|USFDC Home | USF Electronic Theses and Dissertations||| RSS|
This item is only available as the following downloads:
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 001967511
007 cr mnu|||uuuuu
008 081027s2007 flu sbm 000 0 eng d
datafield ind1 8 ind2 024
subfield code a E14-SFE0002091
Whaley, Mirtha Montejo.
Effect of personal and practice contexts on occupational therapists' assessment practices in geriatric rehabilitation
h [electronic resource] /
by Mirtha Montejo Whaley.
[Tampa, Fla.] :
b University of South Florida,
ABSTRACT: Despite considerable debate surrounding an age associated cognitive decline in non-demented elders, recent studies indicate that changes attributable to normal aging affect cognitive processes and fluid abilities. Additionally, studies indicate that factors such as physical illness, depression, neurological damage, medication side effects, drug interactions, and the effects of surgery and anesthesia may also cause varying degrees of cognitive impairment. Impairment of cognitive function is known to affect treatment and rehabilitation outcomes for older persons, and increase their likelihood of institutionalization. Although proper screening and identification of cognitive deficits in geriatric patients are crucial in developing treatment plans, there is evidence in the medical and nursing literature that cognitive decline in older non-demented patients is often not identified.^ ^Proper screening in this case, refers not only to whether or not clinicians engage in assessment behavior, but that they adhere to evidence-based practices and utilize standardized instruments which can identify the type, extent, and implications of the cognitive deficits. This study used an exploratory, non-experimental design and the population of interest consisted of occupational therapists providing physical rehabilitation to patients >65 years of age in the United States. The Ecological Systems Model was chosen as the theoretical framework, because it depicts human behavior as the product of the interaction between the individual's personal attributes and the physical and social environment in which the individual functions. Given the changes in health care, and the limits imposed by third party payers, it would seem important to inquire as to the effect of personal and practice contexts on therapists' assessment practices in geriatric rehabilitation.^ ^Although results of the study indicate that factors in the practice context are stronger predictors of therapists' use of standardized cognitive screening and assessment instruments, the study supports principles of Ecological Systems Models in that both practice and personal contexts contribute to therapists' assessment practices.
Dissertation (Ph.D.)--University of South Florida, 2007.
Includes bibliographical references.
Text (Electronic dissertation) in PDF format.
System requirements: World Wide Web browser and PDF reader.
Mode of access: World Wide Web.
Title from PDF of title page.
Document formatted into pages; contains 194 pages.
Co-adviser: Elizabeth Gulitz, Ph.D.
Co-adviser: James Mortimer, Ph.D.
Age associated cognitive decline.
Cognition and rehabilitation outcomes.
x Public Health
t USF Electronic Theses and Dissertations.
Effect of Personal and Practice Contexts on Occupational Therapists' Assessment Practices In Geriat ric Rehabilitation by Mirtha Montejo Whaley A dissertation submitted in partial fulfillment of the requirements for the degree of Doctor of Philosophy Department of Community and Family Health College of Public Health University of South Florida Co-Major Professor: El izabeth Gulitz, Ph.D. Co-Major Professor: James Mortimer, Ph.D. Karen Liller, Ph.D. Karen Perrin, Ph.D. Date of Approval: June 15, 2007 Keywords: age associated cognitive declin e, cognition and rehabilitation outcomes, cognitive assessment, geriatric re habilitation, occ upational therapy Copyright 2007 Mirtha M. Whaley
Dedication Ancora Imparo (Im still learning) Michaelangelo This dissertation is dedicated to my pare nts and my former patients, who sparked in me the curiosity to inquire; to my da ughter Tanya, my sister Daisy and our family, whose support helped me find the courage to challenge myself and persevere, no matter how difficult the journey became; to my mentors and professors, whose encouragement and wisdom guided me so that I may turn my fire and passion into reasonable science; to my friends and colleagues who have shown me how its done and cheered me on to the finish line; and to Larry, my late husband, who so anxiously awaited my graduation, and was convinced that the time to comp letion was measured in dog years. This dissertation is as much their triumph as it is mine. This has been more than a scholarly journey to a higher de gree; it has been a test of re silience and determination and at times a lifeline. We can all now joyfully say Ph-inally D-one!
Acknowledgements I would like to acknowledge and than k the individuals whose support and assistance were crucial to my success, be ginning with J. Neil Henderson Ph.D., and Magdalene Argiry, who encouraged me to embark on this adventure. To my Co-Chairs, Elizabeth Gulitz Ph.D., whose knowledge, guidance and support throughout this process have been i nvaluable to my personal and professional growth; and James Mortimer Ph.D., for his resp onsiveness, willingness to assist and share his knowledge and for challenging me to st retch beyond what I thought were my limits. To Karen Liller, Ph.D., and Karen Perrin, Ph.D ., for their infinite patience and guidance. To Kay White, Beverly Sanchez and the A cademic Affairs staff, and also to Tom Ross, Paula Geist, Sandhya Srinivasan, We ndy Lay, Carol Williams and the faculty and staff of the College of Public Health for their unconditional support and technical assistance. A special thanks to Chodaesessi e Wellesley-Cole Morgan, Ph.D., for the late nights, the helping hand, the endless tutorials, and her unconditional friendship. To Angela Edney and the occupational ther apy staff at Aegis Therapies; Dyanne Havner and occupational therapists at the Ta mpa V.A. Medical Center; to Pam Toto and Ann OSullivan Chairs of the special interest sections at AOTA; to Kathy Farber, Nancy Allen and all the members of my panel of experts who helped refine my survey instrument; and, finally, to the occupational therapists who contributed to my study and my professional development. Without them, this journey would not have been possible.
i Table of Contents List of Tables__________________________________________________________vi Abstract______________________________________________________________viii Chapter One: Introduction _________________________________________________1 Overview of Aging: A Historical Perspective ____________________________4 Cognition: Issues and Implicat ions for Occupational Therapy _______________6 Determining What to Assess: Clinical Reasoning in Occupational Therapy____10 Statement of the Problem ___________________________________________11 Need for the Study ________________________________________________13 Theoretical Model_________________________________________________16 Implications for Public Health_______________________________________17 Research Questions________________________________________________18 Delimitations_____________________________________________________20 Limitations of the Study____________________________________________20 Assumptions_____________________________________________________21 Definitions_______________________________________________________23 Chapter Two: Review of the Literature______________________________________27 Overview of cognition_____________________________________________27
ii Factors Affecting Cognition: Aging and Disease________________________29 Cognition and Functional Status: Per ceptions of Health Professionals________34 Functional Status as Behavior al or Performance-based______________35 Cognitive Status as a Separate C onstruct from functional status_______35 Cognitive Function Assumed Through Instrumental Activities________35 Cognition as a Dimension of Functional Status____________________36 Geriatric Rehabilitation_____________________________________________37 Occupational Therapy Practice_______________________________________39 Clinical Reasoning in Occupational Therapy______________________42 Ecological Systems Model and Occupational Therapy______________44 Chapter Three: Methods__________________________________________________49 Type of Study, Population, and Sample Selection________________________49 Variables________________________________________________________50 Instrument Development____________________________________________52 Validity___________________________________________________54 Reliability_________________________________________________56 Data Collection_____________________________________________59 Response Rates_____________________________________________63 Data Management___________________________________________64 Data Analyses________________________________________65 Missing Data_________________________________________66
iii Testing Data for Normality______________________________66 Bivariate Analyses____________________________________67 Non-responders ______________________________________68 External Validity of the sample__________________________69 Effect of Social Desirability _____________________________70 Multivariate Analysis __________________________________70 Power Analysis and Sample Size _________________________72 Chapter Four: Results____________________________________________________74 Overview of the Study_____________________________________________74 Description of the Case Selection Process________________________75 Description of Non-responders_________________________________76 Description of Study Sample________________________________________78 Demographic Characteristics of Participants ______________________78 Description of the Clinical Practice_____________________________79 External Validity of the sample________________________________82 Effect of Social Desirability___________________________________83 Results of Research Questions_______________________________________84 Assessing the Logistic Model________________________________________94 Assessing the Logistic Regression____________________________________95 Chapter Five: Discussion_________________________________________________97 Effect of Context on Therapists Assessment Practices ___________________97 Knowledge of Aging and Cognition__________________________________101
iv Beliefs_________________________________________________________103 Practice Setting__________________________________________________105 Implications for Occupational Therapy Practice________________________105 Public Health Implications_________________________________________108 Limitations of the Study___________________________________________109 Strengths of the Study_____________________________________________110 Recommendations for Future Research_______________________________113 References____________________________________________________________117 Appendices___________________________________________________________131 Appendix A: Summary of Research Questions________________________ 132 Appendix B: Expert Panel Members and Credentials___________________ 136 Appendix C: Expert Panel Instructions and Worksheets_________________ 137 Appendix D: IRB Exemption Certificate ______151 Appendix E: Pilot Study Reliability Summary_________________________153 Appendix F: IRB Certificate of Approval for URL______________________170 Appendix G: Electronic a nd Postal Notifications_______________________171 Appendix H: IRB Certificate of Approval for Incentive and Yahoo Account ________________________________________________184 Appendix I: Distribution of Respondents by SIS Membership_____________185 Appendix J: Bivariate Analyses Chi Square__________________________186 Appendix K: Summary of Logistic Regressions________________________187 Appendix L: Summary of Res ponses to Belief Questions_________________188
v Appendix M: Temporal Variables of Participants_______________________189 Appendix N: Summary of Logistic Regressions________________________190 Appendix O: Summary of Regression Diagnostics _____________________194 About the Author_________________________________________________End Page
vi List of Tables Table 1 Geographic Distribu tion of Sampling Frame______________________61 Table 2 Summary of Test for Norm ality for Continuous Variables___________67 Table 3 Comparison of Early and Late Responders_______________________77 Table 4 Summary of t Test of Predictor Variable Age for Early and Late Responders ________________________________________________78 Table 5 Summary of Sample Demographics_____________________________79 Table 6 Years in Clinical Practice_____________________________________80 Table 7 Distribution of Sampling Fr ame and Responders by Geographic Region____________________________________________________80 Table 8 Descriptors of Clinical Practice_______________________________ 81 Table 9 Practice Context by Ty pe of Setting and Ownership________________82 Table 10 Comparison of Demogra phic Characteristics: AOTA 2006 Workforce Study and Current Study on Effect of Personal and Practice Contexts__________________________________________________83 Table 11 Distribution of Therapists Who Almost Always Assess Cognition on Initial Evaluation by Choice of Method________________________85 Table 12 Frequency of Use of Inform al Assessment of Orientation and Standardized Screening and Assessment Instruments_______________86
vii Table 13 Important Characteristics in Therapists Choice of Assessment Instruments________________________________________________87 Table 14 Summary of Logistic Regr ession Analysis Predicting Use of Standardized Assessment Instruments___________________________94 Table 15 Assessment of Logistic Model_________________________________95 Table 16 Correlation Matrix for Predic tor Variables in Logistic Model ________96
viii Effect of Personal and Practice Contexts on Occupational Therapists Assessment Practices in Geriat ric Rehabilitation Mirtha Montejo Whaley ABSTRACT Despite considerable debate surrounding an age associated cognitive decline in nondemented elders, recent studies indicate that changes attribut able to normal aging affect cognitive processes and fluid abil ities. Additionally, studies i ndicate that factors such as physical illness, depression, neurological damage, medi cation side effects, drug interactions, and the effects of surgery and an esthesia may also cause varying degrees of cognitive impairment. Impairment of cognitiv e function is known to affect treatment and rehabilitation outcomes for older pe rsons, and increase their likelihood of institutionalization. Although proper screening and identificati on of cognitive deficits in geriatric patients are crucial in devel oping treatment plans, there is evidence in the medical and nursing literature that cognitive decline in older non-demented patients is often not identified. Proper screening in this case, refers not only to whether or not clinicians engage in assessment behavior, but that they adhere to evidence-based practices and
ix utilize standardized instruments which can id entify the type, extent and implications of the cognitive deficits. This study used an exploratory, non-e xperimental design a nd the population of interest consisted of occupational therapists providing phys ical rehabilitation to patients >65 years of age in the United States. The Ecological Systems Model was chosen as the theoretical framework, becau se it depicts human behavior as the product of the interaction between the indi viduals personal attributes and the physical and social environment in which the individual functions. Given the changes in health care, and the limits imposed by third party payers, it would s eem important to inquire as to the effect of personal and practice contexts on therap ists assessment practices in geriatric rehabilitation. Although results of the study indicate th at factors in the practice context are stronger predictors of therap ists use of standardized c ognitive screening and assessment instruments, the study supports principles of Ecological Systems Models in that both practice and personal contexts contribute to therapists assessment practices.
1 CHAPTER I Introduction The purpose of this study is to dete rmine personal and practice contexts in occupational therapy associated with the use of cognitive screening and assessment tools. The study specifically focuses on the use of su ch instruments by occupational therapists (OTs), with individuals 65 year s of age and older referred for physical rehabilitation. The need for this study is supported by factors such the research literature, individual and focus group interviews of OTs conducted by this researcher during a previous qualitative study and by the rese archers own experience as a geriatric rehabilitation therapist. Two important themes emerged from the focus groups regarding therapists use of cognitive screens or assessments with elderl y patients. The first theme addressed the influence of therapists personal factors (e.g. knowledge and beliefs) on their use of cognitive assessment instruments. The sec ond theme addressed the effect of practice factors (e.g. fixed assessment and treatme nt protocols; increased demands for productivity; cost-containment measures impos ed by third party payers) on therapists assessment practices. The therapists interviews and a review of the literature revealed a number of semantic inconsistencies that highlight the di fficulty involved in reaching consensus as to
2 what constitutes cognitive impairment in elderly rehabilitation patients, and what instruments should be used to determine leve l of impairment. While mild, moderate, and severe all denote degrees of cognitive impairment, the meaning of this classification depends on the instrument used to measure the impairment as well as which components of cognition are measured (Collie & Maru ff, 2002; Petersen et al. 2001). With regard to this researchers clinic al experience, approximately 80% of patients admitted to a rehabilitation f acility in Hillsborough County and assessed using the Large Allen Cognitive Assessment Test (LACL) between 1996 and 1999 were found to experience mild to moderate cognitive deficits at the time of initial evaluation. These deficits often had been missed by nursing and social service st aff using shortened versions of the Mini Mental St atus Questionnaire, because they did not present as deficits of memory or orientation, nor did the patients or their caregivers offer any complaints or awareness of the impairments. In fact, the pa tients difficulties were often attributed to either lack of motivation or obstinacy rather than to limited cognitive capacity. Mild cognitive deficits, according to Allens framework (1992; 1995), are those that affect problem-solving, correcting an e rror, anticipating and identifying a hazard, knowing when to report health problems or side effects of medi cations, and generally maintaining safety. Moderate deficits as described by this framework are those that, while not completely precluding performan ce of activities of daily living (ADLs), interfere with certain aspects of performa nce such as initiating and ending a task, sequencing through the steps of a familiar act ivity, and judging how much pressure to apply or how much quantity of something to use. These deficits are the result of disruptions of fluid processe s that are essential for the acquisition of new learning. The
3 impact of such deficits is pa rticularly important for elderl y persons experiencing the onset of an illness or following an injury, whic h require different ways of engaging in previously familiar tasks, closely following medical recommendations and/or observing safety precautions. The need to identify c ognitive deficits becomes crucial when the individual lives alone, particular ly if he or she experiences he alth conditions that require scheduling and managing prescr ibed medications, special di ets or medical regimens (MacNeill & Lichtenberg, 1997). While there is evidence in the nursing and me dical literature as to the effect of the practice environment on clinical decision making and other aspects of practice, this type of inquiry is not found in th e occupational therapy literatur e. To date, much of the research on clinical reasoning and clinical decision-making in occupational therapy has focused on internal processes, such as the therapists knowledge, be liefs, and attitudes (Mattingly, 1991; Penney, Kasar, & Sinay, 2001; Schell & Cervero, 1993; Unsworth, 2002). Focusing on personal factors alone fails to account for the interaction between the therapist and the envi ronment in which he or she pract ices. Therefore, conducting an inquiry into therapists sc reening and assessment behavi or requires a theoretical framework that takes into account personal fact ors, as well as fact ors in the external environment. This study is framed by the Ecological Systems Model, which explicates human behavior as the product of the inte raction between the individuals personal attributes and the physical and social envir onment in which the individual functions. Ecological models were in troduced to occupational therapy (OT) in the 1970s (Howe & Briggs, 1982) in an attempt to understand and improve patients functional
4 performance. However, this type of model also is also particularly well suited for an inquiry into OT practice in light of re cent changes in case mix and reimbursement affecting the duration and scope of services. Overview of Aging: A Historical Perspective During the 20th century, demographic changes resu lted in record growth in the number of persons 65 years of age and older living in the United States, increasing from 3.1 million to 35 million persons between 1900 and 2000 (Hobbs & Stoops, 2002). Within this group, the fastest rate of growth, six times th at of the general population, occurred among persons 85 years of age a nd older (Tideiksaar, 1997). More rapid growth among elders is predicted by the ye ar 2011, when the first wave of baby boomers will turn 65 years of age (Hobbs & Stoops). The elderly have changed the face of hea lth care and rehabilitation. The caseload, which through the 1970s consisted of a younge r population, shifted in the past 25 years to increasing numbers of elderly persons. Y ounger patients were more likely to require care for birth defects, acute c onditions, or injuries. Todays older rehabilitation patients, however, more commonly experience ch ronic conditions and comorbidity, which increase the complexity of treating any new or acute condition or occurrence. Theyre likely to present with functional deficits, to experience repeated hospitalizations, and to be at risk for institutionalization (Mille r, 2000). Additionally, older individuals may experience declines in cognition due to age-associated frontal lobe changes that are further affected by a number of factors including illness, traumatic stress, sleep disordered breathing, and medications (C ohen-Zion et al. 2004; Cohendy, Brougere, & Cuvillon, 2005; Raz, Rodrigue, & Acker, 2003; Stoner, 1997; Van Boxt el et al. 1998).
5 As the case mix has changed, so has the de livery of health services. Changes in reimbursement for health services over the past 10 to 15 years have impacted service provision by requiring reduced lengths of stay in hospitals and rehabi litation facilities, by emphasizing functional performance that genera lly translates to focusing interventions on regaining basic physical skills, and by requiri ng expeditious discharge to the least costly environments. Concomitantly, there has been an increased demand for patient-centered treatment, for accountability through both the collection and measurement of outcomes of care, and for the justification of health interventions through evidence-based practice (Hinojosa, Kramer, & Crist, 2005). The current health system, predicated on acute care, is driven by market pressures that impact the scope of occupational thera py services and the manner in which they are delivered. Howard (1991) noted that changes in reimbursement and the rise of managed care were redefining the practi ce, the management, and even the professional ethics of occupational therapy. As she further explai ned, both the frequency and nature of the treatment provided by occupational therapists has changed. Perh aps, this is most clearly demonstrated by the focus on utilizing diagnos is-based treatment protocols to ensure reimbursement for services (Howard, 1991). In practice, the current system of care focuses on ameliorating the presenting health problem in order to expedite discharge a nd places demands on practitioners across disciplines for increased productivity. The pa rameters of care imposed on providers by the health care system may indeed preclude occupational th erapists from identifying and addressing important components of function, and/or conditions that have a direct effect on the identified presenting problem. This limitation has the potential for increasing
6 morbidity and mortality and promoting age rela ted discrimination and health disparities. As such, it is in direct conf lict with the goals of Healthy People 2010 and should be cause for alarm to public health practitioners. The convergence of the demographic transi tion, changes in the delivery of services and reimbursement, and increased dema nd for patient-centered treatment and accountability has serious imp lications for health care in general, as well as for occupational therapy and rehabilitation, public health, and community resources. As the number and proportion of individua ls 65 and older increases in the future, there will be an increase in health services utilization, in creased demand on already limited resources, and an increased risk for both excess disability and age related health disparities. Cognition: Issues and Implicati ons for Occupational Therapy In clinical practice and across disciplines, the role of cognition is generally considered in the initial treat ment plan, primarily in cases where the diagnosis indicates the presence or likelihood of a cognitive impairment (e.g. Alzheimers disease, head injury, or stroke). In the course of treatment, cognitive status may be considered when problems arise as a result of the patients beha vior or when the indi vidual appears unable to acquire and retain new skills. Conversely, as noted by occupational therap ists in earlier interviews conducted by this researcher, assessment of cognitive f unction at initial evaluation is actually discouraged in some settings. These occupati onal therapists noted th at any indication of cognitive impairment could raise questions from third party payers as to the necessity and appropriateness of the individua ls referral to and particip ation in rehabilitation, and could ultimately lead to de nial of reimbursement.
