Lipoprotein Level: 35 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 22 Gender: FEMALE Age: 55 Systolic Blood Pressure: 163 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 243 High Density Lipoprotein Level: 44 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 117

108 Patient Number: 23 Gender: FEMALE Age: 65 Systolic Blood Pressure: 117 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 268 High Density Lipoprotein Level: 44 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 24 Gender: FEMALE Age: 54 Systolic Blood Pressure: 120 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 278 High Density Lipoprotein Level: 70 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 118

109 Patient Number: 25 Gender: FEMALE Age: 53 Systolic Blood Pressure: 137 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 311 High Density Lipoprotein Level: 66 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 26 Gender: MALE Age: 54 Systolic Blood Pressure: 136 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 175 High Density Lipoprotein Level: 39 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 119

110 Patient Number: 27 Gender: FEMALE Age: 57 Systolic Blood Pressure: 128 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 175 High Density Lipoprotein Level: 56 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 28 Gender: FEMALE Age: 44 Systolic Blood Pressure: 110 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 188 High Density Lipoprotein Level: 49 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 120

111 Patient Number: 29 Gender: MALE Age: 33 Systolic Blood Pressure: 126 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 236 High Density Lipoprotein Level: 60 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 30 Gender: FEMALE Age: 54 Systolic Blood Pressure: 132 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 279 High Density Lipoprotein Level: 54 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 121

112 Patient Number: 31 Gender: MALE Age: 38 Systolic Blood Pressure: 113 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 162 High Density Lipoprotein Level: 50 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 32 Gender: MALE Age: 46 Systolic Blood Pressure: 130 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 173 High Density Lipoprotein Level: 37 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 122

113 Patient Number: 33 Gender: MALE Age: 55 Systolic Blood Pressure: 116 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 249 High Density Lipoprotein Level: 31 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 34 Gender: FEMALE Age: 34 Systolic Blood Pressure: 112 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 159 High Density Lipoprotein Level: 66 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 123

114 Patient Number: 35 Gender: FEMALE Age: 43 Systolic Blood Pressure: 146 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 191 High Density Lipoprotein Level: 52 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 36 Gender: MALE Age: 55 Systolic Blood Pressure: 112 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 213 High Density Lipoprotein Level: 49 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 124

115 Patient Number: 37 Gender: MALE Age: 47 Systolic Blood Pressure: 122 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 220 High Density Lipoprotein Level: 50 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 38 Gender: FEMALE Age: 64 Systolic Blood Pressure: 150 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 227 High Density Lipoprotein Level: 50 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 125

116 Patient Number: 39 Gender: MALE Age: 65 Systolic Blood Pressure: 127 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 197 High Density Lipoprotein Level: 37 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 40 Gender: FEMALE Age: 55 Systolic Blood Pressure: 159 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 196 High Density Lipoprotein Level: 64 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 126

117 Patient Number: 41 Gender: MALE Age: 54 Systolic Blood Pressure: 115 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 179 High Density Lipoprotein Level: 46 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 42 Gender: MALE Age: 44 Systolic Blood Pressure: 121 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 218 High Density Lipoprotein Level: 34 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 127

118 Patient Number: 43 Gender: FEMALE Age: 46 Systolic Blood Pressure: 121 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 201 High Density Lipoprotein Level: 69 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 44 Gender: FEMALE Age: 42 Systolic Blood Pressure: 111 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 219 High Density Lipoprotein Level: 70 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 128

119 Patient Number: 45 Gender: MALE Age: 35 Systolic Blood Pressure: 114 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 240 High Density Lipoprotein Level: 45 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 46 Gender: MALE Age: 67 Systolic Blood Pressure: 149 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 201 High Density Lipoprotein Level: 60 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 129

120 Patient Number: 47 Gender: MALE Age: 53 Systolic Blood Pressure: 135 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 205 High Density Lipoprotein Level: 33 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 48 Gender: FEMALE Age: 53 Systolic Blood Pressure: 157 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 257 High Density Lipoprotein Level: 28 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 130