7 Although the prospect of identifying cogniti ve decline in elders appears to be a monumental task, there are a number of natu rally occurring opportunities available to health care personnel to screen for cognitive defi cits in this population. For example, in primary care practice, physicians have been en couraged to assess their patients cognitive status during routine visits in an effort to identify those individuals experiencing memory deficits that may be caused by treatable conditions or are in dicative of early stages of Alzheimers disease. In rehabilitation settings, occupational th erapists, by virtue of their training and focus on function, have a unique opportunity to serve as gatekeepers. Assessing their patients cognitive status would allow therap ists to communicate with treating physicians as to the need for further testing and/or referr al to other professionals for identification of the underlying etiology and/or remediation. A ssessment would also allow therapists to engage in preventive interventions, by identifyi ng factors that place th eir patients at risk for adverse events, falls, injuries and non-compliance. Issues regarding the selection and proper use of assessment instruments by occupational therapists are neither new nor insi gnificant. Ina Elfant Asher (1996) in her annotated bibliography of occupational thera py assessment instruments, describes how in 1984 the Representative Assembly of the Am erican Occupational Therapy Association (AOTA) released a statement identifying as a top priority the development of standardized assessments for occupational therap y, and the utilization of such instruments by occupational therapists. In a subsequent document addressing the identified need, AOTA outlined 4 hierarchical competencies regarding therapists use of standardized assessments. Two of
8 the 4 hierarchical competencies charged users with the responsibility to a) recognize the importance of using standardized, reliable, and valid instruments when these are appropriate and b) distinguish the critical di fference between standardized and nonstandardized instruments ( Elfant Asher, 1996). These competencies are particularly pertinent to the findings of this study. Occupational therapists are directed by AOTAs Scope of Practice, which clearly defines the basis for assessment and interven tions. This document delineates the domains and processes of practice for occupational th erapists and assistan ts, and recognizes the importance of assessing components of func tion, including cognitive status (AOTA, 2002). Occupational therapy evaluation marks the beginning of the treatment process, and provides the foundation for the treatment pla n, as well goals indicated to meet the discharge needs of the patient and his or her caregivers. Assessment should be an inquiry that provides a snapshot of (AOTA, 2002; Hinojosa, Kramer, & Crist, 2005): the individuals wishes, plans, and needs related to discharge his/her occupational performance history remaining skills and level and type of assistance and support needed and available after discharge deficits that interfere or have the potential to interfere with functional performance the individuals understandi ng of the health event a nd of his or her current situation
9 The treatment plan and its goals reflect th e therapists best esti mate of the persons capacity to learn and acquire new skills. As part of the process of occupational therapy, assessment and treatment planning should be guided by clinical r easoning, taking into account those occupations that ar e significant for the individual and the roles that he or she occupies within the cont exts of his or her life (AOT A, 2002). Clinical reasoning should be grounded on knowledge of underlyi ng conditions and of limitations likely imposed by the presenting illness or injury a nd also should consider the impact of the individuals physical and so cial environments on his or her ability to function. Changes in occupational therapy in respons e to market pressures have shifted the practice from interventions based on a holistic paradigm, to interventions crafted within a more reductionistic functional/biomedical m odel (Howard, 1991). Within the biomedical model, rehabilitation goals are set taking into account the pa tients pre-morbid level of function; improvement of physical measures from admission to discharge; diagnosisrelated potential for improvement; and constraints imposed by third party payers (e.g. limited length of hospitalizations as dete rmined by the Diagnostic Related Groups, restrictions on services provi ded, and limits on post acute treatment as determined by the Prospective Payment System) (Howard, 1991). The underlying assumption in this model is that of intact cognitive abilities and, as such, it places the burden of assimilation, performance, and compliance on the individual patient. Often cognitive status is determined by a ssessing the patients orientation (to self, place, and time) and his or her ability to fo llow simple commands. Frequently, potential for rehabilitation is estimated by the individua ls ability to communicate, by self report regarding functional performance or by a brief observation of an act ivity of daily living
10 (ADL). However, orientation, following commands, and communication skills are not indicative of an individuals capacity to initiate and sequence him/herself through the steps of an activity, to problem solve or to anticipate, identify, or manage safety hazards (C. Allen, personal communication, September 1998). Similarly, ADLs are not an accurate meas ure of cognitive status, as they are crystallized abilities stored in procedural memory (C Allen, personal communication, September 1998; Ruchinskas, Singer, & Repe tz, 2000). This dissociation between functional status and cognitive abilities wa s supported in an earlier study by Galanos, Fillenbaum, Cohen, and Burchett in 1994. Th ese researchers found that although over 50% of their study participants experienced cognitive impairment, health problems and depression, they were still able to perf orm activities of daily living. A similar dissociation was reported by Ruschinskas et al (2000) in their study of the relationship between ambulation and cognitive abilities. Determining What to Assess: Clinic al Reasoning in Occupational Therapy Although several styles of c linical reasoning have been identified in occupational therapy research, they have been traditiona lly considered to correspond to one of two major categories. In their review of the l iterature, Schell and Cervero (1993) identified these two categories as: scientific reasoning suggesting a methodical, hypothesis-based, cognitive process narrative reasoning describing reflection-in-a ction, in the process of treatment, and serving to help th e therapist understa nd their patients experiences as well as to help patients develop a new future
11 These authors also report having found indi cations of an emergi ng third category of clinical reasoning that had not previously been acknowledged in inquiries about professional practice. This third category, identified as pragmatic reasoning parallels the process identified in cognitive psychology as situated cognition and explains a more complex method of reasoning (S chell & Cervero, 1993). Pragmatic reasoning and situated cognition share in common a belief in the effect of personal and practice contexts on mental activity (Schell & Cerv ero, 1993) and offer a different perspective as to how clinical decisions are made. As described by these authors, personal contexts includ e internal characteristics of the therapist, e.g. his or her values, motivation, knowledge, and available re pertoire and level of skills. Practice contexts exist in the therapists external environment and include the physical environment and its culture, as well as orga nizational, political, and economic factors that can both facilitate or inhibit therapists practices. Inquiring about the therapists personal and practice contexts may increase our unders tanding of the factors involved in their decisions regarding assessment and their ch oice of assessment in struments (Schell & Cervero). Statement of the Problem Although screening for cognitive problems is a reasonable step in developing occupational therapy treatment pl ans in geriatric rehabilitatio n (Barnes, Conner, Legault, Reznickova, & Harrison-Felix, 2004), there is ev idence in the literature that impaired cognition in non-demented elderl y patients is often not iden tified (Ruchinskas, 2002). This is particularly true for patients experiencing mild cogni tive deficits, especially when they retain adequate verbal and social skills.
12 Assessing a patients cognitive status dur ing the initial evaluation, allows the therapist to develop treatment goals based on the individual's capacity and his/her safe performance in areas of occupation (activitie s of daily living, instru mental activities of daily living, work, leisure, etc. The burden in this case is on the clinician to maximize performance and safety during treatment a nd upon discharge by teac hing or training the individual; by modifying the physical, tempor al, and social environment in order to facilitate functional performance; and by providing caregiver training and making appropriate recommendations. To date, much of the research on clinical reasoning and clinical decision-making in occupational therapy has focused on treatment decisions guided by internal processes such as the therapists knowledge, beliefs, and attitudes (Mattingly, 1991; Penney, Kasar, & Sinay, 2001; Schell & Cervero, 1993; Unsw orth, 2002). Little has been published as to other factors involved in occupational th erapists clinical decision-making regarding screening and assessment procedures, despite th e identified need to improve therapists recognition of cognitive deficits in elderl y non-demented patients (Ruchinskas, 2002; Knight, 2000). More significantly, there is a dearth of in formation as to the role of practice and personal contexts in the assessment process. Focusing on personal f actors alone fails to account for the interaction betw een the therapist and the envi ronment in which he or she practices. Therefore, conducting an inquiry into therapists scr eening and assessment behavior required a theoretical framework th at would take into account personal factors, as well as factors in the external environment that may have an effect on or contribute to their assessment behavior.
13 Need for the Study Because older persons are at increased risk for falls, injuries, and adverse events, they are also more likely to experience highe r rates of hospitalization, institutionalization (Tideiksaar, 1997), physical, and fu nctional declines associated with these events. As a result, older persons are also likely to be referred to occupational therapy for rehabilitation to improve or re gain their functional status. Within the next six years, the el derly population in the United States is projected to again reach an unprecedented growth when baby boomers begin to reach age 65 (Hobbs & Stoops, 2002). This projected growth will bring increased health care and Medicare/Medicaid expenditures resulting from concomitant increa ses in utilization of medical and occupational therapy services hospitalizations, and admissions to nursing homes. Adequate provision of services patient education, and prevention of unintentional injuries and adve rse events will require a cl ear picture of the patients capacity for functional performance, including their cognitive status. Under normal conditions, activities of daily living (ADLs) such as bathing, dressing, eating, walking, sitting and risi ng, are performed automatically, without conscious recall (C. Allen, personal communicat ion, September, 1998). These skills are over-learned or crystallized and stored in pr ocedural memory by virtue of the frequency with which they are performed and their longevity. However, when an individual experiences a new disability, or the exacer bation of a chronic condition, these daily activities may require the acquisition of new sk ills (new learning). In other words, the individual may have to lear n new ways of performing ADLs, and may have to do so while integrating adaptive strategies and using adaptive equipment and/or assistive
14 devices. Additionally, individuals may ha ve to learn to manage new medication regimens, observe dietary restrictio ns, and follow safety precautions. New learning, a function of the brains fr ontal lobe, requires th at fluid abilities allow the individual to engage in tasks wh ile managing a changing environment. New learning is dependent on the individuals ability to (Allen, Ea rhart, & Blue, 1992): attend to the task at hand process and catalogue new information store the information in long-term memory When learning occurs, the individual is ab le to apply the newly acquired skills to other situations by first recalling the info rmation from long-term memory and then planning a strategy to fit the ne w situation. This transfer of learning also requires that the individual be able to anticipa te the consequences of his or her actions, problem-solve to achieve the anticipated results, and then ev aluate the outcome for further adjustments (Allen et al. 1992). While mild and moderate degrees of cognitive impairment interfere with new learning, individuals experiencing such impa irment are still capable of acquiring new skills. Doing so, however, requires that su fficient time be allowed for situation-specific training, and that the physical environment and caregiving strategies be modified to support their functional performance and safety. Failure to take cognitive status into account and con tinuing to endorse a strictly functional/biomedical approach in occupationa l therapy carry a number of risks. If, for instance, the patients social and verbal ski lls mask deficits affecting performance and safety, his/her capacity may be overestimated. In this case, the individual may be unable
15 to attain treatment goals, frus trating both patient and therapis t and increasing the risk that the patients behavior will be interpreted as refusal to participate or cooperate. Ultimately, the patient may be labeled unmotivated, manipulative, or non-compliant and may be prematurely discharged from rehabilitation. There is also a risk that an individu al whose history, beha vior, or performance suggests cognitive impairment will be deemed as not having potential for rehabilitation (Barnes et al. 2004). In this case, patients may either not be re ferred to or may be discharged from rehabilitation services, incr easing their risk for excess disability and institutionalization. In either case, there are consequences in terms of the individual s quality of life and the fiscal burden on already limited resources. In the end, the indi viduals capacity for safe functional performance will not be taken into account and safety risks will neither be identified nor addressed, increasing the chan ces for non-compliance, adverse events, and repeated hospitalizations. There are fiscal and policy risks as we ll. Questions should be asked about the efficiency and cost effectiveness of a stri ctly functional approach, which can place the individual at risk for costly hospital read missions and institutiona lization. Concerns should also be raised about th e accuracy and validity of st rictly functional outcomes that are used to guide reimbursement decisions a nd that play a role in the development of aging policies without taking in to account an individuals co gnitive capacity (Challiner, Carpenter, Potter, & Maxwell, 2003). There are also other implic ations, as increasing concerns about medical errors raise new ethical and legal questions and signal a new kind of ri sk related to a functional
16 approach. Errors of omission (i.e. failing to identify risk factors such as a cognitive impairment or to provide need ed services) and errors of co mmission (creating unrealistic expectations of treatment or an unrealistic prognosis) can result in adverse events in the course of rehabilitation and following disc harge (B. Kornblau personal communication, August 2004; L. Andersen personal communi cation, November 2004; Scheirton, Mu, & Lowman, 2003). Just as there are risks associated with the failure to screen or assess the cognitive status of elderly rehabilitati on patients, there may also be concerns about conducting such procedures. An area that appears unexplored in occupational therapy practice is that of the ethical issues associated with assessing the cognitive status of non-demented elders. While identification of cognitive impairment may lead to denials for reimbursement from third party payers, less is known about the ethical implications of identifying such declines in non-demented elde rly and their capacity to consen t to medical procedures. This study explores the effect of prac tice and personal contexts on the use of cognitive assessments in geriatric occupationa l therapy. Understanding the impact of contexts may assist in determining the target and scope of interven tions needed to support patient assessment and improve the car e of older rehabilitation patients. Theoretical Model As a new occupation-based paradigm e volved over several years, a number of practice models closely related to General Systems Theory (GST) have been proposed. Known as ecological or contex tual models, these practice models almost unanimously emphasize the role of personal and external e nvironments or contexts on performance.
17 Although there are minor differences as to what exactly constitutes the environment, these models describe an ex ternal environment that includes physical, social, and cultural elements. As applied to the practice contexts of occupational therapists, these also include reimburseme nt policies and available equipment and resources. Values, knowledge, motivations, and repertoire of skills, constitute the individuals personal internal environment or context (Schell & Cervero, 1993). While these practice models primarily evolved in an effort to improve patient treatment by understanding the patient within his/her in ternal and external contexts, they are particularly suited to study the impact of contexts on therapists performance. A number of studies are found in the occ upational therapy literature focusing on the interaction between therapists and their contex ts and the influence of such interaction on the therapists clinical reasoning. Hoope r (1997) provided evidence supporting the influence of personal contexts on clinical reasoning. Conversely, in their studies on clinical reasoning, Lyons and Crepeau (2001) and Unsworth (2005) reported evidence supporting stronger influence of practice rather than persona l contexts on therapists and assistants clinical reasoning. Implications for Public Health The growth of the older population presen ts a special challenge to health care, rehabilitation, and public health. Failure to identify cognitive deficits will result in a number of missed opportunities fo r therapists working with elderly patients. As an example, conditions amenable to treatment and early stages of dementia may not be identified nor promptly treated. If cognitiv e deficits are not recognized, safety risks may not be identified or addressed, increasing the risks for non-complia nce, adverse events,
18 and unintentional injuries. Ultimately, if oppor tunities to identify cognitive deficits are missed, the risk of institutiona lization and over utilization of costly health and personal care services increases. Interventions and education provided w ithout a clear measure of the patients cognitive capacity assume that the info rmation will be assimilated, successfully processed, and properly utilized by the individual. To the extent that mild degrees of cognitive deficits in the el derly are not easily recognized by medical and rehabilitation personnel, health education messages, medi cal interventions, safety precautions and recommendations may be delivered ineffectively. Research Questions Four primary research questions guide this study: 1. What are the current practices of occupati onal therapists regardi ng screening/assessing the cognitive status of elderly pa tients referred for rehabilitation? a. Do therapists, on initial evaluation, routinely screen or assess the cognitive status of non-demented elderly patient s referred to rehabilitation? b. How frequently do therapists use cogni tive assessment instruments on initial evaluation of older re habilitation patients? c. Are therapists, in the course of treatmen t, likely to assess the cognitive status of patients who fail to improve as anticipate d in the initial evaluation and treatment plan? 2. What is the relationship between context (p ractice and personal) a nd therapists use of standardized instruments to assess the cogni tive status of older, non-demented medical patients referred for rehabilitation?
19 a. Is the type of practice setting (acute hospital, sub acute inpatient rehabilitation unit, home health agency, free-standing rehabilitation unit, and home health) associated with therapists use of cognitive screening/assessment instruments with non-demented elderly rehabilitati on patients on initial evaluation? b. Is there an association between faci lity ownership (non-profit, for profit, VA/military, individual contractor) and therapists use of cognitive screening/assessment instruments with elde rly non-demented patients during initial evaluation? c. Is there an association between prof essional autonomy afforded therapists through the facilities protocols and ther apists use of cognitive screening and assessment instruments? d. Is there an association between em ployer support, supervisor support and availability of resources, and therapists use of cognitive screening/assessment instruments? e. Is there a relationship between knowledge of the effect of aging on cognition and therapists use of screening/assessment instruments? f. Is there a relationship between therapis ts knowledge of how to administer and score a variety of screening/assessment instruments and their use of these instruments? g. Are therapists beliefs (about profe ssional responsibility, aging, or use of cognitive assessment instruments) associated with their use of cognitive screening and assessment instruments in geriatric rehabilitation?
20 h. Is there an association between tempor al characteristics of the therapists (e.g. age, years in occupational therapy practic e, years in geriatric rehabilitation, and length of time in employment at the time of the survey), and their use of cognitive screening/assessment instruments in geriatric rehabilitation? i. Is a therapists level of education associated with use of cognitive screening/assessment instruments? Delimitations The following delimitations were imposed by the researcher: Only occupational therapists (OTs) members of the American Occupational Therapy Association (AOTA), licensed or similarly credentialed by their states, were invited to participate in this study. Individuals were invited to partic ipate in this study, based on having designated their membership in either the Gerontologic (GSIS) or the Home and Community Health Special Interest Sections (HCHSIS) of AOTA. Limitations of the Study A number of limitations beyond the research ers control may prevent generalization to all occupational therapists practicing geriatric rehabilitation in the United States. Therapists who are members of th e American Occupational Therapy Association may differ from non-member therapists. Therapists who agree to particip ate in the study may differ from therapists who decline participation.
21 Successful notification was dependent on whether or not therapists were listed, had enrolled in the special inte rest sections listserves, read and responded to email postings. Assumptions Given that study participants were trained professionals graduates of occupational therapy programs, the following assumptions were made: Study participants would have knowledge of the interact ion between aging, cognition, and disease. Participants would have knowledge of a variety of theoretical perspectives applicable to occupational therapy pr actice, although they may choose to not guide their practice by a particular perspective. Participants would have knowledge of the administration and scoring of a number of assessment instruments, although they may choose to not use them in their clinical practice. Because direct observation of study participants in their practice environments was not feasible, it was further necessary to assume that participants provided honest re sponses in thei r self-reports.
22 Definitions 1. Activities of Daily Living (ADLs) self car e activities such as dressing, bathing, eating, ambulation, and toileti ng, which are part of an i ndividuals daily routine. 2. Assessment in occupational therapy, the process of determining an individuals remaining abilities and problem areas (e .g. muscle strength, range of motion, balance, coordination, cognitive abilities, etc.). 3. Assistive devices devices and/or equipment utilized in therapy to improve problem areas such as balance and ambulation, or to compensate for loss in range of motion, strength, manual dexterity, vision, or memory (e.g. canes, walkers, reachers, long handled self care equipment, weighted utensils, etc.) 4. Biomedical model a mechanistic m odel of care focusing on diagnosis and treatment of the pres enting physical problem. 5. Cognition mental processes that allow i ndividuals to attend to a task, problemsolve, remember, learn, etc. 6. Comorbidity the simultaneous presence of two or more physical illnesses. Generally refers to chronic conditions. 7. Crystallized abilities an individuals abilities based on attained knowledge. 8. Crystallized intelligence intelligence m easured by tests tapping into stored knowledge (e.g. the meaning of words or pr overbs, simple mathematical operations, etc.,) 9. Evaluation comprehensive process where pa tient data is gathered and interpreted to better understand the indi vidual, his or her situation, and the ecology of his or her performance. Ongoing process utilized to determine treatment interventions, to
23 assess the treatment process, and to de termine when interventions should be discontinued. 10. Executive function abilities related to fr ontal lobe function that control and manage other cognitive processes involv ed in processes such as planning, cognitive flexibility, abstract thinking rule acquisition, initiating appropriate actions and inhibiting inappropriate acti ons, and selecting re levant sensory information. 11. Fixed protocol a set format for conduc ting assessment and treatment procedures. Generally dictated by the type of functional measure utilized by the facility for the purpose of assessing and reporting treatment outcomes. Fixed protocols may limit therapists autonomy. 12. Flexible protocol a format for conduc ting assessment and treatment procedures, which affords the therapist the freedom to determine which functional areas and components to assess and treat. Flexib le protocols may allow therapists more autonomy. 13. Fluid abilities an individuals abil ity to reason and to solve problems in unfamiliar situations. Abilities that allow an individual to form concepts, reason, and identify similarities. 14. Fluid intelligence intelligence measured by mental tests requiring on the spot problem solving with unfamiliar materials and problems (e. g. determining what is missing from an unfamiliar drawing, etc.).
24 15. Frame of reference Principles guidi ng the practice of occupational therapy determined by the areas and problems addr essed by therapists and by the processes that therapists utilize to provi de services to their patients. 16. Functional approach approach in occ upational therapy and rehabilitation that focuses on the restoration of physical function such as ambulation and ADLs. 17. Functional performance an individuals ability to engage in activities of daily living such as ambulation, dressing, bathing, etc. 18. Occupational Performance History Client-centered measure developed by Kielhofner, Mallinson, Forsyth, and La i that focuses on the individuals occupational functioning, and his or her routines and habits (Elfant Asher, 1996). 19. Occupational Therapist (OT) therapists specialized on the restoration of function or the use of assistive devices to compensate for functional loss following illness or injury. OTs are approved to practice after completing an accredited program (baccalaureate level or above ) and satisfying internship requirements as well as successful completion of a national exam and state licensure or equivalent credentialing. 20. Occupational Therapy Assistant (OTA) an individual traine d at the associate degree level to implement treatment plans as determined by and under the supervision of an OT. OTAs are also required to successfully complete an internship and exam, and to obtain stat e licensure or equivalent credentialing. 21. Occupational Therapy profession involve d in the restoration of function or compensatory interventions for the purpose of allowing individuals of all ages to
25 engage in age appropriate occupations (devel opmental, physical, and social skills in children; home, community, voc ational, and avocational activities for adults, etc.) 22. Performance in areas of occupation (p erformance areas) broad categories of activities including se lf care, instrumental activities of daily living (shopping, use of transportation, money management, etc.), employment, and avocational activities. 23. Performance skills (components) elements such as sensorimotor, cognitive, perceptual and psychosocia l abilities that underlie performance in areas of occupation. 24. Personal context persona l attributes including knowle dge, values, beliefs, and attitudes, which according to ecological syst em models, determine the extent of an individuals involvement with a number of potential tasks available in his/her environment. 25. Polypharmacy use of multiple medications ge nerally as a result of the presence of comorbidity. 26. Practice context the environment surroundi ng practice that facilitates or limits an individuals performance. Practice c ontext includes the physical, social, and cultural environment; resources available to the individual; and social and cultural values. 27. Pragmatic reasoning mode of reasoning that takes into account both the personal and practice contexts.
26 28. Rehabilitation interventions provided by trained therapists for the purpose of restoring function, or compensating for loss of function as a result of illness, injury, or disability. 29. Screening tests brief procedures utilized in medicine, psychology, and rehabilitation to determine the need for more in depth probing (assessment) as to the presence and extent of impairment. 30. Situated cognition mode of reasoning that takes into account the effect of the situation and its meaning to th e individual on mental activity.
27 CHAPTER 2 Review of the Literature This chapter guides the reader through a review of the literature pertinent to the study. The review is organized into four prin cipal sections as follo ws: 1) an overview of cognition, factors affecting cognition (such as aging and disease), and a review of the literature regarding awareness among profession als of the role of cognition in functional performance; 2) a review of the geriat ric rehabilitation and occupational therapy literature, including OT practice and clinical reasoning in occupati onal therapy; 3) a review of the effect of pers onal and work contexts on medical and nursing practice; and 4) a review of the literature on Ecol ogical Systems Models of practice. Overview of cognition Cognition, a performance component, is one of several elements that play a fundamental role in an individuals ability to function. Cognition encompasses global abilities reflecting numerous and complex proc esses and involving different areas of the brain (Lamar, Zonderman, & Resnick, 2002). Cognitive processes organize and regulate human behavior and are essential to the perf ormance of any task or activity by allowing the individual to experience awareness, at tend and concentrate, recall, understand and learn, store information, make judgments a nd decisions, and problem-solve Cognition delimits individual abilities by determining wh at a person can do; c hoice and preferences
28 influence what a person will do; and the soci al and physical environment delimit what a person may do (Allen, Earhart, & Blue, 1995). While functional performance is the outco me of the intricate interaction between cognition, choice, and the social and physical environments, it is cognition that most significantly influences human functioning, because it provides the outermost boundaries of an individuals ability (Clark et al. 1991). Cognition allows the conscious mind to acquire and process information from the exte rnal environment so th at the individual may engage in motor activity. Thus, safety duri ng the performance of everyday activities is contingent on the individuals cognitive ability so that he or she may adequately process relevant sensory informa tion (Allen et al. 1992). Because humans rely on cognitive processe s to guide their behavior (Clark et al. 1991), a decline in cognitive ab ility may so influence motivation and choice as to endanger the individuals safet y, functional performance, and compliance. In essence, cognition has a direct bearing on the individuals ability to safely engage in age appropriate tasks and to acquire new skills throug hout the life course. In the presence of an injury, illness, or disability, cognitive processes permit the individual to learn adaptive strategies, to co mply with health edu cation information and medical treatment, and to observe safety precau tions. If in fact there is a relationship between normal aging and cognitive decline, and if this relationship is further influenced by chronic illness, comorbid conditions, and ot her factors, then early identification of cognitive deficits in older rehabilitation patients is of critical importance. Early screening of cognitive status is useful in iden tifying patients who can benefit from further neuropsychological or medical a ssessment to determine the cause and extent
29 of a cognitive decline. This would facil itate expeditious treatment of reversible conditions such as depression or delirium, which can adversely affect the individuals participation in rehabilita tion, potentially influencing both outcome and discharge disposition (Lenze et al. 2004; Nedley, Kendrick, & Brow n, 1995; Ruchinskas, 2002). Knowledge of cognitive status allows th e clinician to tailor patient education interventions and messages, thus improvi ng treatment efficiency and outcomes by determining the individuals capacity to acquire and apply new health behaviors; rehabilitate following illness or injury; safely engage in basic and in strumental activities of daily living; use adaptive equipment post rehabilitation; and comply with medical regimens designed to manage chronic condi tions and improve or maintain health. By first identifying the i ndividuals remaining abilities, therapists are able to engage their patients in interventions aimed at promoting the highest possible level of functional performance and safety. These interventions require patient training and modifications in the physical and social envir onments as well as a ppropriate caregiving strategies. According to Alle n et al. (1992), responsible th erapeutic interventions should include teaching others how to f acilitate the patients use of hi s or her remaining abilities, as well as increasing the caregivers awareness of behaviors or events that put the patient at risk for injury or complications Factors Affecting Cogniti on: Aging and Disease Over the years, researchers have de bated a number of fundamental questions regarding cognition and aging. Among these are questions as to whether or not there are cognitive changes associated with normal aging, and what cognitive functions these changes affect. While the discussion about cognitive aging has been considerable and
30 conclusions contradictory, recent stud ies propose hypotheses of age-related neurophysiologic changes that in terfere with cognitive proces ses and fluid abilities, both of which are essential for new learning to occur (Grigsby, Kaye, & Robbins, 1995). The relationship between aging and cognition has be en explored and demonstrated in studies utilizing data from community samples in the MacArthur studies on aging (Chodosh, Seeman, Keeler & Sewall et al. 2004). In spite of contradictory findings by B oone, Miller, Lesser, Hill, and D'Elia in 1991, there is new evidence from neuropsychol ogical research verifying age related physiologic changes in frontal lobe function. These changes are attri butable to a loss of neurons in and decreased blood supply to the frontal cortex of the brain (Grigsby et al. 1995) and account for declines in executive function. Results from a number of studies investig ating frontal lobe function have explained this type of cognitive decline as resulting from deficits in fluid intelligence, while asserting that crystallized intelligence remains relatively unaffected as individuals age (Barberger-Gateau & Fabri goule, 1997; Christensen, Jorm, Henderson, Mackinnon, & Korten, 1994; Kaufman, McMahon, & Becker 1989). This being the case, one would expect that knowledge and abil ity to perform routine tasks (crystallized abilities) would remain intact, while new learning and ability to engage in unfamiliar tasks (fluid abilities) would be compromised. This view was refuted by Tabbarah, Crimmins, and Seeman (2002) in their research based on data from the MacArthur St udies of Successful Aging. Tabbarah et als. study investigated the association be tween cognition and physical performance, focusing on differences between performance of routine and unfamiliar tasks that impose
31 increased attentional demands on the indivi dual. Tabbarahs study found an association between participants cognitive abilities, and changes in performance for both routine tasks (crystallized abilities) and unfamiliar tasks (fluid abilities). As such, these findings contradicted those of Barberger-Gateau a nd Fabrigoule (1997), and indicated that cognitive processes are in eff ect central to the performance of a variety of both routine and novel physical tasks (T abbarah et al. 2002). Electrophysiological studies conducted by Chao and Knight in 1997 provide evidence of frontal lobe changes in normal aging, which specifically result in impaired executive function of the attentional system. This is a particularly important finding in terms of safety and learning given the role of the attentional system in managing sensory input during task performance to facilitate cognitive processing. In this capacity, the attentional system serves as a sort of filte r, allowing the individual to focus on relevant characteristics of sensory stimuli in the envi ronment, while inhibiti ng those which are not relevant to the task at hand (Chao & Kn ight, 1997). Improper functioning of this inhibitory mechanism manifests cognitively and behaviorally as increased distractibility and precludes the individual from experien cing awareness and attending to the task (McDowd, Oseas-Kreger, & Filion, 1995). This ability to focus attention and suppress irrelevant stimulus is a necessary condition for humans to learn, because it permits the transfer information from short term (working) memory to long term memory for pe rmanent storage. Information thus stored can be recalled later to solve problems that arise in an environment that is constantly changing. When learning is compromised as the result of impaired cognition, new skills
32 can be acquired, but require adjustments in th e task, the physical and social environment, and/or the manner in which information is presented (Allen et al. 1992; 1995). Although the literature is replete with conflicting information regarding the cognitive status of elderly persons hospitaliz ed for acute medical events, several studies report an association between medical conditio ns and cognitive decline. For example, Garrett et al. (2004), in their study of vascul ar cognitive impairment identified cognitive problems associated with cardiac and mild cer ebrovascular disease in the absence of dementia. These cognitive problems include d reduced information processing speed, reduced cognitive flexibility, and poor lear ning efficiency (Garrett et al. 2004, p. 746; Kilander et al. 1998; Kirkpatrick & Jami eson, 1993; Waldstein et al. 1996). These results support findings from neur oimaging studies (Gar rett et al. 2004), indicating that, among individua ls experiencing mild forms of cardiovascular disease, there is a significant associ ation between the severity of their cognitive problems and neurologic changes resulting from vascular damage. Neuroimaging studies conducted by DeCarli et al. (2001) and Swan et al. ( 1998) revealed a strong relationship between blood pressure at mid-life, and subsequent development of white matter pathology in the brain. In terms of the prevalence and implica tions of vascular cognitive impairment, a study conducted by Rockwood et al. (2000) uti lizing data from the Canadian Study of Health and Aging revealed that the most prev alent form of vascular cognitive impairment (VCI) among their study subj ects was VCI with no dementia (Vascular CIND). From their study, the authors conclude d that VCI subjects were at higher risk than subjects without cognitive impairment for institutionalization and death, but experienced similar
33 risks when compared to subjects having Al zheimers dementia (Rockwood et al. 2000). Rockwood et al. cautioned against using crite ria for VCI that requires a diagnosis of dementia, as this tends to underestimate the prevalence of VCI, and minimizes the actual burden of cognitive impairment among individuals 65 years of age and older (Rockwood et al.). Research conducted by Elias (1998) and Ru chinskas, Broshek, Barth, Francis, and Robbins (2000) found evidence of cognitive ch anges associated with systemic illness, chronic diseases of the lung, heart, liver, or kidney, and polypharm acy associated with the treatment of these conditions. Studies of older patients undergoing surgical procedures for hip fracture estimate a 30 40 % incidence of cognitive problems post surgery in this popula tion (Herrick et al. 1996; Mast, M acNeill, & Lichtenberg, 1999). In research comparing healthy controls with pa tients diagnosed with peripheral vascular disease the estimated prevalence of frontal lobe dysfunction and attentional impairment was 25% (Rao, Jackson, & Howard, 1999). Gregg and colleagues (2000) in their pros pective cohort study of the effect of diabetes on the cognitive status of older wo men found that controlling for age, education, depression, and a number of comorbidities, di abetic women had lower MMSE scores at baseline than did women without diabetes. Th ese researchers also found the risk of major cognitive decline was greater (57% to 114% greater) for wo men who had been diabetic for over 15 years. Additionally, in a recent st udy of the association between hemoglobin levels and anemia and cognitive function am ong older medical patients, Zamboni et al. (2006) concluded that both conditions were independently associat ed with the cognitive performance of older medical patients treated in acute medical wards.
34 Additionally, studies in the cardiovascular literature provide insights into cognitive decline associated with procedures such as card iac bypass surgery (Seines, Goldsborough, Borowicz, & McKhann, 1999). Ot her studies point to the relationship between socioeconomic conditions, their imp act on the development of cognitive reserve in children, the implications for the child rens life course and for their risks of experiencing cognitive deficits in advanced age (Richards, Shipley, Fuhrer, & Wadswoth, 2004). The majority of these studies conclude th at because of the prevalence of cognitive deficits among older patients w ith medical conditions, there is a crucial need to assess the cognitive status of these indi viduals. These studies furthe r advocate for the development and use of cost effective and easy to admini ster tools capable of assessing executive function and fluid abilities, and sensitiv e to mild forms of cognitive decline. Cognition and Functional Status: Per ceptions of Health Professionals In an extensive examination of the lite rature on functional st atus, Knight (2000) reviewed publications from nursing, psyc hology, and medical databases from the 1960s through 1998 to explore how researchers and he alth professionals viewed the relationship between cognition and function. Her review, undertaken with the primary focus of identifying cognition as an impor tant variable related to func tional status revealed four primary categories of the relationship be tween the two (Knight, 2000, p. 1460). The categories, described below, were identifie d by the author as f unctional status as behavioral; cognitive functi on as a separate construct fr om functional st atus; cognitive function assumed via the measure chosen to measure functional status; and, cognitive function as one domain of ove rall functional status.
35 Functional status as behavi oral or performance-based Twenty four studies were iden tified by Knight (2000) viewing functional status as behavioral or performance-based, and although few in numbers, they were found across a number of professional disciplines and special ties. These studies utilized a variety of performance measures all based on ADL perfor mance. Six of the twenty four studies, according to Knight, acknowledged that othe r (psychological) factors may influence functional performance although none of the studies took cognition into account. Cognitive status as a separate c onstruct from functional status In terms of cognitive status as a separate construct from functional status the author cited 23 studies of which some interp ret cognition as underl ying functional ability, while some conclude there is an association between the two constructs. Knight (2000) comments as to the limitations of the instrume nts used in these studi es, and the lack of sensitivity in detecting cases other than t hose with diagnoses of dementia and other neurobehavioral conditions, which would be e xpected to interfere with the performance of ADLs. However, the author acknowledges th e significance of this group of studies, as they represent the beginning of an inquiry as to the relationship between the two variables. Cognitive function assumed thr ough instrumental activities The category of cognitive function assumed through instrumental activities is described in 21 studies in whic h, while not directly assessing cognitive status, the role of cognition is subsumed through the capacity to perform IADLs. Measures utilized in these studies included the Func tional Activities Questionnaire, the Direct Assessment of Functional Status, Duke Universitys Ol der Americans Resources and Services
36 Multidimensional Functional Assessment (OARS) the Duke UNC Health Profile, and the Sickness Impact Profile. The author cautio ns against use of these instruments with certain populations, because while they measur e whether or not functional tasks can be performed, they fail to iden tify other factors which may support or interfere with performance (Knight, 2000). One additional cri ticism the author did not include is that these measures rely on self report, which ha s been found to be inaccurate in many cases. Cognition as a dimension of functional status Knights last category revealed an additional 42 studies which account for cognition as a dimension of functional status. As an example, in occupational therapy, Allens work with cognitive levels (1985, 1992, 1995) and Fishers Assessment of Motor and Processing Skills (AMPS) acknowledge th e relationship between impaired cognition and functional performance at a ll levels, including the social and interpersonal levels. Studies in this category almost unanimously recommend the inclusion of assessments of memory, learning and problem-solving into f unctional assessments of elderly persons. Knight concludes her review by s upporting the recommendations that any measure of functional status should also addr ess cognitive, behavioral, and psychological components of function. She further suggest s closely looking at aspects of cognition such as attention, memory, and problem-solvi ng with an understanding that interventions for individuals with cognitive impairment must determine which of these components should be targeted for remediation (Knight, 2000). A review of the professional practice literature yielded a number of studies investigating awareness among h ealth professionals, and indi cated that physicians and nurses consistently underestimate the preval ence of cognitive impairment in elderly
37 patients ( Pisani, Redlich, McNicoll, Ely, & Inouye, 2003). In a subsequent study of recoverable cognitive dysfunc tion in older acute care pa tients, Inouye and colleagues (2006) concluded that a type of cognitive de cline, not characterized by dementia or delirium, is both prevalent in the targ et population and often undetected. These researchers proposed that older adults hospitali zed for acute illness be considered at risk for recoverable cognitive declin e and screened so that appr opriate interventions can be developed and implemented to tr eat this revers ible condition. Only one study was found addressing o ccupational and physical therapists failure to identify cognitive problems in their elderly patients (Ruchinskas, 2002). The dearth of information regarding identification of cognitive problems by rehabilitation therapists may well be a reflection of the therapists beliefs as to what constitutes functional status. Geriatric Rehabilitation There is growing evidence in the literature of a higher rate of cognitive decline among geriatric rehabilitation patients as comp ared to the general population of the same age. According to Ruchinskas and Curyto ( 2003), geriatric rehabil itation patients are at risk of experiencing cognitive impairment as a result of a number of factors, including age associated changes in attention and c ognitive processing. Ot her researchers have found associations between cognitive impa irment and medical conditions including chronic hypertension (Elias, 1998); chronic systemic and major organ disease and the polypharmacy associated with their treatmen t (Clarnette & Patters on, 1993; Ruchinskas et al. 2000); and surgical re placement of knee and hip join ts following traumatic fracture (Herrick et al. 1996; Mast et al. 1999).
38 Rao et al. (1999) reported that, contrary to earlier beliefs, as many as 25% of individuals having peripheral vascular disease demonstrate frontal lobe and attentional dysfunction. Similarly, research conducted by Mast et al. (1999), Ruchinskas, Singer, and Repetz (2000), and Tatemichi et al. (1994) indicated that one to two thirds of individuals with recent onset cerebrova scular accidents experienced cognitive dysfunction. Cognitive impairment has been linked to a number of adverse outcomes. For instance, one such outcome is a limited or lack of functional improvement in rehabilitation, because cognitive tasks such as memory, visuo-spatial skills, cognitive processing, and motor speed play an importa nt role in ADL performance (MacNeill & Lichtenberg, 1997). Additionally, because li ving alone requires successful performance of ADLs and the ability to e ngage in instrumental activitie s of daily living (IADLs, e.g. cooking, driving or using public trans portation, shopping, finances, adhering to medication regimens, and handling schedules and appointments), impaired cognition is also associated with increased risk of institutionalization following rehabilitation. In a study of 900 urban patients admitted to a geriatric rehabilitation unit between 1991 and 1994, MacNeill and Lichtenberg (1997) s ought to identify pred ictors of return to independent living. Patients in this study ra nged in age from 60 to 99 years of age, and had varied admitting diagnoses. Fifty three perc ent of these individuals had diagnoses of arthritis or peripheral-vascular disease; 33% had diagnoses of pelvic, hip, or leg fractures; and 13% had diagnoses of stroke. Results of the study indicated th at motor performance was not related to cognitive status and that in fact, motor scores may lead providers to overestimate an individuals capability to return to independent living.
39 While medical severity and demographic va riables did not contri bute to predicting discharge disposition in M acNeill and Lichtenbergs st udy, the authors did find an association between higher cognitive function at admission and being discharged to independent living. Findings from this st udy verify the importance of assessing the cognitive status of geriatric patients and emphasize that physi cal abilities alone are not an adequate measure of ability to return to independent living. Occupational Therapy Practice According to the most recent information from the Bureau of Labor Statistics, 92,000 occupational therapists were employed in 2004, most working in hospitals, with 1 in 10 therapists holding more than one job. Therapists were employed in a variety of settings, including physicians offices, home h ealth care services, out patient care centers, community care facilities for th e elderly, government agencies educational services, and nursing homes. A small number reported be ing in private practice, and providing services on referral from physicians or cons ultation to nursing ho mes, adult day care programs, and home health agencies (Bureau of Labor Statistics, 2006). As of April 2006, there were 102,000 licensed therapists in the United States as reported to the American Occupational Ther apy Association (AOTA) by the licensing boards of forty six states. Of this number, 24,000 were members of AOTA (Karen Bingham, Membership Director, AOTA, pe rsonal communication by phone, April 2006). Michigan, Indiana, and Hawaii do not curren tly require licensure as a condition of employment, but do have CE requirements in orde r to practice. Colorado is the only state that has no licensure or CE requirements (Kar en Smith, Associate Director of Regulatory Affairs, AOTA, personal communication by phone, April 2006).
40 Occupational therapists are health profe ssionals involved in the development and restoration of function in patie nts ranging from neonates to th e aged and provide services on a continuum that includes acute and post-acute stages of an illness or injury. In their work with neonates and toddlers, occupa tional therapists engage in preventive interventions when physical problems thre aten normal physical, cognitive, and/or psychosocial development. In their work w ith adults, occupational therapy interventions are designed to restore function and prevent the onset of disability resulting from a medical condition or injury and/or to redu ce the effects of a disabling condition. Within the domain of occupational ther apy, the concept of occupation describes those developmental tasks that are age appropr iate, essential for the individuals identity and ability to function and culturally mean ingful for the patient and the profession (Kielhofner, 1997). Occupation is a core construct that addresses tasks throughout the life-span and has provided the foundation for the profession since it s inception during the Moral Treatment movement in the late 1800s (Howard, 1991). The practice of occupational therapy, as in other professions, has been defined by the paradigm within which it functions, i.e., its perspective, values, roles, and tasks. Through role definition, shared meanings ar e generated that characterize the practice domain of therapists. Shared meanings, in tur n, clarify for therapists and others what the profession does, the population it serves, wh ich problems therapists address, and how services are provided (Kielhofner, 1997). However, shared meanings are not static entities; they are in fact an expression of the professions paradigm and, as paradigms change, so do shared meanings, values, ro les, and tasks. Thus, for the purpose of survival, a profession that at its core believed in the holis tic nature of the individual
41 responded to market pressures by adopting a redu ctionistic, biomedical framework. In so doing, it allowed the practice c ontext (e.g., corporate culture and reimbursement issues) to influence its roles and tasks (B aum, 1985; Foto, 1988; Howard, 1991). While national and state level leaders and some clinicians actively engaged these practice contexts (e.g. third party payers and pol icy makers) to ensure quality of services, individual therapists although frustrated adapted to ex ternal demands and, at least partially, redefined their personal contexts (e .g. values, beliefs, and attitudes) (Burke & Cassidy, 1991). As Burke and Cassidy (1991) indicate, emphasis on productivity, efficiency, and cost-containment has altered both the frequency and the ty pe of services provided by occupational therapists. These authors discuss how, in an effort to ensure reimbursement for services, therapists have had to pr ovide diagnosis-based treatment protocols regardless of whether or not they addressed the patients needs. One qualitative study by Walker (2000) inquired into the effect of managed health care on the practice of occupational thera py and the ways in which occupational therapists adapted to the ch anging health care environment in the 1990s. One adaptive technique Walker described was to become more businesslike and mostly focused on the economic value of occupational therapy. She reports that therapis ts in this category adapted by changing their pace and focusing on efficiency. Therapists aligned with the culture of managed care by ch anging their assessment practices, the interventions in which they engaged, and their documentation in order to comply with changing rules and expectations (Walker, 2000).
42 How external contexts affect the practic e of occupational therapy was discussed by Howard (1991), as she described the shif t towards the biomedical model and the emphasis on research for what she believed to be the wrong causes: Research, therefore, become s not just a measure of efficacy, but a method to ju stify occupational therapy according to the dominant model in health care practice. Because reimbursement rewards the unifactorial medical model, it becomes difficult to survive economically while clinging to a philosophy based on multifactorial causes of disease. Compromise by assimilating aspects of the medical model allows for survival, but limits options for social effectiveness (p. 879). The longevity of issues related to the e volution of occupational therapy in response to changes in the health care climate and reimbursement are ch ronicled by this literature review. The impact of cost-containment on the provision of services and the tension it created for therapists is evident as well. Clinical Reasoning in Occupational Therapy Initially described as fitti ng into one of two major cate gories, clinical reasoning in occupational therapy has undergone consider able scrutiny for over 20 years. One category has been described in a number of ways as technical rationality, scientific problem-solving, and instrumental reasoning, an d refers to a methodical, instrumental, hypothesis-generating scientific approach. A ccording to Rogers and Masagatani (1982) and Mattingly (1991), therapists engage in th is type of reasoning when they use the
43 medical diagnosis to guide their assessments and frame their treatment decisions. As instrumental reasoning, it is congruent w ith the biomedical model and is mostly concerned with prediction and control (Matti ngly, 1991). This assumes that knowing the medical diagnosis one can exp ect to find certain types of dysfunction, which respond to specific interventions. The second major category, described by Fleming (1991) and Mattingly (1991) as narrative reasoning, refers to a mode of thi nking in which therapists use narratives or stories when thinking about or discussing therapy with clie nts and caregivers or with other professionals. Narrativ e clinical reasoning promotes patient-centered interventions and is congruent with the models of occupati on that have emerged in the last few years. Through narratives, the patient and therapis t can create meaning from the illness or disabling event, and move on to create ne w futures incorporating the patients new situation (Mattingly, 1991; Sc hell and Cervero, 1993). In their 1993 review of the literature Schell and Cervero became aware of the emergence of a third category of clinical reasoning that had not been previously documented in occupational therapy research. This third category, identified as pragmatic reasoning is analogous to the process iden tified in cognitive psychology as situated cognition and serves to explain a more complex method of reasoning. Pragmatic reasoning and situated cogniti on take into account the effect of the particular situation on mental activity (Schell & Cervero, 1993), and the meaning afforded to the situation by the individual(s) Pragmatic reasoning is concerned with personal context, which as described by the authors, include the therapists values, motivation, knowledge, and availabl e repertoire of skills, and al so with external contexts,
44 which are those factors exte rnal to the individual and include the physical, social, cultural, and economic factors that facil itate or hinder the individuals behavior. Pragmatic reasoning is used by therapists whe n, in the process of treatment, they take their practice and personal c ontexts into account and consid er the impact that these factors have on potential inte rventions (Unsworth, 2005). It is possible that this latter form of reasoning had not been acknowledged in earlier research, because as Unsworth (2005), Sche ll and Cervero (1993) indicate, contextual factors were interpreted as ba rriers to clinical reasoning rather than being understood as part of the process in clinical decision-m aking. Pragmatic reasoning is congruent with the emerging paradigm in occupational ther apy based on ecological models of practice. Ecological Systems Model and Occupational Therapy Ecological models have b een proposed in occupational therapy literature since the 1970s. These models reflect the evolution of the profession, as it searched to develop a theoretical base that would assist in both de fining its similarities w ith, and distinctiveness from other health professions and public he alth (Howe & Briggs, 1982). Perhaps more importantly, these models signaled the evoluti on of an emerging paradigm fashioned after General Systems Theory. Ecological models provide a conceptual structure that views the individual as an open system, and explicate the reciprocal rela tionship of the individual with his or her environment, each contributing to and infl uencing the other (Howe & Briggs, 1982). Ecological models serve as a reminder to oc cupational therapists that the individuals performance cannot be considered without taki ng into account his or her environment,
45 which simultaneously affords opportunities fo r performance and constrains by pressing for particular behaviors. Environment or context, as defined by a number of researchers, is a common thread connecting several contemporary models in occupational ther apy. For instance, Kielhofner and Burke (1980) based the M odel of Human Occupation (MOHO) on GST, expanding the original model by adding s ubsystems of volition, habituation, and performance. Schkade and Schultz (1992) pr omoted a holistic appr oach to occupational therapy in which therapists assessments a nd interventions would take into account the role of the person, the environment, and th e interaction of the two. Dunn, Brown, and McGuigans (1994) Ecology of Human Perfor mance (EHP) was developed as a way to both recognize the role of context in tr eatment and as a way to improve patient treatment by facilitating collabo ration across disciplines (Dun n et al. 2003). Finally, and very closely related to th e Ecology of Human Performance, Christiansen and Baums (1997) Person-Environment Performance Mo del and Laws et al. (1996) PersonEnvironment Occupation Model focus on the relationship between performance and the interaction of the individual with his/her environment. Several of the models have been cited in a number of articles in the medical and rehabilitation literature. I ndividually, however, the Ecol ogy of Human Performance has been cited in 56 publications covering su ch diverse areas as assistive technology, disability and rehabilitation, occupational and physical ther apy, nursing and public health nursing, stroke, geriatric medi cine, psychology, psychiatry, science of human movement, and special education.