121 Patient Number: 49 Gender: MALE Age: 72 Systolic Blood Pressure: 102 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 198 High Density Lipoprotein Level: 38 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 50 Gender: MALE Age: 35 Systolic Blood Pressure: 133 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 221 High Density Lipoprotein Level: 43 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 131

122 Patient Number: 51 Gender: MALE Age: 72 Systolic Blood Pressure: 144 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 183 High Density Lipoprotein Level: 39 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 52 Gender: MALE Age: 50 Systolic Blood Pressure: 122 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 199 High Density Lipoprotein Level: 40 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 132

123 Patient Number: 53 Gender: MALE Age: 66 Systolic Blood Pressure: 178 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 215 High Density Lipoprotein Level: 47 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 54 Gender: FEMALE Age: 45 Systolic Blood Pressure: 106 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 195 High Density Lipoprotein Level: 64 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 133

124 Patient Number: 55 Gender: FEMALE Age: 64 Systolic Blood Pressure: 131 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 241 High Density Lipoprotein Level: 40 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 56 Gender: MALE Age: 45 Systolic Blood Pressure: 123 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 215 High Density Lipoprotein Level: 41 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 134

125 Patient Number: 57 Gender: MALE Age: 32 Systolic Blood Pressure: 131 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 204 High Density Lipoprotein Level: 37 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 58 Gender: FEMALE Age: 34 Systolic Blood Pressure: 109 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 145 High Density Lipoprotein Level: 58 Smoker: YES Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 135

126 Patient Number: 59 Gender: FEMALE Age: 34 Systolic Blood Pressure: 98 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 201 High Density Lipoprotein Level: 52 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 60 Gender: FEMALE Age: 46 Systolic Blood Pressure: 126 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 195 High Density Lipoprotein Level: 54 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 136

127 Patient Number: 61 Gender: MALE Age: 64 Systolic Blood Pressure: 156 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 176 High Density Lipoprotein Level: 46 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 62 Gender: MALE Age: 42 Systolic Blood Pressure: 160 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 176 High Density Lipoprotein Level: 35 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 137

128 Patient Number: 63 Gender: MALE Age: 44 Systolic Blood Pressure: 126 Left Ventricular Hypertrophy: YES Total Cholesterol Level: 174 High Density Lipoprotein Level: 34 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 64 Gender: MALE Age: 55 Systolic Blood Pressure: 133 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 246 High Density Lipoprotein Level: 32 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 138

129 Patient Number: 65 Gender: FEMALE Age: 45 Systolic Blood Pressure: 159 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 183 High Density Lipoprotein Level: 58 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 66 Gender: FEMALE Age: 56 Systolic Blood Pressure: 116 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 242 High Density Lipoprotein Level: 64 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 139

130 Patient Number: 67 Gender: MALE Age: 50 Systolic Blood Pressure: 111 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 181 High Density Lipoprotein Level: 36 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 68 Gender: FEMALE Age: 66 Systolic Blood Pressure: 137 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 178 High Density Lipoprotein Level: 58 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 140

131 Patient Number: 69 Gender: MALE Age: 46 Systolic Blood Pressure: 144 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 229 High Density Lipoprotein Level: 30 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 70 Gender: MALE Age: 53 Systolic Blood Pressure: 128 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 185 High Density Lipoprotein Level: 58 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 141

132 Patient Number: 71 Gender: FEMALE Age: 57 Systolic Blood Pressure: 144 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 179 High Density Lipoprotein Level: 64 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 72 Gender: MALE Age: 33 Systolic Blood Pressure: 124 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 169 High Density Lipoprotein Level: 34 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 142

133 Patient Number: 73 Gender: FEMALE Age: 63 Systolic Blood Pressure: 170 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 233 High Density Lipoprotein Level: 69 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 74 Gender: FEMALE Age: 53 Systolic Blood Pressure: 149 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 219 High Density Lipoprotein Level: 56 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 143

134 Patient Number: 75 Gender: MALE Age: 67 Systolic Blood Pressure: 176 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 230 High Density Lipoprotein Level: 34 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 76 Gender: MALE Age: 55 Systolic Blood Pressure: 129 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 225 High Density Lipoprotein Level: 42 Smoker: YES Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 144