46 By way of explanation as to the critical need to consider the effect of environment in treatment, Dunn et al. (1994; 2003) propose the following two central assumptions of the EHP framework: the interaction between person and e nvironment affects human behavior and performance performance cannot be understood outside of context Dunn et al. (1994; 2003) further conte nd that while environment (external context) has been recognized as an im portant element of performance, occupational therapy practice has focused more on perf ormance in areas of occupation (e.g. ADLs, IADLs) and performance skills (cognitive, percep tual, and sensory skills) that reside in the personal context. In their chapter on the Ecological Model of Occupation, these authors further indicate that this lack of attention to contex t is not limited to occupational therapy, but has been noted in other huma n service professions (Dunn et al. 2003). Context, according to Dunn et al, (2003) refers to factors proximate to the individual and encompasses the physical, so cial and cultural environments, as well as temporal factors related to the individual (e.g. age, developmental stage, place in important life cycles). They further explain th at each individual has a distinctive but not exclusive contextual experience, because while each experience may be personal, contextual factors are shared with othe r individuals in the same space and time. Similarly, Howe and Briggs (1982) conceptu alize a system as comprised of nested layers with the centermost layer representi ng the inner life space (p ersonal context) of the individual, and encompassing the persons psychological, cognitive, and physiological dimensions. The outer layers, according to Howe and Briggs, represent the
47 individuals extended life space (external cont ext), which the authors describe as the space in which the person functions within his or her lifes roles, through the performance of life tasks (Howe and Bri ggs 1982, p. 323). Howe and Briggs (1982) immediate setting, defines the first e nvironmental layer, which includes home, neighborhood, family, and others who have regu lar contact with the individual. Within this layer, roles are principally related to family and community and activities mostly deal with personal care. Social networks comprise the sec ond environmental layer, which includes peers, schools, social groups, transpor tation, and social institutions ranging from health care to government. Activities and ro les in this layer cover a wide range and include the persons role as a worker. The th ird layer, defined by Howe and Briggs as the ideological layer, holds the societal and cultural values that instill meaning and motivation to the other layers. For the purpose of treatment, therapists w ould likely be concerned with interactions between their patients and the patients imme diate setting and social networks (Howe and Briggs, 1982). This study, however, focused on the interaction between the therapists personal factors within his or her inner life space (personal context) and the second (social) and third (ideological) layers as described in Ecological Systems Models. This study is supported by concerns expr essed by occupational therapy researchers such as Barris (1987), Fondiller et al. (1990) and Howard (1991) as to the effect of context on practice, as well as by others in th e medical literature. As an example, Landon et al. (2000), in a study of market influe nces on physicians practices surveyed 4,825 primary care physicians providing services to adult patients. These authors inquired as to the impact of physicians characteristics, pa tient factors, and ch aracteristics of the
48 practice setting and the organization on clini cal decisions. Their study revealed that personal characteristics of the physician and ch aracteristics of the pr actice setting (rather than organizational constraint s) were more predictive of physicians assessment and treatment behaviors (Landon et al. 2000). In question is the impact that external contexts may have on therapists beliefs, which ultimately determine how therapists c onduct their practice. Such inquiry is the basis for the proposed study and for the selectio n of an Ecological Systems Models as its framework.
49 CHAPTER 3 Methods This chapter describes the methods that were utilized in this study, and is organized into ten sections: 1) a de scription of the st udy, population, and sample selection; 2) research questions; 3) variables; 4) in strument development; 5) data collection; 6) response rates and representativ eness of the sample; 7) effect of social desirability; 8) data management; 9) data analyses; 10) power analysis and sample size. Type of Study, Population, and Sample Selection The study used an exploratory, non-expe rimental design and the population of interest consisted of occupational therapis ts (OTs) providing phys ical rehabilitation services to patients >65 years of age in the United States. The study used a purposive sample of o ccupational therapists (OTs) with primary enrollment in either the Gerontologic (G SIS) or the Home and Community Health Special Interest Sections (H CHSIS) of The American Occupational Therapy Association (AOTA). The two special interest sections we re selected as they were more likely to include therapists who provided physical reha bilitation services to persons 65 years of age and older. Names and addresses of 2000 occupational therapists with primary membership in the GSIS, and 1000 therapists with primary membership in the HCHSIS were obtained
50 from AOTA. Although therapists may enroll in several special in terest sections, AOTA limits primary membership to a single secti on. Accordingly, the chairpersons of the GSIS and the HCHSIS were contacted to obt ain their permission to post pre-notification announcements as well as subsequent reminde rs on their respective listserves. This contact was also important to garner the support of the ch airpersons and enlist their assistance in the proce ss of recruitment. Variables The outcome variable of interest in the study was the routine use of cognitive screening and assessment instruments by occ upational therapists on initial evaluation to assess the cognitive status of non-demented medical patients 65 years and older referred for rehabilitation These patients were selected ba sed on findings from prior interviews conducted by this researcher with occupationa l therapists and by the research literature, which indicated that therapists, as do other he alth professionals, tend to underestimate the prevalence of cognitive deficits among non-demented elderly medical patients (Ruchinskas & Curyto, 2003; Ruchinskas, 2002; Ruchinskas et al. 2000; Whaley, 2000). Thirty four predictor variables were initially identified based on constructs identified in the Ecological Systems Models and the research literat ure (Dunn et al. 2003; Howe and Briggs, 1982). Ecological Systems Mo dels provide a conceptual structure that views the individual as an ope n system with personal attri butes (personal contexts) in reciprocal interaction with hi s or her environment (external contexts). Fo r the purpose of this study, the external contexts comprise the occupational thera py practice environment including the physical and social work envi ronment, as well as patient factors.
51 Independent variables within the prac tice contexts, as described by Howe and Briggs (1982), inhabit the social networks and ideologic layer. While the social networks layer contains work, peers, social groups a nd social institutions, th e social and cultural values that infuse meaning to these networks are inherent in the ideologic layer. Although according to Howe and Briggs (1982), individuals participate in a number of activities and occupy a number of roles within thei r social networks, this study specifically focused on the therapists role as wo rkers. Thus, construc ts of the theoretical framework were operationalized through pr edictor variables based on the existing literature on Ecological Systems Models, as we ll as the research literature in nursing, medicine, and rehabilitation (Christi ansen and Baum, 1997; Dunn, Brown and McGuigan, 1994; Dunn et al. 2003; Law et al. 1996). Predictor variables asso ciated with each context were operationalized as follows: Personal context 1. Knowledge of the effect of ag ing on cognition (5 variables) 2. Knowledge gained from formal training in the administration and scoring of a variety of screening and assessment instrument 3. Beliefs regarding the use of cognitive screening and assessment instruments (6variables) 4. Beliefs regarding professional responsibility (2 variables) 5. Beliefs about aging (2 variables) 6. Age 7. Years in occupationa l therapy practice 8. Years in geriatri c rehabilitation
52 9. Years in employment with company 10. Gender 11. Race/ethnicity 12. Education (2 variables) 13. Special interest section Practice context Social layer 1. Type of work site or setting 2. Facility ownership Practice context Ideologic layer 1. Resources available to therapists 2. Supervisor support 3. Employer support 4. Professional autonomy afforded therapists (3 variables) 5. Patient factors A summary of the research questions, asso ciated variables, and survey questions is provided in Appendix A. Instrument Development Questions utilized in the initial version of the su rvey were derived from information in the research li terature (Dunn et al. 2003; Howe & Briggs, 1982) as well as from the focus groups and individual inte rviews conducted earlier with occupational therapists. The survey instrument was developed following recommendations from Dillman (2000), McDermott & Sarvela, (1999), and Thomas (2004).
53 Questions were designed according to Dillmans principles (2000) so that they would be applicable to each participant and worded in a manner that would clearly require a response. Additionally, as recomm ended by Dillman, questions were crafted so as to avoid excessive mental effort on the part of respondents; e.g., keeping recall simple and recent and providing response categories th at were carefully ordered, clear, and mutually exclusive (Dillman, 2000). There are a number of additional elements that contribute to creating a questionnaire that is resp ondent friendly, thereby improving the likelihood of a substantial response rate. In cluded in these elements are: clear and easy to understand ques tions ordered to indicate high salience to the respondent ; and a questionnaire layout in accordance with visual princi ples of design for comprehension and easy response (Dillman, 2000, p. 150). As such, the first question in the survey inquired as to the mo st recent clinical experience of respondents, and was designed to have high salience, to be non-threatening, and easy to answer. A skip patt ern was applied to this firs t question, to serve as a filter and ensure that only therapis ts carrying an active caseload at the time or having treated elderly patients within the previous 6 m onths would respond to the questionnaire. Visual appearance and accessibility are al so important concerns in designing web based surveys. Thus, as recommended by Dillman (2000) and Thomas (2004), graphics were avoided to prevent dist ractions, avoid leng thy download time, and ensure that the online questionnaire would display properly in different operat ing systems.
54 Question layout and background color we re selected to present a crisp and professional appearance, and a matrix layout was adopted for six items to maintain a reasonable length and avoid redundancy. Open-e nded responses were used sparingly as was use of a drop-down menu, which was limited to questions such as state of residence with many response options. Clear instruc tions were provided preceding a question or section when indicated, and a progress bar at the top of each page was utilized so that respondents would have an idea of ho w near completion they were. Maintaining a sense of context can be more difficult for respondents of web based questionnaires than it is for those responding to printed surveys. Therefore, following Dillmans recommendations, respondents were permitted to navigate back to previous questions much as they would if completi ng a paper and pencil questionnaire (Dillman, 2000). Survey features were also activated to allow individuals to resume responding in the event that their connection to the internet was disrupted or if they found it necessary to stop prior to completion. Several recommendations by McDermott a nd Sarvela (1999) were incorporated when assembling the questionn aire. Items that had sim ilar content (e.g. choice of instrument, frequency of use) or response options (e.g. multiple choice, true/false, or yes/no) were grouped together. Demographic items were placed together at the end of the questionnaire. Validity Validity, the most importan t consideration in survey de sign, refers to whether or not an instrument measures what it pur ports to measure and whether appropriate, meaningful, and useful inferences can be ma de from the obtained results (Ary, Jacobs, &
55 Razavieh, 2002; McDermott & Sarvela, 1999; Thomas, 2004). Face validity substantiates the appropriatene ss of the instrument for the target audience and that it measures the constructs of interest (McDer mott & Sarvela, 1999). Content validity is based on evidence obtained from a number of e xperts and attests to two important facts: 1) that each question addresses an objective of the study and 2) that questions provide sufficiently broad coverage to obtain meaningful information (Thomas, 2004). McDermott & Sarvela (1999) recommend establ ishing content validity when developing an instrument for data collection, particularly when the instrument will be utilized to measure social behavior. To verify face and content validity, a pa nel of experts was assembled and asked to review the survey instrument and all communica tions that would be pr ovided to potential participants (McDermott & Sarvela, 1999). The panel consisted of a University of South Florida (USF) College of Public Health facu lty member, a faculty member from the USF College of Nursing, a faculty member from the School of Occupational Therapy at the University of Florida, a masters prepared occupational therapist owner of a dementia specialty educational practice, two bachelors level occupational therapists practicing in geriatric rehabilitation, two Ph.D prepared researchers from the Patient Safety Center at the James A. Haley V.A. Medical Center in Tampa, and a Masters prepared finance specialist not familiar with occupational thera py practice. Inclusion of a panel member who was unfamiliar with occupational therapy co ntributed to an unbiased analysis of the constructs and questions (Thomas, 2004). A more complete list of panel members and their credentials is in cluded in Appendix B. A copy of the instructions provided the panel is provided in Appendix C.
56 These experts were instructed to review all documents for readability and clarity of instructions and to indicate if any items or wording might be offensive to potential participants. They were asked to review re sponse options to ensure these were adequate for the questions and that th e range of responses provide d was sufficient. Panel members were also asked to indicate if, in their estimation, any questions should be eliminated or included in the survey. In order to identify potential problems, panel members were provided the URL to the survey site and asked to comment on the appearance of the surv ey instrument, and on their experience navigating through the questions. They were also asked to indicate if they encountered any difficulties with accessi bility or display. Based on suggestions from several panel members and as reco mmended by Dillman (2000), respondents were not required to provide an answer as a condition for being allowed to respond to subsequent questions. Finally, using the classification appro ach to establish c ontent validity, panel members were provided with a form listing all 63 survey questions, a description of the objectives of the study, and a de scription of the th eoretical constructs. Panel members were then instructed to assign each survey que stion to a theoretical construct. Results obtained from the panel of experts were then reviewed to establish the representativeness and relevance of each item in order to determ ine what changes to the final instrument were indicated. Reliability Prior to conducting the p ilot study and, as required by the University of South Florida for the protection of human subjects in volved in social and behavioral research,
57 an Application for Initial Review was submitted to the Institutional Review Board (IRB). Because the involvement of human participants was restricted to the use of a survey questionnaire and subjects were not at risk of being identified nor subjected to any risk of liability or damage, an Exemption Certificat ion Request was also submitted. On October 5, 2005, an Exemption Certification for I RB Protocol #IRB 104071G was issued by the Institutional Review Board. A copy of this certificate is included in Appendix D. In order to pilot test an instrume nt, McDermott and Sarvela (1999) recommend identifying 20 to 50 subjects who are representa tive of the target population. Pilot tests are conducted in order to estab lish reliability, i.e., the degr ee of consistency with which the instrument measures whatever it intends to measure (Ari et al. 2002). The pilot test for this study was conducted to determine the test-retest stability of the instruments items by having therapists answer the su rvey questionnaire at two points in time (McDermott & Sarvela). Occupational therapists from the James A. Haley V. A. Medical Center in Tampa, Florida and Aegis Therapies in Florida, Alabama, and Arkansas were invited to participate in the pilot test of the in strument. As recommended by McDermott and Sarvela (1999), therapists were instructed to not take part in the research study if they agreed to participate in the pilot test. Potential participants were provided with information about the survey, a link to the in strument, and instructions as to how to access the survey site. An iden tical copy of the survey was created in a second Surveyor site, and the link to that site later provided to participants in a reminder message sent so as to allow at least 5 days to elapse between the first (T1) and second (T2) administration. Forty-four therapists responded to the survey at T1 and 22 responded at T2. A total of 18
58 matched surveys completed at both points we re identified as viable for the test-retest reliability analysis. Percent agreement and coefficient kappa between responses at each administration were calculated for all of the 21 dichotomous variables. Eighteen items with a kappa reliability coefficient of .50 or greater remain ed consistent 80% of the time or better and were retained for the final survey (Ari et al. 2002). Ten of those items, related to assessment instruments generally used in occupational therapy were changed from a matrix format to individual questions. The stability of ranked variables was ex amined using the Spearman rho correlation coefficient. A correlation coeffi cient of .30 or above was used as the criteria for retention of any item in the final survey. The rationale for that decision was based on a number of factors. Reliability estimates may be affect ed by sample size and reliability coefficients decrease as the homogeneity of the samp led group increases (Ari et al. 2002). Furthermore, decisions to retain or purge it ems should not be entirely based on reliability estimates but should also take into account the value of the question. Although items with a correlation coeffici ent below .30 could be interpreted as unstable, there are a number of plausible e xplanations that may account for such low correlations. It is possible that having res ponded to the survey once, therapists may have reconsidered their previous responses and replie d differently during the second administration. It is also po ssible that by responding once th ey may have been cued to the purpose of the inquiry, and subsequent re sponses may be reflective of the social desirability associated with the response.
59 Another possibility is that response options based on a four point scale may have offered limited choices and resulted in inflated differences. One last consideration is that despite the short period betw een administrations, therapis ts may have gained some awareness or knowledge or had some exposure, which could account for the variability in their responses. Results of the reliability anal ysis appear in a detail ed report presented in Appendix E. Data Collection The survey instrument was uploaded to a University of South Florida secure site using Ultimate Surveyor software. Ultimate Surveyor is an IIS Microsoft ASP based software, written in Visual Basic using secure socket layers for tr anslation, and storing the information behind firewalls. The inform ation between the participants computer and the server is encrypted so that it cannot be intercepted. To further ensure easy access to the surve y, several individuals were asked to access the site and test the questionnaire. This fi nal step revealed that manually entering the long and complex URL to the survey site was problematic and made access to the survey inconsistent, increasing the threat of non-re sponse. Although the URL was provided in both the letter of invitation and the electroni c message posted to the special interest sections listserves, there was a risk that onl y those therapists rece iving and reading the electronic notification would be able to access the questionnaire di rectly. Thus, as recommended by an expert panel member and to facilitate access and increase participation, a separate web page was crea ted through the University of South Floridas Health Information Technology Department, to se rve as a portal to the survey site. The letter of invitation and the el ectronic pre-notification messa ge were amended to include
60 the URL to the web site ( http://www.hsc.usf.edu/~mwhaley ), and approval to modify the study was obtained from IRB. A copy of the approval letter is included in Appendix F. Survey implementation followed Dillmans Tailored Design Method (2000), making several contacts with potential participants, utilizing first class mail for the prenotification letter and post-card reminder, and offering an incentive. A total of 6 postal and electronic contacts were made over a period of 6 weeks, beginning with the electronic pre-notification message posted on both AOTAs Gerontological and Home and Community Health Special In terest Sections listserves, advising therapists that they would be receiving letter s of invitation to participate in the study. The electronic pre-notification message was posted to the GSIS and HCHSIS listserves on April 17, 2006, and sent to member therapists via the respective listserves on April 18, 2006 (Appendix G). The electronic message was followed on April 25 by 3000 mailed letters of invitation. Table 1 shows a geographic distribution by region, of the therapists invited to participate in the study The letters of invitation included a descri ption of the study, an informed consent as required by the Institutional Revi ew Board at the University of South Florida (USF), and the link to the web page where the survey questionnaire could easily be accessed. A signed informed consent was not required. By responding to the surv ey, therapists were giving their consent to participate. On April 28, follow-up email thank you notes and reminders were posted to the GSIS and HCHSIS listserves, and on May 4, post card reminders were sent by mail, excluding 12 therapists whose letters of invita tion had been returned undeliverable as of that date. A second thank you note and email reminder was posted to the listserves on
61 May 16 and a final email reminder posted on Ma y 21 advising therapists that the data collection period had been extended through May 26. Table 1. Geographic Distribution of Sampling Frame Geographic Location Letters of invitation Sent Invitations by Region as % of Sampling Frame Region 1 Northeast 788 26.2 Region 2 Midwest 845 28.2 Region 3 South 753 25.1 Region 4 West 611 20.4 Puerto Rico 3 0.1 Total 3000 100.0 Other than multiple contacts with potential participants, one other factor associated with an increased response rate in survey res earch is the use of some form of incentive. Miller & Salkind (2002) discu ss how the use of incentives ha s been well researched in the response rate li terature on mailed questionnaires a nd, although there is inconclusive evidence as to how the value of the incentive impacts response rate, there appears to be agreement on two issues. One is that incentive s are effective in rais ing response rate in mailed survey research. The second issue is that incentives which are enclosed with a mailed survey are more effective than those th at are promised contingent on participation. The latter finding is congruent with Social Exchange Theory, which suggests that people are likely to engage in an action to redu ce what they perceive as an obligation.
62 Therefore, even a small incentive included with a mailed survey will likely persuade individuals to respond (Miller & Salkind, 2002). A promised incentive, according to Dillman (2000), becomes an economic rather than a social exchange and makes participation contingent on the perceived value of the incentive. Declining to participate in the latter type of exchange is also more culturally acceptable. Use of incentives with Web and Internet-b ased questionnaires pos es special logistic problems. As an example, determining how the incentive will be offered and delivered, especially if maintaining the confidentiality and anonymity of respondents is a concern. Thomas (2004) suggests offering a redeemable printable coupon which becomes visible at the time respondents submit the comp leted questionnaire. However, as Thomas indicates, the determination of whether to use an incentive with an electronic survey depends to some extent on the degree of mo tivation of potential pa rticipants. Thomas suggests that surveys addressi ng a subject that is meaningful and interesti ng to potential respondents may not require an incentive. O ffering to share the results with respondents, as well as appealing to pote ntial respondents altruism, ma y be sufficient motivation to promote participation (Thomas, 2004, p.123). The decision to use an incentive in this study, the manner in which potential respondents were notified of the incentive, and the means utili zed to deliver the incentive were borrowed from the literature of both maile d and electronic survey research. As an example, the letter and electronic postings appealed to the therapists sense of professional involvement and pr ofessional responsibility. Po tential partic ipants were advised that survey results would be shared with them through the USF website and that they would be notified through the SIS lists erves when results were available.
63 To protect the anonymity and confiden tiality of participants, a Yahoo email account was established where participants c ould provide their contact information to participate in the random drawing. Instruct ions as to how to access the Yahoo site and submit the information for the drawing was included on the last page of the questionnaire. As this represented a change to the original research plan, a modification to include an incentive was submitted to th e IRB and approval obtained to proceed with the study. A copy of the approval lette r is included in Appendix H. The incentives offered participants cons isted of two, $ 50 winners choice gift certificates. Two College of Public Health staff members who were not familiar with the study or with the roster of participating therapis ts, were recruited to as sist with this last phase. One individual was asked to create and maintain the Yahoo email account used for the drawing, while the second individual conducted the blind, ra ndom drawing at the conclusion of the data collection on May 26, 2006. A total of 248 of the 349 respondents participated in the random drawing. Response Rates As indicated in the literature, response rates for online surveys vary widely and are often lower than those for mailed questionnair es. Rates for 12 onlin e surveys of health professionals conducted between 1999 and 2002 ranged from 9% to 94% (Braithwaite, Emery, de Lusignan & Sutton, 2003). Response rates often depend on factors such as how the survey was deployed and the method for obtaining the sampling frame. Cook, Heath & Thompson (2000) cauti on against accepting response rates of published web based studies as the standard because studies with small response rates may not be submitted for publication, or may not be published if submitted. These
64 authors further assert that basing the ade quacy of response rates solely on published studies may lead to an overestimation of response rates. The literature also indicates that external validity, i.e. the repr esentativeness of the sample, rather than sample size is often mo re important to web ba sed and online survey research. And, in cases where the response ra te is small (<40%) every effort should be made to use other sources of data to compar e demographics of the survey respondents to those of the target population (Thomas, 2004). Our study followed Dillmans Total Design Method to maximize response rate, and yielded a response rate of 12%. In surv ey research, homogeneous samples (such as members of professional groups responding about issues that ar e salient to th eir practice) are less likely to present a source of bias with low response rates, such as the rate obtained in this study (Braithwaite et al. 2003; Cook et al. 2000; Tracy, Dantas, Moineddin & Upshur, 2005). Data Management The software package utilized for data management and analyses was SPSS version 15.0. Data were analyzed using logistic regr ession with a dichotomous outcome variable coded so that a value of 1 identified ther apists in the response category (therapists who almost always use any of the standardized instruments to assess cognition on initial evaluation ), and 0 would identify thos e in the reference category ( therapists who utilize an instrument frequently, sometimes, or almost never ). The almost always category was selected to ensure that the frequency of utilization of assessment instruments approximated routine use of an instrument in the process of initial evaluation.
65 Although levels of measurement for predicto r variables in a logi stic regression may vary from nominal to ratio, nominal (classifica tion) level predictor variables must first be transformed using dummy-coding or effect -coding (Munro, 2001). Thus, nominal variables were transformed into dichotomous variables as signing a value of 1 to the response category ( presence of the attribute or category of interest ) and 0 to the reference category ( absence of the attribute or category of interest ). Knowledge of assessment instruments wa s measured by therapists scores on a scale based on a selection of assessment instruments. The score for knowledge increased by one point for every instrument the therapist was trained to use. Data Analyses Prior to any analyses, the data were re viewed to ensure there were no duplicate cases. Univariate descriptive analyses were then conducted to screen the data for errors and to ensure that the data was read corre ctly by the computer program. Descriptive analyses are helpful to understand the shap e of the data and determine appropriate measures of central tendency, to detect marked departures from normality, to determine suitable statistical tests, and to answer rese arch questions (Hatcher & Stepanski, 1997). Statistics obtained with this procedure we re the mean (for age only), median, mode, standard deviation, skewness, and kurtosis. Based on results of th e univariate analysis, two variables, race/ethnicity and location of training were dropped from further analyses because their distributions were highly skew ed and variable transformations to achieve normality were impractical (Munro, 2001).
66 Missing Data An exploration of the data revealed th at missing responses for key variables ranged from .3 to 3% and only 1 predictor variable, frequency of use of the Cognitive Performance Test had 4% missing data. Because the percentage of missing data was small (<5%) and the pattern of missing data appe ared random, all variable s were retained. Additionally, since cases containing missing data would be excluded from pairwise comparisons in the analyses, all 303 cases were retained. Testing Data for Normality Tests of normality were conducted in order to determine suitable statistics to be used in subsequent analyses. Continuous variables (e.g. age, years in employment with company or agency, years in occupational therapy practice, years in geriatric rehabilitation ) were tested to determine if their distribution was significantly different from that which is found in a normally dist ributed population. Distribution of the data was tested using the Kolmogorov-Smirnov test of normality, which compared a set of scores in the sample to a normally distribu ted set of scores having the same mean and standard deviation (Field, 2001). The test considered significant with a p value < 0.05, yielded a p value of .003 for age, and .000 for th e other temporal vari ables. This value verified that the distributi on in question did demonstrate a significant departure from normality; therefore, the null hypothesis that the data was drawn from a normally distributed sample was rejected (Field, 2001; Hatcher & Stepanski, 1997). Table 2 summarizes the results of the normality test.
67 Kolmogorov-Smirnov Table 2. Summary of Test of Normality for Continuous Variables. Statistic df Sig. Age .067 295 .003* Years with company .278 295 .000** Years in OT practice .296 295 .000** Years in geriatric rehabilitation .190 295 .000** *Significant at p < .05, ** Significant at p <.001 Reliability analyses were undertaken in order to determine if variables could be combined in multiple-item add itive scales to reduce the number of variables. Cronbachs alpha coefficient was selected for this pro cedure and a criterion of >.70 established to identify items that belonged toge ther (Munro, 2001). Two items ( how therapists determine areas of performance to assess, and how therapists determine performance skills to assess ) with a Cronbachs alpha of .728 were found to be component measures of the scale Autonomy. All other items remain ed as individual variables for further analyses. Bivariate Analyses Potential confounders were identified through bivariate an alyses. All significance tests were two-tailed and based on a 0.05 level of significance. Chi-square ( ) tests of independence were conducted to estimate asso ciations between the outcome variable and ten categorical pred ictor variables ( gender education, setting, ownership, special interest
68 section, responders, percentage of caseload >65, years in OT pr actice, years in geriatric rehabilitation, years in employment ). The Chi-square, significan t at p <.05, tested the nu ll hypothesis of no significant association between two categorical variable s. Strength of th e association between variables was indicated by th e value of the Chi square, w ith larger values denoting stronger associations between the variables in the sample. For each of the bivariate associations the Chi square analyses provide d frequencies, sample size, missing data, pvalue, and level of significance. Spearman correlations were obtained between age and other temporal variables of the therapists ( years in practice, years in geriatric rehabilitation, years with company ) to explore the associations between these vari ables. Spearman corre lation was selected, because it is a non-parametric st atistic suitable for data that did not meet the assumption of normality. The effects of twenty th ree variables addressing know ledge, beliefs and support on the outcome variable were explored through logistic regressions controlling for age, gender and education. Based on these analyses thirteen pr edictor variables displaying significant (p<.05) associations with the outco me variable were reta ined for the larger multivariate analysis. Non-responders Non-response poses a threat to the validity of a study (non-response bias), because of the risk that there may be importa nt differences between responders and nonresponders on crucial aspects of the study. For that reason, it is recommended that all feasible measures be taken to reduce non-re sponse rates and that information about non-
69 responders be obtained in order to discern si gnificant differences betw een the two groups. When information about non-responders is not available, as was the case in this study, it is possible to make inferences about this group through extrapolat ion. This method is utilized by researchers who c onceptualize non-response as part of a continuum ranging from early responders through non-responders (Center for Survey Research, 2004; Walonick, 2004). Extrapolation is based on the assump tion that non-responders have more in common with late responders than with those who participate earlier in the process, and that late and non-responders would perform similarly with regards to the outcome of interest. Thus, a dichotomous variable ( Responders ) was created and therapists assigned to one of two groups based on their date of pa rticipation. Therapists who participated in the study through May 18 were classified as early responders (coded 1) and those responding between May 19 and May 26 were classified as late responders (coded 0). This cutoff date was designated because it ma rked the beginning of a one-week extension for data collection, and followed one last electr onic reminder in an effort to recruit more participants. External validity of the sample The external validity of a ny study is enhanced by the degree to which the sample is representative of the population of interest Additionally, in studies with low response rates, response bias can be minimized if the study sample is representative of the population of interest. To determine the representativeness of the study sample, demographic data from the current study was compared with data from the 2006 AOTA Workforce and Compensation Survey. The cross-sectional study conducted by AOTA
70 had a sample size of 3003 respondents. Pa rticipants included members and non-members of AOTA, practicing in a variet y of settings with all age gr oups and disabilities. This provided a comprehensive view of the profession (AOTA, 2006). Effect of social desirability Five items from the Marlowe-Crowne Social Desirability Scale (Crowne & Marlowe, 1960) were incorporated in the survey to discern if social desirability had an effect on therapists responses. To obtain a social desirability score according to the scale instructions, three items ( I have given up doing something because I doubted my ability; I have felt like rebelling against people in authority; it is sometimes hard to go on with my work if Im not encouraged ) were reversed scaled and the sum of the five items obtained. This summated scale had values ranging from 0 to 5, with higher values indicating increasingly higher so cially desirable responses. Multivariate Analysis Logistic regression predicts the probab ility of an outcome occurring, given known values of one or more predictor variables. This procedure is flexible, does not assume that predictor variables are normally distribut ed, and is easy to interpret (Moss, Wellman, & Cotsonis, 2003). Logistic regression does re quire a dichotomous outcome variable, and assumes that observations are independent of each other (Tabachnick & Fidell, 2001). Predictor variables were selected for the logistic regression based on results obtained in the bivariate analyses. The outcome variable was coded so that success, i.e. almost always using an assessment instrument, would be coded 1, and not almost always using an assessment instrument coded 0. Categorical predic tors, with the exception of gender, were coded for the logi stic regression analysis assign ing 1 to positive attributes
71 such as more knowledge, more experience, a nd more positive beliefs, and 0 otherwise. Gender was coded 0 for females and 1 for males. A thirteen predictor logistic model was then fitted to the data to answer research questions regarding the relationship between therapists use of assessment instruments and his or her practice and personal contex ts. The backward elimination method was selected as this was an exploratory procedure seeking the best fitting model for the data. Backward elimination is useful in the absen ce of theoretical research and for exploratory model building (Field, 2001). In the backward method, the model begi ns with all predictor variables included, and the software program then assesses the contribution of each predictor against the criterion value for removal. If a variable is not making a significant contribution to the to the models ability to predict the outcome, th at variable is removed, and both the model and the remaining variables reassessed (Moss, Wellman, & Cotsonis, 2003; Field, 2001). This process is repeated until the most parsimonious array of relevant predictors for the full regression model is obtained. Eight models were tested by SPSS to determ ine the best fitting model for the data. The final model consisted of six predictor va riables, two from the practice context, and four from the personal context. The logi stic model was assessed by examining tests results of the logistic regression, statistical significance of predic tor variables in the model and results of the Homer and Lemeshow goodness-of-fit test Regression diagnostics in cluding Cooks Statistic, le verage, studentized and standardized residuals, devian ce, and DFBetas were conducte d to determine whether the model fit the data well or was influenced by a small number of cases, and to ascertain if
72 the model could generalize to other samples. In an average sample with a normal distribution, 95% of the standa rdized residuals sh ould lie between -2 and +2 and 99% should lie between -2.5 and 2.5. Therefore, st andardized residuals greater than 2.5 in more than 1% of the sample should be examin ed to determine the le vel of error in the sample. Residual statistics were thus examined to determine if extreme cases were exerting undue influence on the model. Multicollinearity in the model was assessed with a pr eliminary review of the correlation matrix, followed by a linear regres sion producing VIF and tolerance values. Collinearity in the data would be detected if correlat ion coefficients were >.60, VIF values >10, an average VIF considerably gr eater than 1, and tole rance values < 0.2. Tabachnick and Fidell (2001) recommend a minimum case-to-predictor ratio of 10 to 1, with a minimum sample size of 100 cases in order to ensure an adequate sample size for the data analysis. In this study, after cases with missing data were eliminated listwise, the sample size for the logistic regression was 271, the number of predictors 13, and the case-to-predictor ratio 20.8 to 1. Power Analysis and Sample Size Cohens power analysis formula was applie d to determine the sample size needed to detect an effect in this study. The formula used Cohens effect size ( R ), an effect size index ( L = 17.8) based on 13 predictor variables ( u ), an alpha (significance) level = .05 and power = .80 (Cohen, 1988; Munro, 1997). N = L (1 R ) + u + 1 R
73 Based on Cohens formula, a sample size ( N ) of 1572 would detect a small effect in this study ( R = 0.02); N =226 would be sufficient to de tect a moderate effect ( R = 0.13); and N =88 would detect a large effect ( R = 0.30). The study sample was deemed to provide sufficient power to de tect a medium size effect, i.e. to reject a false null hypothesis.
74 CHAPTER 4 Results This chapter discusses the results of th e data analyses and is divided into six sections: 1) overview of the study; 2) descri ption of non-responders; 3) description of the study sample; 4) external validity of the sample; 5) effect of social desirability; 6) results for each research question. Overview of the Study The study was designed to explore the assessment practices of occupational therapists engaged in the physic al rehabilitation of patients 65 years of age and older. Specifically, the study inquired as to the a ssociation of the therapists personal and practice contexts with their use of cognitiv e screening and assessment instruments with older non-demented medical patients. The theo retical framework uti lized for this study was the Ecological Systems Model. The study sample was obtained from the memb ership rosters of tw o special interest sections of AOTA. Three thousand occupati onal therapists throughout the United States were invited to participate in the study. Da ta were collected by means of a web based survey questionnaire and analyzed ut ilizing SPSS version 15.0 for Windows. Therapists were instructed to exclude fr om consideration patients with diagnoses of stroke, Alzheimers disease, and head injury in their responses. These conditions were
75 omitted because they are diagnostically asso ciated with signifi cant impairment in cognitive function, and therefore more likely to prompt therap ists to screen for cognitive deficits or to assess the patients cognitive status. Conversely, cognitive decline in nondemented medical patients is often not id entified and can result in less than optimal rehabilitation outcomes and missed opportunitie s to address reversible conditions. If cognitive impairment is identified, it could be used to provide early intervention to patients who may be at risk for developing dementia. The study sought to answer three research questions. 1. What are the current practices of occupationa l therapists regarding assessment of the cognitive status of older, non-demented medical patients referred for physical rehabilitation? 2. What effect do the therapists contexts (p ractice and personal) have on therapists utilization of standardized screen ing and assessment instruments? 3. Which factors of the therapists contexts (p ractice and/or personal) are predictive of their utilization of standardiz ed assessment instruments? Description of the Case Selection Process Of three hundred forty nine therapists who participated in the study, 27 were excluded because they were in academic or managerial positions and had not been involved in clinical practice for 6 months or longer prio r to the survey. Nineteen additional respondents also failed to meet the inclusion criteria as they either did not identify their special interest section (SIS), or they indi cated their primary membership was in other than the Gerontological (GSI S) or the Home and Community Health SIS (HCHSIS). The remaining study sample of 303 therapists resulted in a response rate of
76 12% and included 208 therapists (68.6%) with primary membership in the GSIS and 95 therapists (31.4%) with primary membersh ip in the HCHSIS. A summary of all respondents by SIS membership is presented in Appendix I Description of Non-Responders Six therapists contacted the principal i nvestigator during the data collection and cited reasons for not par ticipating in the study, incl uding being retired (N=1), involvement in other than clinical practice fo r over a year (N=2), survey questions not applicable to a non-traditional practice (N=1), therapist in a variety of settings abroad (N=1), and forgetting to reply prior to closi ng of the survey site (N=1). Three other potential responders requested assistance accessing the survey site, but only one acknowledged having successfully co mpleted the questionnaire. Because a follow-up study of non-responders was not feasible, characteristics of non-responders were estimated by extrapol ation as described in chapter 3. Early responders (N=255) participated in the study through May 18, and late responders participated between May 19 and May 26. Th is cutoff date was designated because it followed the last electronic reminder and a llowed for a one week extension for data collection. As illustrated in Table 3, late responders te nded to be older, with a mean age of 46.2 years compared to 42 years for early res ponders. As with early responders, a higher proportion of late responders had primary memberships in the Gerontological SIS (64.6%) compared to membership in the Home and Community Health SIS (35.4%). A higher percentage of males were in the la te responder group (12.5 %) compared to early male responders (7.9%). In terms of educat ion, baccalaureate and master level therapists
77 showed similar early and late responder distributions. All P h.D prepared therapists were in the early responder group. Table 3 Comparison of Early and Late Responders (N=303) Variable Total Sample Early Responders (84.2%) Late Responders (15.8%) Age Range 24-75 25-76 Mean 42 46.2 Median 42 46 Mode 45 52 SIS Gerontological 208 (68.6%) 177 (69.4%) 31 (64.6%) Home /Community 95 (31.4%) 78 (30.5%) 17 (35.4%) Missing 0 Education Baccalaureate 170 (56.1%) 143 (56.0%) 27 (56.3%) Master's 119 (39.3%) 98 (38.4%) 21 (43.8%) Doctoral 12 (4.0%) 12 ( 4.7%) 0 Missing 2 (0.6%) Gender Male 26 (8.6%) 20 ( 7.9%) 6 (12.5%) Female 275 (90.8%) 233 (92.1%) 42 (87.5%) Missing 2 (0.6%) A t test conducted to compare the age means of early and late responders indicated that late responders were si gnificantly older than early re sponders. A summary of the t test is provided in Table 4. A Chi square test of independence wa s conducted to determine if there were significant differences between early and la te responders on the outcome variable, in order to make further inferences about non-re sponders. A p >.05 revealed there were no significant differences on the outcome variable between the two groups. By extrapolation, this would i ndicate that non-responders would likely have performed similarly with regards to the outcome of interest.
78 p <.05 Description of the Study Sample Demographic Characteristics of Participants Participating therapists ranged in age fr om 24 to 76 years with a mean age of 42.7, a median age of 42.5 years, a mode of 45 year s, a standard deviation of 11.1, skewness of 0.163, and kurtosis of -0.731. Approximately 91% of the sample was female, and 56.5% of the sample had a bachelors degree; 1.3% percent had obtained Bo ard Certification in Gerontology, 2.6% held a Cer tificate in Gerontology from a university OT program, and 94.3% did not have any special ce rtification. Table 5 provides a detailed summary of the demographic characteri stics of the sample. Table 4 Summary of t Test of Predictor Variable Age for Early and Late Responders Responders Age (24-76) M (SD) Early (n=250) 41.98 (11.149) Late (n=48) 46.19 (10.346) t = -2.419*
79 Table 5 Summary of Sample Demographics ( N = 303) Characteristic Frequency Percent Gender 301 Female 275 91.4 Male 26 8.6 Race and ethnicity 295 Caucasian 276 93.6 African American 6 2.0 Hispanic 6 2.0 Alaskan-Pacific Islander 7 2.4 Highest level of education attained 301 Bachelor 170 56.5 Master 119 39.5 Doctoral 12 4.0 Location of professional training 303 Trained in U.S. 294 97.0 Foreign Trained 9 3.0 Special certification 298 Board Certification in Gerontology 4 1.3 Board Certification in Neurology 4 1.3 Board Certification in Pediatrics 1 0.3 Univer sity OT Certificate in Gerontology 8 2.6 None 281 94.3 Description of the Clinical Practice Almost half of the therapists in th e sample (47.4%) had be en in occupational therapy practice over 15 years and 1% had prac ticed one year or less. Approximately one third (30.6%) reported having 15 years or more of geriatric experience, while slightly more than 50% had 10 years or less of geriat ric experience. Table 6 summarizes the descriptive statistics for the temporal variables of the clinical practice.
80 Table 6 Years in Clinical Practice Variable Range* Median Mode Skewness Kurtosis Years with company <1 to >10 2-5 years 2-5 years 0.307 -0.933 Years in Occupational Therapy practice <1 to >15 11-15 years >15 years -0.423 -1.393 Years in geriatric rehabilitation <1 to >15 6-10 years >15 years -0.106 -1.335 *Mean scores not obtained as th ey were not suitable for ranges Geographic distribution of par ticipants was fairly consistent with that of the letters of invitation. Table 7 compares the geogra phic distribution of the sampling frame with the responders. Table 7 Distribution for Sampling Frame and Responders by Geographic Region Geographic Location Letters of invitation Sent Number of Letters Returned as Undeliverable Study Sample Responses Region 1 Northeast 788 (26/2%) 12 85 (28.1%) Region 2 Midwest 845 (28.2%) 11 91 (30.0%) Region 3 South 753 (25.1%) 13 63 (20.8%) Region 4 West 611 (20.4%) 7 61 (20.1%) Puerto Rico 3 (0.1%) 0 0 Missing 3 (1.0%) Total 3000 (100%) 43 303 (100%)
81 Therapists were asked whether they were so lely engaged in clini cal practice or held positions which, while primarily non-clinical, also carried a patient caseload. Two hundred forty-seven participan ts (81.5%) reported being predominantly engaged in clinical practice. Almost 80% of the study sample indicated that over 75% of their caseload consisted of patients 65 years of ag e and older. Seventy seven percent of participants reported a treatment caseload of twelve patients or less. Only twenty six therapists (8.6%) reported util izing one particular theoretical perspective to guide their clinical practice. A descript ion of the clinical practice ha s been summarized in Table 8. Table 8 Descriptors of Clinical Practice N = 303 Characteristic Frequency Percent Percent of patients >65 years of age in caseload 0 25% 3 1.0 26 50% 11 3.6 51 75% 48 15.8 > 76% 241 79.5 Number of patients in caseload <9 126 42.0 10-12 106 35.3 13-15 33 11.0 >15 35 11.7 Use theoretical perspective to guide practice Yes 26 8.6 No 277 92.4 Over half of the sample (51.5%) was em ployed by for profit companies; 35.3% was employed by non-profit companies. Forty th ree percent of responde nts were based in skilled nursing facilities, while 29.7% reported working in ho me health. A full summary of the practice setting is included in Table 9.
82 Table 9 Practice Context by Type of Setting and Ownership (N=303) Frequency Percent Setting Acute hospital 25 8.2 Subacute Inpatient Rehabilitation 18 5.9 Skilled Nursing Facility 130 42.9 Outpatient Rehabilitation 10 3.3 Assisted Living Facility 6 2.0 Home Health 90 29.7 Other 23 7.6 Missing 1 0.4 Ownership For profit 156 51.5 Non-profit 107 35.3 VA or military 7 2.3 Independent contractor 18 5.9 Other 11 3.6 Missing 4 1.3 External Validity of the Sample To determine whether the study sample was representative of the population, data from the current study was compared to results obtained from the 2006 American Occupational Therapy Association (AOTA) Workforce and Compensation Survey. The AOTA national survey was mailed to 8998 member and non-member therapists, practicing in a variety of sett ings with all age groups and disabilities, which provided a comprehensive view of the profession. Demographic data from the current study on the effect of context on therapists assessments practices revealed similar tren ds to the AOTA survey. Results of both studies are compared in Tabl e 10, showing similar distributi ons for median age, gender, and education.
83 Table 10 Comparison of Demographic Characteristics: AOTA 2006 Workforce Survey and Current Study on Effect of Personal and Practice Contexts AOTA Survey 2006 (N=3003) Current Study (N=303) Overall Sample AOTA Members Nonmembers AOTA Members Gender Female 95% 96% 94% 91.40% Male 5% 4% 6% 8.60% Median Age 42 44 36 42.5 Education Baccalaureate 63.80% N/A N/A 56.50% Ma ster's Degree 31.90% N/A N/A 39.50% Doctoral Degree 2.40% N/A N/A 4% Median Years Experience 13 15 9.5 11-15 (range) Race/ethnicity African American 1.4% 2% 2% American-Indian, Alaska Native, Asian-Pacific Islander 7.9% 7.2% 2.4% Ca ucasian 88.3% 81.4% 93.6% Hispanic 1.5% 2.4% 2.0% Multi-ethnic 1.1% 1.7% N/A No response 3.8% 5.3% 2.6% Data unavailable Effect of Social Desirability In order to ascertain if social desirability had an effect on therapists reported use of assessment instruments, the summated soci al desirability scale was regressed on the outcome variable ( almost always use an inst rument to assess cognition ). Results
84 indicated that, for therapists who reported al most always using an instrument on initial evaluation to assess cognition, social desirabili ty did not have a significant effect. Results of Research Questions Research Question 1 What are the current practic es of occupational therapi sts regarding screening or assessing the cognitive status of non-de mented elderly patients referred for rehabilitation? 1. a. Do therapists, on initial evaluation, routinely screen or assess the cognitive status of non-demented elderly patients referred to rehabilitation? Therapists were asked to indicate how frequently they assessed specific performance areas and skills during their initial evaluation (balance and mobility, coordination, cognition, and sensation). Two hundred thirty one therapists (76.2% of the sample) indicated they almost always assessed cognition on initial evaluation ; 57 (18.8%) reported assessing c ognition very frequently; 12 (4%) sometimes; and 2 (0.7%) almost never. Next, a Chi square test of independence was conducted to expl ore the proportion of therapists who reported using standardized instruments among th e 231 therapists who indicated almost always assessing cognition on initial evaluation. Results are summarized in Table 11 and revealed that 214 of the 231 respondents who almost always assess cognition on initial evaluation used a brief informal assessment of orientation and 73 used a standardized assessment instrument.