135 Patient Number: 77 Gender: FEMALE Age: 65 Systolic Blood Pressure: 135 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 291 High Density Lipoprotein Level: 46 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 78 Gender: FEMALE Age: 57 Systolic Blood Pressure: 137 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 247 High Density Lipoprotein Level: 57 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 145

136 Patient Number: 79 Gender: MALE Age: 44 Systolic Blood Pressure: 107 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 176 High Density Lipoprotein Level: 34 Smoker: NO Diabetes: NO On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No Patient Number: 80 Gender: FEMALE Age: 71 Systolic Blood Pressure: 165 Left Ventricular Hypertrophy: NO Total Cholesterol Level: 317 High Density Lipoprotein Level: 50 Smoker: NO Diabetes: YES On a scale from 0% to 100%, Estimate this patient's risk for CHD within the next 10 years. Estimated Risk is ______% Would you refer this patient to a cardiologist? _Yes _No

PAGE 146

About the Author Kelly D. Stamp received a Bachelor of Science in Nursing Degree from Southeast Missouri State University, Cape Girardeau, Missouri in 1998 and a Master of Nursing Degree from the University of South Florid a, Tampa, Florida in 2004. She was a staff nurse in a cardiac surgical unit and medical intensive care unit from 1998 to 2002 before returning to the University of South Florida to complete her Master and Doctoral degrees. Ms. Stamp has remained active in criti cal care nursing and is currently an Instructor at the University of South Florida. After th e completion of her Doctoral education, Ms. Stamp remained in academia and research.


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 001929724
003 fts
005 20080305140455.0
006 m||||e|||d||||||||
007 cr mnu|||uuuuu
008 080305s2006 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0001811
035
(OCoLC)212850051
040
FHM
c FHM
1 100
Stamp, Kelly D.
0 245
Advanced registered nurse practitioners' judgments of coronary heart disease risk
h [electronic resource] /
by Kelly D. Stamp.
260
[Tampa, Fla] :
b University of South Florida,
2006.
520
ABSTRACT: Coronary heart disease (CHD) is the single largest killer of American males and females in the United States. According to the American Heart Association, (2005) approximately 41% of Americans that experience a coronary attack in a given year will die from it (AHA, 2005). To combat this growing problem, strategies need to be evaluated to assess how the identification of actual and potential CHD risks are made. This study utilized the Social Judgment Theory to gain insight into nurse practitioner's decision-making strategies. Sixty family or adult specialty nurse practitioners affiliated with the University of South Florida (USF) College of Nursing volunteered to take part in a pretest-posttest experimental design. They were randomly assigned to one of three conditions. Condition 1 and 2 received educational interventions and Condition 3 served as the control group, which received no education. This design was used to assess the effects of educational feedback on improving judgment accuracy, achievement, and insight. The findings indicated nurse practitioners agreement with the Framingham prediction model of CHD risk did improve significantly for the two intervention groups from Time 1 to Time 2 (p < .05). the participants also showed a relatively high degree of cognitive control when judging and performing the policy-capturing task (average Rs = .88) as compared to Framingham (Re = .96). Significant amount of unconditional bias (F(2, 57) = 9.85, p < .01) and conditional bias (F(2, 57). 5.42), p < .05) was present in this sample. Nurse practitioners overall performed well when compared with the Framingham Heart Study risk equation, however, nurse practitioners showed little insight into their judgment process. The results of this study may provide the opportunity for nurse practitioners to offer patients more appropriate medicinal and diagnostic treatments. Future cardiac events may be avoided through evidenced-based CHD education for nursepractitioners.
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 136 pages.
Includes vita.
590
Adviser: Mary Webb, Ph.D., R.N.
653
Decision-making.
Social judgment theory
Clinical assessment.
Nursing.
Patients.
690
Dissertations, Academic
z USF
x Nursing
Doctoral.
049
FHMM
090
RT41 (ONLINE)
773
t USF Electronic Theses and Dissertations.
4 856
u http://digital.lib.usf.edu/?e14.1811