85 Table 11 Distribution of Therapists Who Almost Always A ssess Cognition on Initial Evaluation by Choice of Method (231) Almost Always Assess Cognition on Initial Evaluation Sig Almost Always Use Informal Assessment 1.093 .296 Yes 214 No 17 Total 231 Almost Always Use Standardized Assessment Instrument 8.362 .004* Yes 73 No 158 Total 231 *Significant at p < .05 1. b. How frequently do therapists utilize specif ic cognitive assessment instruments on initial evaluation of elderl y rehabilitation patients? This question included an informal asse ssment of orientation, nine standardized screening or assessments instruments, and one fictitious assessment. Respondents were asked to indicate the frequency with which they used any of the listed assessments. Response options to this question were on a f our point Likert scal e and items were not mutually exclusive. Three standardized instruments topped the list as the most widely used among respondents using an instrument almost always very frequently or sometimes One hundred ninety nine therapists (65.7%) repo rted using the MMSE with any frequency, and of those, 34 reported using that instrument almost always (11.2 %). The Allen Cognitive Level Screen (ACL) was the sec ond most frequently selected, with 127
86 therapists (41.9%) using this instrument with any frequency, and 23 therapists using the instrument almost always (7.6%). Third on the list, The Assessment of Motor and Processing Skills, was used with any frequency by 91therapists (30%) therapists and of those, 35 (11.6%) reported almost always using this instrument. In comparison, 302 respondents (99.7%) reported using an inform al assessment of orientation with any frequency, while 278 (91.7%) reported almost always using that method of assessment. Conversely, 271 ther apists reported almost never using the Short Portable Mental Status Questionnaire (89.4%); 266 therapists reported almost never using the Lowenstein Occupational Therapy Assessment (87.8%); and 213 reported almost never using the Cognitive Performance Test (70.3%). A summa ry of the frequency of use of cognitive screens and assessments is provided in Table 12. **Fictitious, not a cognitive assessment instrument Table 12 Frequency of Use of Informal Assessment of Orientation a nd Standardized Screening and As sessment Instruments (N=303) Assessment Instrument Almost Always Very Frequently Sometimes Almost never Informal assessment 278 (91.7%) 22 (7.3%) 2 (0.7%) 1 (0.3%) **Canadian Occupational Performance Measure 4 (1.3%) 4 (1.3) 34 (11.2%) 255 (84.2%) Clock Drawing Test 3 (1.0%) 17 (5.6%) 125 (41.3%) 151 (49.8%) Cognitive Performance Test 11 (3.6%) 15 (5.0%) 52 (17.2%) 213 (70.3%) **Nutritional Assessment of Elderly 3 (1.0%) 10 (3.3%) 20 (6.6%) 259 (85.5%) Assessment of Motor and Processing Skills 35 (11.6%) 29 (9.6%) 27 (8.9%) 207 (68.3%) Routine Task Inventory 17 (5.6%) 21 (6.9%) 45 (14.9%) 211 (69.6%) Allen Cognitive Level Screen 23 (7.6%) 34 (11.2%) 70 (23.1%) 172 (56.8%) Mini Mental Status Exam 34 (11.2%) 51 (16.8%) 114 (37.6%) 97 (32.0%) Lowenstein Occupational Therapy Assessment 0 (0%) 4 (1.3%) 28 (9.2%) 266 (87.8%) Short Portable Mental Status Questionnaire 7 (2.3%) 6 (2.0%) 11 (3.6%) 271 (89.4%)
87 Insight into therapists utilization of a ssessment instruments would be incomplete, without also inquiring about the extent to which they believ ed certain characteristics of the instruments to be important in their selection. This i nquiry is particularly valuable given the emphasis on engaging in evidencebased practice, and th e increasing demands for productivity. Based on responses as to which characteristics were very important therapists selected having the knowledge and skills to administer and score an instrument (88.7%) and instruments that can be administered quickly (73.1%) as the top two characteristics, followed by supported by research (66.4%). Standardization (34.1%), on the other hand, ranked below is accepted by the team (57.9%) and complies with evaluation form (51.0%). A summary of therapists re sponses is provided in Table 13. Table 13 Important Characteristics in Therapists Choice of Assessment Instruments Characteristic Frequency Very Important Somewhat Important Not Important at All Having knowledge and skills in administering and scoring instrument 302 268 (88.7%) 30 (9.9%) 4 (1.3%) Instrument is standardized 302 103 (34.1%) 173 (57.3%) 26 (8.6%) Is supported by research 298 198 (66.4%) 95 (31.9%) 5 (1.7%) Complies with evaluation form 300 153 (51.0%) 115 (38.3%) 32 (10.7%) Can be administered quickly 301 220 (73.1%) 77 (25.6%) 4 (1.3%) Is accepted by the team 299 173 (57.9%) 114 (38.1%) 12 (4.0%) 1. c. Are therapists, in the course of treatm ent, likely to assess the cognitive status of patients who fail to improve as anticipated in the initial evaluation and treatment plan? A frequency analysis revealed that therapists in this sample were more likely to assess a patients cognitive status than to discharge a patient who was not improving as
88 anticipated in the initi al evaluation. Of 286 therapists re sponding to this case scenario, 258 (85.1%) indicated they would assess the pa tients cognitive status. However, the sample was divided as to how they woul d code the assessment for the purpose of reimbursement. Slightly more than half of those res pondents (51.4%) indicated they would assess cognition and code it as su ch; 38.8% replied that they would assess cognition, but code it as a treatment to avoi d hassles. Only 11 participants (3.8%) responded they would discharge the patient as having received maximum benefit from therapy and 17 (5.9%) indicated they woul d discharge the patient because of poor potential for rehabilitation. Research Question 2 What is the relationship between personal and practice context and therapists frequency of use of standardized cognitive assessment instruments? Associations between covari ates and the outcome variab le were tested through Chi square tests and logistic regressions, controlling for age, education, and gender. Predictors exhibiting significan t associations (p <.05) with the outcome variable were retained for the multivariate analysis. Resu lts of the Chi square tests are summarized in Appendix J. A summary of results of the logi stic regressions is in cluded in Appendix K. 2. a. Is the type of setting (inpa tient settings, home-based se ttings, or other) associated with frequency of use of c ognitive assessment instruments? The type of setting was not significantly associated with the outcome variable, frequency of use of assessment instruments ( =2.995, p=0.224). 2. b. Is facility ownership associated with frequency of use of cognitive assessment instruments?
89 Ownership was not significantly associ ated with the outcome variable, frequency of use of assessment instruments ( =1.891, p=0.388). 2. c. Is there an association between pr ofessional autonomy afforded therapists through the facilities protocols, and frequency of use of cognitive and assessment instruments? Professional autonomy addressed whether ther apists were able to determine areas of performance and performance skills to assess on initial evaluation or if they had to adhere to fixed evaluation protocols, or had restrictions imposed by third party payers. To assess the association between predictor and outcome, pred ictor variable Autonomy was regressed on the outcome vari able controlling for age, ge nder and education. A Wald of 1.855 (CI .845, 2.542) and p >.05 verified that Autonomy was not significantly associated with the outcome variable. 2. d. Is there an associa tion between employer su pport, supervisor support, or facility having sufficient resources and therapists use of assessment instruments? Individual logistic regr essions controlling for age, education, and gender were conducted to determine the association of each of these predictors with the outcome variable. Based on the Wald =.841 (CI .348, 1.467) p >.05, for supervisor support it was determined that this predictor was not significantly associated with the outcome variable. Employer requires assessment on the other hand, was significantly associated with use of assessment instruments based on the Wald = 17.192 (CI 1.947, 6.426) p <.05, and was thus retained for the multivariate analysis. Results of the logistic regre ssion between predictor variable facility has sufficient resources and the outcome variable, controlling for age, educa tion, and gender, yielded a
90 Wald = 12.873 (CI 1.579, 4.738), p <.05. This predic tor was significantly associated with the outcome variable and was also retained for the multivariate analysis. 2. e. Is there a relationshi p between knowledge of the e ffect of aging on cognition and therapists use of cognitive assessmen t instruments in their practice? Only one predictor variable associated with knowledge of the effect of aging on cognition ( fluid intelligence declines with age ) was retained for the multivariate analysis with a Wald = 3.619 (CI .984, 3.006), despite a border line level of significance (p =.057). 2. f. Is there a relationshi p between therapists knowle dge of how to administer and score a variety of screening/a ssessment instruments and their use of these instruments? Having knowledge of how to administer and score standardized assessment instruments significantly increa sed (p <.05) the probability that a therapist would use such instruments. This variable was retained for the logistic regression. 2.g. Are therapists beliefs regarding aging, professional responsibility, or use of screening and assessment instruments associat ed with use of standardized instruments in their initial evaluations of elderly rehabilitation patients? The association between beliefs and use of assessment instruments was tested by regressing each of the three in dividual predictors on the outco me variable, controlling for age, education, and gender. Although in genera l, therapists in this sample held positive beliefs regarding aging and cogniti on, professional responsibility, and use of cognitive screening and assessment instruments, these predictor variables were not significantly
91 associated with the outcome variable at p <.05. A summary of par ticipants responses to belief questions is presented in Appendix L. 2.h .. Is there an associa tion between temporal fact ors of the therapists (e.g. age, years in occupational therapy practice, ye ars in geriatric rehabilitation, length of time in employment at the time of th e survey), and their use of cognitive screening/assessment instruments in geriatric rehabilitation? Results of Chi square tests indicated that years in occupationa l therapy practice and years in geriatric rehabilitation were significantly associated at p <.01 level with the outcome variable. Both predic tor variables were retained for the logistic regression. A frequency distribution of therapists tempor al variables is included in Appendix M. 2.i Is level of education associated with use of assessment instruments? A Chi square test of independence conducte d to test the associ ation between this dichotomous predictor (baccalaureate=0, post-baccalaureate=1) and the outcome variable was not significant at p <.05 (2-tailed), indicat ing that education was not predictive of use of standardized assessment instruments. Research Question 3 What is the effect of co ntext (practice and/or personal) in predicting therapists utilization of standardized assessment instruments? A multivariate logistic regression was used to explore the effect of context on predicting therapists utilizat ion of standardized assessment instruments. Based on results of bivariate analyses, thirteen predic tor variables were entered into the logistic regression to obtain the best fitting model for the data. The analysis was carried out using the logistic procedure in SPSS versi on 15.0 in the Windows XP environment.
92 A backward elimination method was selected wi th p = .05 for entry into the regression, p =.10 for removal, p =.50 for cut-off, and a ma ximum of 20 iterations. Eight models were tested by SPSS to determine the best fitting model for the data. Summaries of the models are provided in Appendix N. The final mode l consisting of six pred ictor variables (two from the practice context, and four from th e personal context), is summarized in Table 14. The coefficient represents the change in the logit of the outco me, per unit change in the corresponding predictor vari able. Significance values of the Wald statistic (p < .05 and p < .01) indicate that the coefficient for each predicto r is significantly different from zero in this model, i.e. all predictors ( assess and charge for trea tment, knowledge of assessment instruments, employer requires asse ssments, knowledge of fluid intelligence, sufficient resources, and years in geriatric rehabilitation experience ) are making significant contributions toward pred icting the outcome (Field, 2001). The odds ratio in the model gives an i ndication of the change in odds per unit change in each predictor. Results of the logistic regression identified Employer requires assessment as the strongest predictor of use of asse ssment instruments. An odds ratio of 3.39 indicated that therapists whose employers required them to use cognitive assessment instruments were 3.4 times more likely to ut ilize them such instruments than were therapists who did not have that type of support in their practice. Predictor variable, facility has sufficient resources, was found to be the second strongest predictor of use of standardized instruments. With an odds ratio of 2.62, therapists whose facilities ha d sufficient available resources were slightly better than 2.5
93 times as likely to use standardized assessment instruments as their counterparts in facilities lacking sufficient resources. The likelihood of a therapist almost al ways using an assessment instruments on initial evaluation increased by 1.19 times for every unit change in knowledge of assessment instruments (i.e. every additional assessment instrument he or she was trained to administer and score). Similarly, therap ists who had knowledge of changes in fluid intelligence as a result of normal aging were tw ice as likely to use standardized cognitive assessment instruments as therapis ts who lacked this knowledge. The association between years in geri atric practice and use of assessment instruments had an odds ratio of 1.51. As this variable had 4 categories (<1 to 5 years; 610 years; 11-15 years; and >15 years), this odd ratio increases with each ascending category of years in geriatric practice. In ot her words, therapists who had 6-10 years of geriatric experience were 1.5 times as likely to use cognitive assessment instruments as therapists with 5 years or less of geriatric e xperience. Therapists with >15 of geriatric experience were 1.5 times as likely to use a ssessment instruments as therapists with 1115 years experience, but 2.25 times as likely as therapists with 6-10 years geriatric experience, and 3.38 times as likely as therap ists with 5 years or less of experience. As for responses to the case example, a significant Wald statis tic (2.38*), associated with a negative B coefficient (.696) and an odds ratio of .50, reveals that therapists who replied they would assess cognition and charge for a treatment to avoid hassles are half as likely to use a standardized instrument as are therapists who would charge for the procedure as a cognitive assessment. Confid ence intervals identify the boundaries within which 95% of samples measuring the same vari ables as the present study would fall.
94*p<.05, ** p<.01 Assessing the Logistic Model The model was assessed by examining test s results of the logistic regression, statistical significance of predictor variables in the model, and results of the Homer and Lemeshow goodness-of-fit test. Goodness-of-fit statistics in Table 15 assess the fit of the logistic model against observed outcomes. The insignificant results of the Hosmer & Lemeshow statistic are desirable, as they suggest that the model was fit to the data. Cox & Snell and Nagelkerke provide R values that repr esent the proportion of the variance in the outcome variable explained by the model, 16.6% and 24.3% respectively. Table 14 Summary of Logistic Regression An alyses Predicting Use of Standardiz ed Cognitive Assessment Instruments 95% Confidence Interval Variable Odds Ratio P value Lower Upper Assess and charge for treatment .50 .036* .260 .96 Has knowledge of instruments 1.19 .041* 1.00 1.40 Employer requires assessments 3.39 .000** 1.71 6.73 Knowledge fluid intelligence 2.00 .034* 1.05 3.78 Sufficient resources 2.62 .004** 1.37 5.00 Geriatric rehabil itation experience 1.51 .003** 1.15 1.98
95 Table 15 Assessment of Logistic Model Test Goodness-of-fit test df p Homer & Lemeshow 14.634 8 .067 Cox & Snell R Nagelkerke R .166 .243 Multicollinearity was assessed with a prel iminary review of the correlation matrix (Table 16). Correlation coefficients < .60, indicated there was no collinearity among the predictor variables in the model. VIF values were we ll under 10, the average VIF (1.016) was not considerably greater than 1, and to lerance values were well above 0.2 indicating there was no collinearity in the model. Assessing the Logistic Regression Regression diagnostics were conducted to determine whether the model fit the data well or was influenced by a small number of cases, and to ascertain if the model could generalize to other samples. Residual statis tics were first examined to determine if extreme cases were exerting undue influence on the model. Seven cases (3%) in the sample had standardized residuals >2.0, and one had a standardized residual >3.0. All cases had studentized residua ls ranging from 2.05 to 2.28, and deviance ranging from 2.03 to 2.26, so they were all below 2.5 and were not cause for concern. All of the cases had a Cooks statistic and DFBetas <1, indicatin g there are no influe ntial cases having an effect on the model (Field, 2001). Expected value of leverage was computed using the formula k+1/N, where k=number of predictors, and N=sample size. For this analysis, th e expected value of
96 leverage was (6+1/271) 0.026, and the range of leverage values for these cases was from .009 to .02. Leverage close to 0 indicates no undue influence by any case (Field, 2001). Appendix O summarizes the regression diagnostics. Table 16 Correlation Matrix for Predictor Variables in Logistic Model Constant Case Example (assess and charge treatment) Score for knowledge of assessment instruments Employer requires assessment Knows fluid intelligence declines with age Facility has sufficient resources Yrs geriatric rehabilitation Constant 1.00 Would assess and charge as treatment -.203 1.00 Score for knowledge of assessment instruments -.653 .096 1.00 Employer requires assessment -.227 -.019 .000 1.00 Knows fluid intelligence declines with age -.296 -.008 .030 .027 1.00 Facility has sufficient resources -.084 -.053 -.202 -.060 .218 1.00 Yrs experience in geriatric rehabilitation -.767 .007 .258 .164 .131 .051 1.00
97 CHAPTER 5 Discussion Effect of Context on Therapi sts Assessment Practices The purpose of this study was to dete rmine personal and pr actice contexts of occupational therapy associated with the use of cognitive screening and assessment tools. The study specifically focused on the use of su ch instruments by occupational therapists with non-demented medical patients 65 year s of age and older referred for physical rehabilitation. Data were collected through a web-based survey questionnaire constructed using findings from the resear ch literature, individual and focus group interviews of occupational therapists c onducted by this researcher, and by the researchers own clinical experi ence in geriatric rehabilitation. Themes that evolved from the therapists interviews were used to select the theoretical framework and the constructs of in terest for this study. Survey questions were utilized to explore personal and practice factors and their a ssociation with therapists use of screening and assessment instruments. Determining whether therapists assess the cognitive status of their elderly medical patients is important, given the evidence in the literature rega rding age associated frontal lobe changes which affect executive functions (Chao & Kn ight, 1997; Grigsby, Kaye & Robbins, 1995; Herrick et al. 1996; Mast et al. 1999). As impor tant, is determining what
98 tools therapists use and the fr equency of use, to ensure that selected instruments are sensitive to mild forms of impairment and provide information which can be translated into functional treatment goals. Furthermore, concerns regarding the selection and proper use of assessment instruments by occupational therapists are neither new nor insignificant. This issue became a priority for AOTA in 1984, when its Representative Assembly released a statement identifying as a top priority for the association the development of standardized assessment instruments for occupational thera py and the utilization of such instruments by occupational therapists (Elfant-Asher, 1996) Three questions were designed to obtain information about participants assessment practices. Therapists were asked: 1) how fr equently they screened or assessed cognition on initial evaluation; 2) what instruments or processes they utilized to conduct the screens and assessments; and 3) how they would likely proceed if a patient failed to progress as anticipated in the initi al treatment plan. In response to the question, how frequently therapists screened or assessed cognition on initial evaluation 231 therapists (76.5% of the sample) reported they almost always assessed cognition. As to specific instruments or processes utilized to conduct the screens and assessments 214 participants reported usi ng an informal assessment of orientation and only 73 reported usin g a standardized assessment. Among therapists indicating almost always utilizing any of the three most frequently utilized tools on initial evaluation, 35 respondent s (11.6%) reported utilizing the AMPS; 34 (11.2%) the MMSE; and 23 (7.6%) the ACL. Conversely, among therapists reporting almost never using a standardized assessment, 266 therapists reported
99 almost never using the Lowenstein Occupati onal Therapy Assessment (87.8%); 213 (70.3%) almost never used the Cogni tive Performance Test; and 211 (69.6%) almost never used the Routine Task Inventory. With the exception of the Nutritional A ssessment of the Elderly and the Canadian Occupational Performance Measur e, all the standardized inst ruments listed in the survey questionnaire were cognitive screens or a ssessments. Five of these instruments (Cognitive Performance Test, ACL, Lowenstein Occupational Therapy Assessment, Routine Task Inventory, and AMPS) were de veloped by and for occupational therapists (Elfant-Asher, 1996), yet only the ACL and the AMPS were frequently used by therapists in the study sample. These findings, combined with the prevalence of use of the MMSE, indicate that while AOTAs goals of increasi ng therapists use of in struments have been somewhat met, therapists preferences of a ssessments are limited and heavily influenced by instruments used by other professions, such as the MMSE. Therapists were asked questions as to whether they had received training in the administration and scoring of a number of in struments. Of the top three instruments which therapists reported having been trai ned to administer and score (MMSE 82.2%, ACL 78.5%, and Clock Drawing Test 62.4%), th e ACL is the only occupational therapy assessment developed for the purpose of a ssessing a patients cognitive capacity for functional performance. The ACL screen pr ovides information about the individuals level of cognitive function at the time of the assessment, is sensitive to mild forms of cognitive impairment, economical, and quick to administer. Additionally, the ACL highlights approaches to maximi ze performance and safety and is easy to incorporate into the treatment plan. The MMSE, on the other hand, was developed as a quick screen to
100 detect moderate to severe cognitive declin e and is not sensitive to mild forms of impairment (Mitrushina & Satz, 1994; TengWai, Knopman, Geda, Edland et al. 2003; Tombaugh & McIntyre, 1992). The Clock Dr awing Test, while a good indicator of cognitive impairment, does not provide cl ear guidelines specif ic to functional performance. In order to explore factors which may in fluence choice of instruments, therapists were also asked to indicate the degree to wh ich certain characteristics were important in their selection of assessment instruments. Therapists responses, summarized below, point to the need to emphasize the selection and use of standa rdized instruments that are congruent with evidence-based practice rath er than driven by external influences. having the knowledge and skills to administer and score the instrument (88.7%) assessment can be administered quickly (73.1%) assessment is supported by research (66.4%) assessment is accepted by team (57.9%); assessment complies with evaluation form (51.0%) assessment is standardized (34.1%). It is possible, given the high percentage of therapists who were trained to use the MMSE and the prevalence of use of this instru ment, that therapists practices are highly influenced by their professional preparation. It is also possible that therapists, based on participants choices of impor tant characteristics, tended to use the MMSE because it could be administered in a brief amount of time.
101 Although a number of occupational ther apy assessment instruments have been developed since 1984, findings from this st udy raise questions re garding how much progress has been made in promoting ther apists knowledge and us e of standardized cognitive assessment instruments. This is especially true regarding mild deficits associated with aging. Participants in this study were given a case scenario and asked to indicate how they would proceed if a patient, during the c ourse of treatment, failed to improve as anticipated. Overwhelmingly, therapists responded they would assess the patients cognitive status. Only 9.7% of respondent s indicated they would discharge the patient; 51.4% reported they would assess cognition and charge as such, and 38.8% indicated they would assess cognition but code it as a treatment to avoid hassles. While it is encouraging that 90.2% of respondents would choose to assess the patients cognitive status, the fact that almost 39% would choose to code the procedure as a treatment to avoid hassles raises questions regarding the therapists ac tual or perceived degree of autonomy to make clinical decisions. Knowledge of Aging and Cognition A number of survey questions addressed therapists knowledge a bout the effects of aging on cognition, yielding the following results: less than one-third of respondents agreed that impairm ent in cognitive function is part of normal aging slightly more than half of the sample agreed that a knowledge able therapist can effectively assess cognitive status by conversing with the patient
102 two-thirds of the sample agreed that ab ility to perform ADLs is a good measure of a persons cognitive status only one-third agreed that fluid in telligence deteriorates with age three-quarters of the sample agreed that fluid intelligence affects the ability to learn in rehabilitation Responses to these questions indicate th at, although 75% of the study sample were aware of the role of fluid intelligence on l earning and rehabilitation, two-thirds of the sample or more were not aware that cognitive function and fluid intelligence decline with normal aging. Therapists responses rega rding aging and cognition were unexpected given the extensive literatu re and evidence of age-related neur ophysiologic changes affecting cognitive processes and fluid abi lities (Chao & Knight, 1997; Chodosh et al. 2004; Grigsby et al. 1995), and the impact of illness, chronic conditions and polypharmacy on fluid and executive abilities (Elias, 1998; Ruchinskas et al, 2000). For example, studies of older patients undergoi ng surgical procedures for hip fracture estimated a 30 40 % incidence of cognitive problems post surgery in this population (Herrick et al. 1996; Mast et al. 1999). Add itionally, research comparing healthy controls with patients diagnosed with peripheral vascul ar disease reported th e latter group had an estimated prevalence of 25% of frontal lobe dysfunction and atten tional impairment (Rao et al. 1999). Published studies support assessing the c ognitive status of ol der medical patients due to the prevalence of cognitive deficits in this group (Garrett et al. 2004; Kilander et al. 1998; Kirkpatrick & Jamieson, 1993; Waldstei n et al. 1996). These studies advocate
103 for the development and utilization of instru ments sensitive to mild forms of cognitive decline for assessing executive function and fl uid abilities in older medical patients. Therapists responses regarding their use of ADL performance as a proxy for cognitive status were also unexpected, as ADLs are crystallized abilities stored in procedural memory (C. Allen, personal communication, September 1998; Ruchinskas, Singer, & Repetz, 2000) and therefore not a good measure of mild to moderate cognitive impairment. This is supported by evidence fr om several studies of frontal lobe function indicating that, while fluid abilities are aff ected by age-related changes, crystallized abilities remain unaffected (Barberger-Gateau & Fabrigoul e, 1997; Christensen et al. 1994; Kaufman et al. 1989). Similarly, in a study of cognitive impairment and functional performance, Galanos et al. (1994) concluded that despite a 50% prevalence of cognitive impairment, health problems, and de pression, participants were still able to perform activities of daily living. Evidence-based practice is not limited to the selection of appropriate therapeutic techniques and interventions, but should al so be informed by sound research on issues regarding the medical, psychological, and develo pmental status of pa tients. Therapists misconceptions of aging and cognition have im plications for the pr ofessional preparation of occupational therapists, for continuing pr ofessional education, and for evidence-based practice because they are likely to affect th erapists assessment and treatment practices. Beliefs In general, therapists in this sample he ld positive beliefs about the use of cognitive screenings and assessment instruments, agi ng, and professional responsibility. Of 302 therapists who responded to why assessi ng cognition was problematic, only 4.6%
104 strongly or moderately agreed that it may lead to costly referrals for further evaluation while 20.5% strongly or moderately agreed that it is difficult to translate that information into the treatment plan. The latter is partic ularly true of assessments such as the MMSE, which was the most widely use instrument and the Short Portable Mental Status Questionnaire. A number of occupational th erapy assessments such as the ACL and other instruments reported as being used less fr equently (the CPT, RTI, etc.) are linked to functional performance and provide guidelines so that a patients score and his or her approach to the task(s) required by the instrume nt can be easily translated into functional treatment goals. Most therapists indicated that initial ev aluations should not be limited to physical function and assessment of cognitive status s hould not be restricted to diagnoses of stroke, head injury, or Alzheimers disease. Most also agreed that occupational therapy education should include extensive training in assessing the cognitive status of older patients. This further suppor ts the issue of up-to-date evidence-based professional preparation that incorporates the physical, cognitive and psychosocial aspects of the patient. As to responses to questions regardin g therapists beliefs about professional responsibility, 87% of respondents agreed that assessing cognitive status is the responsibility of the occupationa l therapist. When asked if only a psychologist or other licensed mental health prof essional should assess cogniti ve status, only 20.8% of participants responded in agreement.
105 Practice Setting Results indicated that neither ownership nor type of setting were associated with the outcome variable ( frequency of use of assessment instruments) There was also a lack of association between autonomy afforded therapists by the practice setting and the outcome variable. One plausible explanation fo r this is that the e ffect of autonomy may be diminished since the strongest predictor of use of standardized assessment instruments was employers requiring ther apists to assess cognition ( odds ratio 3.5). Another, although perhaps less likely explan ation, is that any restrictio ns imposed by the type of ownership or by fixed facility protocols may be circumvented by coding services in such a manner as to avoid denial of reimbursement, (i.e. assess cognition and code it as a treatment to avoid hassles). If the latter is the case, th is would contradict Burke and Cassidys (1991) conclusions that emphasi s on productivity, efficiency, and costcontainment were altering both the freque ncy and type of services provided by occupational therapists. Implications for Occupat ional Therapy Practice Results of this study are congruent with the tenets of the se lected theoretical framework, which posits that: 1) human behavior is best understood as the product of the interaction between the indi viduals personal attributes and the physical and social environment in which the individual functions ; and 2) performance cannot be considered without taking into account the individua ls environment (Howe & Briggs, 1982). Research question three explored the e ffect of context on predicting therapists utilization of assessment instruments. Results of the multivariate analysis showed that four predictor variables fr om the personal context ( knowledge of assessment instruments;
106 knowledge of aging and cognition; years in geriatric practice; and assess cognition and charge as treatment ), and two from the ideologic layer of the practice context ( employer requires assessment and facility has sufficient resources ) were found to contribute significantly to therapists us e of standardized assessment instrument. Although the odds of utilizing standardized assessment instrument s were highly associated with the practice context neither the personal nor the practi ce context alone accounted for therapists assessment practices. While therapists know ledge of a number of instruments and of the effect of aging on cogn ition ensures that he or sh e has the skills to conduct appropriate assessments, employer requirements and availability of sufficient resources appear to be necessary for ther apists to both utilize and impr ove their assessment skills. There are also implications for promo ting the use of standardized assessment instruments among therapists in order to atta in AOTAs goals. Establishing a task force to investigate therapists assessment practices in geriatric rehabil itation would be a good starting point. Developing a series of white papers addressing the educational preparation of occupational therapists woul d be a reasonable progr ession. Additionally, AOTA should engage in convers ations with educational prog rams to ensure that new practitioners enter the field with sufficien t knowledge of aging and cognition and with skills to appropriately utilize a variety of instruments. Similarly, AOTA could promote evidence-based continuing education opportun ities for clinicians to enhance their knowledge of aging and cognition and to acquire or refine their assessment skills. As it did in 1984 to promote development and ut ilization of appropriate assessment instruments, AOTA could make identification of cognitive de ficits through appropriate selection and utilizati on of standardized instruments a priority.
107 AOTA has been successfully involved w ith a number of policy issues, including their recent efforts to ensure that Medi cares current procedur al terminology (CPT coding) would not preclude therapists from utilizing cognitive screening and assessing instruments (J. Thomas, AOTA, personal communication, May 2006). Changes to CPT coding proposed in 2006 would have made it more difficult for therapists to get reimbursed for the administration and interp retation of assessments that could be construed as neurobehavioral te sting, therefore therapists would be less likely to engage in the utilization of these types of assessments. Medicares attempted changes to CP T terminology should provide a compelling reason for AOTA, schools of occupational ther apy, and individual therapists to promote training in and utilizing appr opriate cognitive screening a nd assessment instruments. Otherwise, the scope of practice and doma ins of occupational therapy risk being redefined by third party payers and other disciplines. AOTAs continued efforts with this issue should promote policy cha nges to reimbursement so that occupational therapists in geriatric rehabilitation can rout inely use standardized cogniti ve screening and assessment tools as part of their evalua tions and treatment planning. As indicated by the results of the logistic regression, therapists in facilities with limited resources or lack of employer require ments were less likely to engage in cognitive assessments of their elderly patient s than were therapists who had support from their employers. Therefore, promoting utiliz ation of standardized cognitive assessments at the practice context level will require increased advocacy efforts and education by individual therapists and AOTA of agencies rehabilitation compan ies, and third party payers.
108 Public Health Implications Within the next four years, the elderly population in the United States is projected to reach an unprecedented growth when the ba by-boom generation begins to reach age 65 (Hobbs & Stoops, 2002). This projected gr owth will inevitably result in additional health care, Medicare and Medicaid expenditu res as a result of increased hospitalizations, utilization of health services and admissions to nursing homes As such, the growth of the older population will presen t a special challenge to rehabilitation and public health. Adequate provision of services, patient e ducation, and prevention of unintentional injuries and adverse events will require a clea r picture of patients capacity for functional performance, including th eir cognitive abilities. From the standpoint of treatment outcome s, identification of patients experiencing mild cognitive deficits can assist therapists to design realistic treatment plans and engage in interventions that will enhance their patient s safety and functional performance. From the perspective of public hea lth, conditions amenable to tr eatment and early stages of dementia can be identified and treated promp tly and safety risks identified and addressed. In this way, risks for non-compliance, adve rse events, unintentiona l injuries and over utilization of costly health and person al care services may be minimized. Interventions and education provided w ithout a clear measure of the patients cognitive capacity place the burden of assim ilation on the individual and assume that the information will be successfully processed and properly utilized. To the extent that mild degrees of cognitive deficits in the elderly are undetected by medical and rehabilitation personnel, then health education messages, medical and rehabilita tion interventions, and safety precautions and recommendations may be ineffectively delivered.
109 Efforts should be made to educate therap ists across discipline s and clinicians in general, as to the public health implications of their practice. The c linical perspective of treating each patient as an isolated entity within a medical context driven by costcontainment and market pressure must be re placed by a new paradigm. Practitioners must understand that each individual intervention ultim ately affects the health and function of a community and a nation, especially when procedures are bypassed or abridged to satisfy third-party payers. Assessing cognitive status must be seen as a preventive strategy, similar to interventions in which therapists engage to prevent joint damage to arthritic joints, or prevent development of pressure sores in non-ambulatory patients. Continued emphasis on a practice driven solely by economics, while cost effective in the short term, can result in increased costs and higher degrees of exce ss physical and cognitive disabilities in the long term. Limitations of the Study A response rate of 12% may be cons idered a limitation by mailed questionnaire standards. However, there is evidence in the litera ture of similar response rates for online surveys of health professiona ls (Braithwaite et al. 2003). Use of the online survey limited partic ipation in the study to AOTA members who had opted for primary membership in one of two special interest sections, and activated their membership in the sections listserve It is possible th at, although the rosters obtained from AOTA indicated there were 3000 members in the two SIS, not all of those therapists had enrolled in the email listse rve and therefore did not have access to the sections electronic messages. Participati on was thus contingent on therapists receiving
110 and reviewing the special in terest sections email messages and having sufficient knowledge of the Internet to successfully acces s and manage the survey questionnaire. While practice guidelines pr eclude occupational therapy a ssistants from engaging in patient assessment and treatment planning, assistants are allowe d to utilize some instruments to collect patient data to assist the occupational therapist in formulating a treatment plan. It is therefore possible that, in practice, the use of standardized cognitive assessment instruments is higher than was re flected in this study, because none of the survey questions addressed this possibility. Although demographic characteristics of participants in this study compare favorably with results from the AOTA (2006) study on practice, it is possible that therapists who are members of AOTA may differ from non-member therapists in terms of their knowledge, beliefs, and assessment prac tices. Thus, therapists who agreed to participate in the current study also may have differed along these dimensions, from therapists who declined participation. One last limitation of the study was relate d to the survey questionnaire. Providing response options to the question regarding fr equency of use of specific instruments or processes to assess cognition th at were not mutually exclusive made it more difficult to draw comparisons between individual assessments. Strengths of the Study The study reflects findings of a previous qualitative study of occupational therapists practicing in home health (Whaley, 2000) The current study verified constraints regarding productivity requirements and limits imposed by the practice context and third party payers on how therapists practice, wh ich were identified by participants in an
111 earlier focus group and two individual intervie ws. As in the current study where 80% of participants reported that ove r three fourths of their caseload or more consisted of patients 65 years and older, therapists interv iewed reported 80 to 100% of their caseloads were of that age group. All 10 participants in the focus group and interviews identified ex ternal factors (i.e. demands for productivity, limited treatment time, limited resources, and a focus on function), which they felt discouraged their use of standardized instruments to assess cognitive status. Thirty percent of part icipants in the focus group and interviews expressed concern that usi ng a standardized assessment instrument (vs. observing functional performance) would affect thei r productivity. One focus group participant reported not assessing cogniti on to avoid denials by Medi care. Information obtained from the focus groups and interviews suppor t findings from the current study, which indicated that therapists whose employers required them to assess the cognitive status of their older patients, and ther apists whose practice environments had sufficient resources were significantly more likely to use standa rdized cognitive screening and assessment instruments than were their counterparts. Therapists in the focus groups shared sim ilar beliefs with partic ipants in the current study regarding cognition and agi ng. Fifty percent of focus group participants reported their home health patients were not able to transfer skills learned in inpatient rehabilitation and all ten focus group partic ipants described problems their patients experienced with judgment, recall, non-complia nce and poor safety awareness. In spite of that, focus group participants linked those de ficits with patient ch aracteristics such as stubbornness or lack of motivation, rath er than with limited capacity.
112 All focus group participants interpreted cogniti ve impairment as severe deficits associated with trauma, dementia or other neurologic events. When cognitive deficits and cognitive impairment were redefined fo r focus group and indivi dual participants, only two therapists (20%) cha nged their estimates of the prevalence of such deficits among their patients to 50%. As in the current study, therapists in the focus group predominantly relied on informal assessments of memory and orientation, and believed observation of ADLs to be a useful way to screen cognitive status. The study presents new information This is possibly the fi rst study in occupational therapy, adopting a theoretical pe rspective utilized in practice, to consider the influence of both personal and external contexts on therapists clinic al practices. The study also presents new information for public health, an d the need to engage in further research regarding the public health implicatio n of medical and clinical practice. The study is relevant. As Walker (2000) cautione d, changes in response to managed care were not and will not be a one time event. Instead, they signaled the beginning of a constant state of flux wher e change is rapid and driven by cost containment, efficiency, and competition. Ho w occupational therapists adapt to these changes while continuing to provide eviden ce-based care and engage in sound clinical decision-making is perhaps even more relevant today. The study is timely Within two weeks of comple ting data collection for this study, AOTA sent its own inquiry thr ough all special interest list serves urgently requesting information as to the type of cognitive assessment instruments therapists were using in the field. This request was made as AOTA was preparing to respond to proposed changes to Medicares Curre nt Procedural Terminology (C PT) coding, which threatened
113 to limit the use of any type of cognitive or neuropsychological assessment to licensed psychologists. The proposed change would have precluded therapists from assessing their patients cognitive status and conducting comprehens ive assessments in their evaluations, by denying reimbursement to disciplines other than psychology. Such changes also have the potential for increasing costs and limiting patients access to needed and appropriate services. Additionally, new Medicare guidelines for reimbursement require that therapists provide more thorough documentation of their patient assessments and that they utilize specific standardized instruments in their pr actice. Medicares list of standardized assessments include several instruments such as the Lowenstein Occupational Therapy Assessment, The Routine Task Inventory, and the Cognitive Performance Test, which were almost never utilized by therapists in this study and which fewer therapists reported being trained to use in their professional prep aration or in continui ng education courses. The study is also timely because of the m ounting awareness in th e literature of the prevalence of cognitive impairment among older, non-demented, medical patients and the failure of medical, nursing, and rehabilitation professionals to identify these patients. Recommendations for Future Research Of the six predictor variables associated with use of standardized assessment instruments, one ( years in geriatric rehabilitation ) is a function of time, but predictors within knowledge and support are modifiable. This is particularly important when we consider that the highest odd ratios in the model for use of assessment instruments were associated with employer requirements (3.4) and with the facility having sufficient
114 resources (2.6). The effect of context on pr actice should be an area of continued research for AOTA and other researchers in order to improve the practice of occupational therapy and provide therapists with the knowledge and to ols to advocate for and facilitate change. As such, a larger scale study of members and non-members would be useful to more fully understand the assessment practices of therapists in general, the factors associated with therapists use of standardized cognitive screening and assessment tools, and the perceived barriers to their use of such instruments. Additionally, supervisor and corporate knowledge and beliefs regarding use of cognitive assessments should be studied in or der to determine ways of bolstering support for therapists assessment practices. Th en, along with continuing education and university OT programs, AOTA could promot e the development, standardization, and skillful utilization of cognitive screening and assessment instruments to identify mild forms of cognitive impairment in older medical patients. Whether therapists responses to questi ons regarding age-associated changes in cognition represent an actual gap in knowledge or speak to how ther apists conceptualize cognition and its relationship to functional perf ormance, this is certainly an area that warrants further investigation and intervention. Thus, future research exploring the impact of practice and personal contexts on th erapists assessment practices should also inquire as to the therapists perceptions regardin g the relationship between cognition and functional performance, as described in Kni ghts literature review (2000). Such research would inform as to additional personal factor s that may influence therapists decisions to screen/assess the c ognitive status of elderly nondemented medical patients.
115 Research on the impact of screening/assessing cognitive status on treatment outcomes, functional performance, cost factor s, safety, caregiver burden, and patient and caregiver satisfaction with delivery of occupational therapy services is also recommended. This type of research would provide useful information in the following domains: 1) determining whic h, of a variety of screening and assessment instruments, is/are sensitive to milder forms of cognitive impairment; 2) establishing which screening and assessment instruments provide informati on which can be translated into goals of functional performance and safety; and 3) expl oring the efficiency and cost-effectiveness of identifying cognitive deficits in older non-demented medical patients; Studies such as the one conducted for th is dissertation, combined with actual observation of therapists assessment practices and/or chart reviews, would be useful in order to control for potential effects of social desirability on self-report. Although the current study focused on occupational therapists, the need for further research goes beyond the practice of occupati onal therapy and the responsibility of AOTA. The issue of under-recogni tion of cognitive deficits in older patients has been identified across health care providers (Pisani, et at., 2003; Ruchinskas, 2002) and, as such, should be of concern to the National Institutes of Health and the Centers for Disease Control and Prevention. As in this study, the focus of future research should be on barriers to identifying cogni tive deficits among older pa tients hospitalized for acute conditions in order to determine suitable interventions to en hance professionals awareness of the issue as well as increas e their screening and assessment practices Results of the recommended research should serve to guide evidence-based practice; to educate students of occupational therapy, othe r practitioners, consumers,
116 family caregivers, and payers; and to provide practitioners with information and tools to influence health care policy, particularly as it affects delivery of serv ices to elder clients.
117 References Allen, C. (1985). Occupational therapy for psychi atric diseases: Measurement and management of cognitive disabilities. Little Brown, Boston. Allen, C., Earhart, C., & Blue, T. (1992). Occupational therapy treatment goals for the physically and cognitively disabled. Rockville, Md: American Occupational therapy Association. Allen, C., Earhart, C., & Blue, T. (1995). Occupational therapy treatment goals for the physically and cognitively disabled (2nd ed.). Rockville, Md: American Occupational Therapy Association. American Occupational Therapy Asso ciation. (2002). Occupational Therapy Practice Framework: Domain and process. American Journal of Occupational Therapy 56, 609-639. American Occupational Therapy As sociation. (2006). 2006 AOTA Workforce and Compensation Survey: Occupational therap y salaries and job oppor tunities continue to improve. OT Practice, September 25, 2006 Ary, D., Jacobs, L., & Razavieh, A. (2002). Introduction to research in education. Belmont, California: Wads worth/Thomson Learning.
118 Barberger-Gateau, P., & Fabrigoule, C. (1997). Disability and cognitive impairment in the elderly. Disability and Rehabilitation, 19 (5), 175-193. Barnes, C., Conner, D, Legault, L., Reznickova, N., & Harrison-Felix, C. (2004). Rehabilitation outcomes in cognitively impair ed patients admitted to skilled nursing facilities from the community. Archives of Physical Medicine and Rehabilitation, 85, 1602-1607. Barris, R. (1987). Clinic al reasoning in psychosocial occupational therapy: The evaluation process. Occupational Therapy Journal of Research, 7, 147-162. Baum, C. (1985). Growth, renewal, and challenge: An important era for occupational therapy. American Journal of Occupational Therapy, 39, 778-784. Boone, K. B., Miller, B. L., Lesser, I. M., Hill, E., & D'Elia, L. (1991). Performance on frontal lobe test s in healthy, older individuals. Developmental Neuropsychology, 6, 215-223. Braithwaite, D., Emery, J., de Lusign an, S., & Sutton, S. (2203). Using the internet to conduct survey s of health professionals : a valid alternative? Family Practice, 20(5) 545-551. Bureau of Labor Statistics, U.S. De partment of Labor. (2006). Occupational Therapists. Occupational Outlook Handbook, 2006-07 Edition (On-line). Available: http://www.bls.gov/oco/ocos078.htm (obtained May 29, 2006 ). Burke, J., & Cassidy, J. (1991). The issu e is Disparity between reimbursementdriven practice and humanistic va lues of occupational therapy. American Journal of Occupational Therapy, 45, 173-176.
119 Center for Survey Research, Indian a University. (2004). Survey research methods. Topics in methods and measurements (Online). Available: http://www.indiana.edu/~survmeth/survey_coop_nonrespon_04.htm Challiner, Y., Carpenter, I., Potter, J., & Maxwell, C. (2003). Performance indicators for hospital services for older people. Age and Ageing, 32 (3), 343-346. Chao, L., & Knight, R. (1997). Prefrontal deficits in attention and inhibitory control with aging. Cerebral Cortex, 7 (1), 63-69. Christiansen, C., & Baum C. (Eds.). (1997). Occupational therapy: Enabling function and well-bein g, pp 46-71. Thorofare, NJ: Slack. Christensen, H., Jorm, A., Henderson, A., Mackinnon, A., & Korten A. (1994). The relationship between health and cognitive functioning in a sample of elderly people in the community. Age and Ageing, 23 (3), 204-212. Chodosh, J., Seeman, T., Keeler, E., Se wall, A., Hirsch, S., et al., (2004). Cognitive decline in high functioning older persons is associated with an increased risk of hospitalization. Journal of the American Geriatrics Society, 52, 1456-1462. Clark, F., Parham, D., Carlson, M., Fra nk, G., Jackson, J., & Pierce, D., et al., (1991). Occupational science: Academic i nnovation in the service of occupational therapy's future. American Journal of Occupational Therapy, 45 (4), 300-310. Clarnette, R. M., & Patterson, C. J. (1 993). Hypothyroidism: Does treatment cure dementia? Journal of Geriatric Psychiatry, 6, 23-27. Cody, R., & Smith, J. (1991). Applied Statistics and the SAS Programming Language (3rd ed.). Englewood Clif fs, New Jersey: Prentice Hall.
120 Cohen, J. (1988). Statistical power analysis for the behavioral sciences (2nd ed.). Hillsdale, New Jersey: Lawrence Erlbaum Associates, Inc. Cohen-Zion, M., Stepnowsky, C., Johnson, S., Marler, M., Dimsdale, J., AncoliIsrael, S., et al., (2004). Cognitive changes and sl eep disordered breathing in the elderly: Differences in race. Journal of Psychosomatic Research, 56 (5), 549-553. Cohendy, R., Brougere, A., & Cuvillon, P. ( 2005). Anesthesia in older patients. Current Opinion in Clinical Nutrition and Metabolic Care, 8 (17), 17-21. Collie, A., & Maruff, P. (2002). An analysis of systems of classifying mild cognitive impairment in older people. Australian and New Zealand Journal of Psychiatry, 36, 133-140. Cook, C., Heath, F., & Thompson, R. (2000). A meta-analysis of response rates in web or internet-based surveys. Educational and Psycholog ical Measurement, 60 (6), 821-836. Crowne, D & Marlowe, D (1960). A new s cale of social desirability independent of pathology. Journal of Consulting Psychology, 24 (4), 349-354. DeCarli, C., Miller, B., Swan, G., Ree d, T., Wolf, P., & Carmelli, D. (2001). Cerebrovascular and brain morphologic correlat es of mild cognitive impairment in the National Heart, Lung, and Blood Institute Twin Study. Archives of Neurology, 58 (4), 463-467. Dillman, D. (2000). Mail and internet surveys: The tailored design method (2nd ed.). New York: John Wiley & Sons, Inc.
121 Dunn, W., Brown, C., & McGuigan, A. (1994). The Ecology of Human Performance: A framework for cons idering the effect of context. American Journal of Occupational Therapy, 48 (7), 595-607. Dunn, W., Brown, C., & Youngstrom, M. (2003). Ecological Model of Occupation. In P. Kramer, J. Hinojosa, C. Brasic-Royeen (Ed.), Perspectives in human occupation (pp. 222-263). Baltimore, MD: Li pppincott Williams & Wilkins. Elias, M. (1998). Effect of chr onic hypertension on cognitive function. Geriatrics, 53 (Supp 1), S49-S52. Elfant Asher, I. (1996). Occupational therapy assessmen t tools: An annotated index (2nd. Edition). Bethesda, MD; American Occupational Therapy Association. Field, A. (2001). Discovering statistics using SPSS for windows. Thousand Oaks, CA. Sage Publications, Inc. Fleming, M. (1991). The therap ist with the three-track mind. American Journal of Occupational Therapy, 45, 1007-1014. Fondiller, E., Rosage, L., & Neuhaus, B. (1990). Values influencing clinical reasoning in occupational thera py: an exploratory study. Occupational Therapy Journal of Research, 41-55. Foto, M. (1988). Managing changes in reimbursement patterns, Part 1. American Journal of Occupational Therapy, 42, 463-465. Galanos, A., Fillenbaum, G., Cohen, H., & Burchett, B. (1994). The comprehensive assessment of community dwellin g elderly: Why func tional status is not enough. Aging Clinical and Experimental Research, 6, 343-352.
122 Garrett, K., Browndyke, J., Whelihan, W., Pa ul, R., DiCarlo, M., Moser, D., et al., (2004). The neuropsychological profile of vascular cognitive impairment-no dementia: Comparisons to patients at risk for cere brovascular disease an d vascular dementia. Archives of Clinical Neuropsychology, 19, 745-757. Gregg, E., Yaffe, K., Cauley, J., Rolka, D., Blackwell, T., Narayan, K., et al. (2000). Is diabetes associated with cognitiv e impairment and cognitive decline among older women? Archives of Inter nal Medicine, 160 (2), 174-80. Grigsby, J., Kaye, K., & Robbins, L. (1995). Behavioral disturbance and impairment of executive functions among the elderly. Archives of Gerontology and Geriatrics, 21, 167-177. Hatcher, L., & Stepanski, J. (1997). A step-by-step approach to using the SAS System for univariate and multivariate statistics (552 pp). Cary, N.C.: SAS Institute, Inc. Hennekens, C., & Mayrent, S. (1987). Epidemiology in Medicine (1st ed.). Boston: Little, Brow n and Company. Herrick, I., Ganapathy, S., Komar, W., Kirby, J., Moote, C., Dobkowski,, W., et al., (1996). Postoperative cognitive impair ment in the elderly: Choice of patientcontrolled analgesia opioid. Anaesthesia, 51, 356-360. Hinojosa, J., Kramer, P., & Crist, P. (Eds.). (2005). Evaluation: obtaining and interpreting data. Bethesda, MD: American Occupational Therapy Association. Hobbs, F., & Stoops, N. (November 2002) Demographic trends in the 20th century. U.S. Census Bureau, Census 2000 Special Reports, Series CENSR-4, Washington, D.C: U.S. Government Printing Office.
123 Hooper, B. (1997). The relationship between pretheoretical assumptions, and clinical reasoning. American Jour nal of Occupational Therapy, 51. 328-338. Howard, B. (1991). How high do we jump ? The effect of reimbursement on occupational therapy. The American Journal of Occupational Therapy, 45 (10), 875-881. Howe, C., & Briggs, A. (1982). Ecol ogical Systems Model for occupational therapy. American Journal of Occupational Therapy, 36 (5), 322-327. Inouye, S., Zhang, Y., Leo-Summers, L., Jones, R., Marcantonio, E. (2006). Recoverable cognitive dysfunc tion at hospital admission in older persons during acute illness. Journal of General In ternal Medicine, 21 (12), 1276-1281. Kaufman, P., McMahon, R., & Becker, L. (1989). The psychophysiological investigations of myocardial-ischemi a (PIMI) study objective, methods, and variability of measures. Psychosomatic Medicine, 60 (1), 56-63. Kielhofner, G., & Burke, J. (1980). A model of human occupation, Part 1. Conceptual framework and content. American Journal of Occupational Therapy, 34, 572581. Kielhofner, G. (1997). Conceptual Foundations of Occupational Therapy. Philadelphia, F. A. Davis Company. Kilander, L., Andren, B., Nyman, H., Lind, L., Boberg, M., & Lithell, H. (1998). Atrial fibrillation is an independent determinant of low cognitive function: a cross-sectional study in elderly men. Stroke, 29 (9), 1816-1820. Kirkpatrick, J., & Jamieson, M. (1993) A critical review of cognitive and memory assessment tools: Implications for occupational therapists working in cardiac rehabilitation. Occupational Therapy in Health Care, 8 (4), 19-45.
124 Knight, M. (2000). Cognitive ability and functional status. Journal of Advanced Nursing, 31 (6), 1459-1468. Lamar, M., Zonderman, A., & Resnic k, S. (2002). Contribution of specific cognitive processes to executive f unctioning in an aging population. Neuropsychology, 16 (2), 156-162. Landon, B., Reschovsky, J., Reed, M., and Blumenthal, D. (2000). Personal, organizational, and market level influe nces on physicians practice patterns. Medical Care, 39 (8), 889-905. Law, M., Cooper, B., Strong, S., Stewar t, D., Rigby, P., & Letts, L. (1996). The person-environment-occupation model: A tr ansactive approach to occupational performance. Canadian Journal of Occupational Therapy, 63, 9-23. Leidy, N. (1994). Functional status and the forward progress of merry-go-rounds: toward a coherent analytical framework. Nursing Research, 43 (4), 196-202. Lenze, E., Munin, M., Dew, M., Rogers, J., Seligman, K., Mulsant, B., et al., (2004). Adverse effects of depression and cognitive impairment on rehabilitation participation and recovery from hip fracture. International Journal of Geriatric Psychiatry., 19, 472-478. Lyons, K., & Crepeau, E. (2001). The clinical reasoning of an occupational therapy assistant. American Journal of Occupational Therapy 55, 577-581. MacNeill, S., & Lichtenberg, P. (1997). Home alone: The role of cognition in return to independent living. Archives of Physical Me dicine and Rehabilitation, 78, 755792.
125 Manias, R. (2005). How graduate nurs es use protocols to manage patients medications. Issues in Clinical Nursing, 14, 935-944. Mattingly, Cheryl (1991). Wh at is clinical reasoning? American Journal of Occupational Therapy, 45 (11), 979-986. Mast, B., MacNeill, S., & Lichtenberg, P. (1999). Geropsychological problems in medical rehabilitation: Dementia and depr ession among stroke and lower extremity fracture patients. Journal of Gerontology: Medical Sciences, 54A, M607-M612. McDermott, R. & Sarvela, P. (1999). Health education evaluation and measurement: a practi tioner's perspective.. Madison, Wisconsin: Brown & Benchmark. McDowd, J., Oseas-Kreger, D., & Filion, D. (1995). Inhibitory processes in cognition and aging. In Dempst er, F., Brainerd, C (Ed.), New perspectives on interference and inhibition in cognition (pp. 363-400). New York: Academic Press. Miller, D. (2000). Effectiveness of acut e rehabilitation services in geriatric evaluation and management units. Clinics in Geriatric Medicine, 16 (4), 775-782. Miller, D., & Salkind, N. (2002). Handbook of research design and social measurement (6th ed.). Thousand Oaks CA: Sage Publications. Moss, M., Wellman, A., Cotsonis, G. (2003). An appraisal of multivariate logistic models in the pulmonary and critical care literature. Chest, 123 (3), 923-928. Munro, B. (1997). Statistical Methods fo r Healthcare Research (3rd ed.). Philadelphia: Lippincott. Mitrushina, M., & Satz, P. (1994). Utility of mini-mental state examination in assessing cognition in the elderly. Aging, 6 (6), 427-432.
126 Nedley, A., Kendrick, B., & Brown, R. (1995). Health and memory in people over 50: a survey of a single-GP practice in England. Journal of Advanced Nursing, 21, 646-651. Penney, N., Kasar, J., Sinay, T. (2001). Student attitudes towa rd persons with mental illness: The influence of course work and Level I Fieldwork. American Journal of Occupational Therapy, 55 (2), 217-220. Petersen, R., Doody, R., Kurz, A., Mohs R., Morris, J., Rabins, P., et al. (2001). Current concepts in mild cognitive impairment. Archives of Neurology, 58, 19851992. Pisani, M., Redlich, C., McNico ll, L., Ely, E., Inouye, S. (2003). Underrecognition of preexisting cognitive impairment by physicians in older ICU patients. Chest 124 (6), 2267-2274. Poon, L., Fozard, J., Cermak, L., Are nberg, D., Thompson, L. (Ed.). (1980). New Directions in Memory and Aging: Proceed ings from the George A. Talland Memorial Conference.. Hillsdale, NJ: Lawrence Erlbaum. Rao, R., Jackson, S., & Howard, R. ( 1999). Neuropsychological impairment in stroke, carotid stenosis, and peripheral vascular disease: A comparison with healthy community residents. Stroke, 30, 2167-2173. Raz, N., Rodrigue, K., & Acker, J. (2003). Hypertension and the brain: Vulnerability of the prefrontal regions and executive function. Behavioral Neuroscience, 117 (6), 1169-1180.
127 Richards, M., Shipley, B., Fuhrer, R., Wadsworth, M. (2004). Cognitive ability in childhood and cognitive decline in mid-lif e: Longitudinal birth cohort study. British Medical Journal, dol:10,1136/bmj.37972.513819.EE (published 3 February, 2004). Rockwood, k., Rogers, J., & Masagatani, G. (1982). Clinical Reasoning of occupational therapists during the initial assessmen t of physically disabled patients. Occupational Therapy Journal of Research 2, 195-219. Ruchinskas, R. (2002). Rehabilitation therapists' recognition of cognitive and mood disorders in geriatric patients. Archives of Physical Me dicine and Rehabilitation., 83, 609-612. Ruchinskas, R. A., & Curyto, K. J. (2003). Cognitive screening in geriatric rehabilitation. Rehabilitation Psychology, 48 (1), 14-22. Ruchinskas, R., Broshek, D., Barth, J., Francis, J., & Robbins, M. (2000). A neuropsychological normative database for lung transplantation candidates. Journal of Clinical Psychology in Medical Settings, 7, 107-112. Ruchinskas, R., Singer, H., & Repetz, N. (2000). Cognitive status and ambulation in geriatric rehabilita tion: Walking without thinking? Archives of Physical Medicine and Rehabilitation, 82, 920-924. Sands, L., Yaffe, K., Covinsky,, K., Ch ren, M., Counsell, S., Palmer, R., Fortinsky, R., & Landefeld, C. (2003). Cognitive screening predicts magnitude of functional recovery from admission to 3 mont hs after discharge in hospitalized elders. Journal of Gerontology, 58A (1), 37-45.
128 Scheirton, L., Mu, K., & Lowman, H. ( 2003). Occupational therapists' responses to practice errors in physical rehabi litation settings. American Journal of Occupational Therapy, 1957(3), 307-314. Schell, B., & Cervero, R. (1993). Clinical reasoning in occupational therapy: an integrative review. American Journal of Occupational Therapy, 47 (7), 605-610. Schkade, J., & Schultz, S. (1992). Occupa tional adaptation: Toward a holistic approach for contemporary practice, part I. American Journal of Occupational Therapy, 46, 829-837. Seines, O., Goldsborough, M., Borowicz, L., & McKhann, G. (1999). Neurobehavioral sequelae of cardiopulmonary bypass. The Lancet, 353 (May 8), 16011606. Stoner, S. (1997). Drug-induced cognitive dysfunction. Medical Update for Psychiatrists, 2 (3), 81-85. Swan, G., DeCarli, C., Miller B., Reed, T., Wolf, P., Jack, L., et al. (1998). Association of midlife blood pr essure to late-life cognitive decline and brain morphology. Neurology 51 (4), 986-993. Tabachnknick, B., & Fidell, L. (2001). Using multivariate statistics (4th.ed.). Needham Heights, MA: Allyn and Bacon. Tabbarah, M., Crimmins, E., & Seeman, T. (2002). The relationship between cognitive and physical performance: MacArthur studies of successful aging. Journal of Gerontology: Medical Sciences, 57A (4), M228-M235.
129 Tatemichi, T., Desmond, D., Stern, Y., Paik, M., Sano, M., & Bagiella, E. (1994). Cognitive impairment after stroke: Freque ncy, patterns, and relationship to functional abilities. Journal of Neurology, Neur osurgery, and Psychiatry., 57 (202-207). Teng-Wai, D., Knopman, D., Geda, Y., Ed land, S., Smith, G., Ivnik, R., Tangalos, E., Boeve, G., & Petersen, R. (2003). Comparis on of the short test of mental status and the mini-mental state examination in mild cognitive impairment. Archives of Neurology, 60 (12), 1777-17781. Thomas, S. (2004). Using Web and Paper Questionnaires for Data-Based Decision Making: From Design to Interpretation of the Results. Thousand Oaks, CA: Corwin Press. Tideiksaar, R. (Ed.). (1997). Falling in old age (2nd. ed.). New York: Springer Publishing Co. Tombaugh,T. & McIntyre, J. (1992). The mini-menta l state examination: A comprehensive review. Journal of the American Geriatrics Society,40 (9), 922-935. Tracy, C., Dantas, G., Moineddin, R., Upshur, R. (2005). Contextual factors in clinical decision making. Canadian Family Physician 51. www.cfpc.ca/cfp Unsworth, C. (2005). Using a head-mount ed video camera to explore current conceptualizations of clinical reasoning in occ upational therapy. American Journal of Occupational Therapy 59 (1), 31-40. Van Boxtel, M., Buntinx, F., Houx, P., Mets emakers, J., Knottnerus, A., Jolles, J., et al. (1998). The relation between morbid ity and cognitive performance in a normal aging population. Journal of Gerontology. Series A, 53 (2), M147-154
130 Waldstein, S., Ryan, J., Muldoon, C., Shapir o, M., & Polefrone, J., et al., (1996). Hypertension and neuropsychological performan ce in men: Interactive effects of age. Health Psychology, 15 (2), 102-109. Walker, K. (2000). Adjustments to manage d health care: Pushing against it, going with it, and making the best of it. American Journal of Occupational Therapy, 55 (2), 129-137. Walonick, D. (2004). Survival statistics (On-line). Available: http://statpac.com/surveys/surveys.doc Wentzel, C., Hachinski, V., Hogan, D ., MacKnight, C., & McDowell, I. (2000). Prevalence and outcome of vascular cognitive impairment: Vascular Cognitive Impairment Investigators of the Ca nadian Study of Health and Aging. Neurology, 54 (2), 447-451. Whaley, M. (2000, September). Dissecting clinical reasoning. Paper presented at the Great Southern Occupational Therapy Symposium in Indianapolis, Indiana. Zamboni, V., Cesari, M., Zuccala, G., O nder, G., Woodman, R., Ranzini, M., et al. (2006). Anemia and cognitive performan ce in hospitalized older patients: results from the GIFA study. International Journal of Geriatric Psychiatry, 21 529-534.
132 Appendix A Summary of Research Questions Summary of research question #1 and associated survey items. What are the current practices of occ upational therapists, regarding screening or assessing the cognitive status of nondemented older patients referred for rehabilitation? a. Do therapists rou tinely screen/assess the cognitive status of non-demented elderly patients referred to rehabilitation on initial evaluation? Q 6 When you conduct initial evaluations of patients 65 years of age and older, how frequently do you assess cognition? b. How frequently do therapists use specific cognitive assessment tools on initial evaluation of elderly rehabilitation patients? Q 25 35 How frequently do you use the _____ screen or assessment as part of your initial evaluation of rehabilitation patients 65 years and older? c. If patients fail to impr ove as anticipated in the initial assessments, are therapists likely to assess their cognitive status? Q 74 After one week of treat ment, your patient's functional status (ADL's, wheelchair mobility, etc) does not seem to improve to meet your long term goa ls as you anticipated in your initial treatment plan. Which are you most likely to do? Summary of research question # 2 and associated survey questions. What effect does the practice context have on therapists decisions to assess the c ognitive status of non-demented elderly patients during their initial evaluation? a. Is the type of practice setting (acute hospital, subacute inpatient rehabilitation unit, home health agency, free-standing rehabilitation unit) associated with therapists' cognitive screening/assessment of elderly non-demented patients on initial evaluation? Q 78In which type of setting is your practice physically located? b. Is there an association between facility ownership (non-profit, for profit, VA) and therapists' cognitive screening/assessment of elderly non-demented patients during initial evaluation? Q 75 Please indicate the type of ownership of the agency or company where you are employed c. Is there an association between autonomy afforded therapists through the facility protocols, and their use of cognitive screening and assessment instruments? Q 55 If, in the course of treat ment, a rehabilitation patient is identified as having a cognitive impairment, my facility will be denied reimbursement. Q 63 If a patient is not learning in occupational therapy, I am required to discharge him or her. Q 72 Which of the following best describes how you determine which performance in areas of occupation (i.e. ADLs, IADLs, etc.) to assess on initial evaluation? Q 73 Which of the following best describes how you determine which performance skills (motor, process, etc.) to assess on initial evaluation?
133 Appendix A Summary of Research Questions (Cont.) Summary of research question #2 (Cont.) d. Is there an association between employer/supervisor support for conducting cognitive screenings/assessments, and available resources, and therapists use of cognitive screening/assessment instruments? Q 53 The company for which I work believes that assessing the cognitive status of all elderly patients on initial evaluation is a poor way to use billable units. Q 57 My employer requires use of cognitive assessments as part of our initial evaluati on and for discharge planning. Q 59 My supervisor encourages comprehensive assessment of our elderly patients, including cognitive status, as part of our initial OT evaluation. Q 60 Because of productivity requirements I lack the time to use standardized cognitive screens and assessments. Q 65 My facility has sufficie nt resources (screening and assessment instruments, sufficient funds for training) to support my use of standard cognitive screens and assessments.
134 Appendix A Summary of Research Questions (Cont.) Summary of research question #3 What effect does the personal context have on therapists' use of cognitive screening and assessment tools? a. Is there a relationship between therapists' knowledge of the effect of aging on cognition and their use of cognitive screening and assessment tools? Q 39Impairment in cognitive function is part of normal aging. Q 40 A knowledgeable therapist ca n effectively assess the cognitive status of a patient by conversing with the person. Q 42 A person's ability to perform ADLs is a good measure of his/her cognitive ability. Q 43 Fluid intelligence deteriorates with age. Q 45 Fluid intelligence affects the ab ility to learn new skills in therapy b. Is there a relationship between therapists' knowledge of how to administer and score a variety of screening and assessment tools and their use of these tools? Q 8 18 Received formal training (through classroom instruction or specialized continuing education c ourses in the administration and scoring of the following screening and assessment instruments. c. Are therapists' beliefs about professional responsibility associated with their use of cognitive screening and assessment instruments in geriatric rehabilitation? Q 38 Cognitive status should only be assessed by a psychologist or other licensed mental health professional. Q 47 Assessing cognitive status is th e responsibility of the occupational therapist. d. Is there an association between therapists' beliefs about use of cognitive screening and assessment instruments and frequency of use? Q 49 Assessing the cognitive status of elderly patients is problematic, b ecause it may lead to costly referrals for further evaluation. Q 50 Assessing cognitive status is problematic, because it is difficult to translate the information into treatment goals. Q 46 Assessing the cognitive status of elderly patients in every initial evaluation is a poor use of billable units. Q 36 A patient's cognitive status need only be addressed on initial evaluation if the individual carries a di agnosis of stroke, head injury, or Alzheimer's disease. 37 -Initial assessment should only address a patient's physical function. Q 41 Occupational therapy educati on should include extensive training in assessing the cognitive status of adults and elderl y patients.
135 Appendix A Summary of Research Questions (Cont.) Summary of research question #3 (Cont.) e. Are beliefs regarding aging associated with use of cognitive screening and assessment instruments in geriatric rehabilitation? Q 44 Non-compliance in older patient s is caused by lack of cognitive capacity. Q 51 Lack of motivation is the primary reason for non-compliance f. Is there an association between temporal factors of the therapists (age, years in OT practice, years in geriatric rehabilitation, length of time in employment at the time of the survey) and thei r use of cognitive screening and assessment tools in geriatric rehabilitation? Q 80 What is your age? Q 86 How many years have you worked in occupational therapy? Q 87 How many years have you wo rked in geriatric rehabilitation? g. Is level of education associated with use of cognitive screening/assessment instruments? Q 84 What is the highest degr ee you have attained to date?
136 Appendix B Expert Panel Members and Credentials Nancy Allen, OTRL Geriatric Occupational Therapist Tampa, Florida Larry Branch, Ph.D Professor, Health Policy and Management College of Public Health University of South Florida Professor, Aging and Mental Health Department of Aging and Mental Health University of South Florida Tampa, Florida Celinda Evitt, R.P.T Ph.D Associate Professor School of Physical Therapy University of South Florida Tampa, Florida Katherine Farber, OTRL Geriatric Occupational Therapist Tampa, Florida Deborah Gavin-Dreschnak, Ph.D Health Science Researcher VSN 8 Patient Safety Center James A. Haley V.A. Medical Center Tampa, Florida Sean Hillary, M.A. Director of Business Planning Tampa Electric Company Tampa, Florida Cecilia Jevitt, CNM, Ph.D Assistant Professor of Midwifery and Nursing University of South Florida Tampa, Florida Kim Warchol, OTRL President, Dementia Care Specialists Allen Cognitive Disabilities Expert Advisor Specialist in Dementia Care Chesterfield, Missouri Wendy Stav, Ph.D, OTRL, CDRS Research Assistant Professor Department of Occupational Therapy College of Public Health and Health Professions University of Florida Gainesville, Florida
137 Appendix C Expert Panel Instructions and Worksheets Dear Expert Panel member, This study looks at current screening and assessm ent practices in occupational therapy. Your assignment, if you should accept it, is to match the areas of interest to the researcher, to the questions in the survey instrument. Below, you will find information as to the two main categories and the sub-categories found in each. Please label each question as to which subcategory, in your opinion, it matches. Occupational Therapy Practice Environment 1. Type of practice site or setting 2. Facility ownership 3. Patient factors (patient characteristics whic h may influence the therapists assessment or treatment behavior). 4. Autonomy (degree of afforded therapists regarding assessment and treatment decisions and choices, fixed vs. flexible a ssessment and treatment protocols). 5. Resources (funding, availability of assessment tools, staffing patterns and levels, caseload, etc). 6. Supervisor/employee support. Therapists Personal Context 1. Demographic information (age, gender, race/ethnicity, education) 2. Therapists knowledge (of aging, disease, cognition) 3. Self-efficacy (how therapists perceive th eir level of skill in using screening and assessment instruments). 4. Therapists beliefs (regarding benefits Vs. disadvantages in using cognitive screening or assessment tools; professional responsibility). 5. Therapists attitudes regarding the use of cognitive screening and assessments. 6. Therapists values. 1. Provided physical rehabilitation to patients >65 yrs in past 6 months? (Yes, no). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 2. Percentage of patients 65 years or older in caseload past 6 months? ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
138 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 3. Which of the following assessments do you conduct in every initial evaluation of patients >65 referred to occupational therapy for rehabilitation? Select all that apply. Exclude patients who, at admission, have dx of stroke, Alzheimers disease, or head injury. (List of assessments provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 4. How familiar are you with the protocols for administering and scoring the following assessments? (List of assessments provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 5. Please indicate how frequently you use the following assessments as part of your initial evaluation of patients 65 years and older? (List of assessments provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 6. How would you rate the importance of the following characteristics in your selection of assessments? (List provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 7. Patient assessment should only address areas of physical function as indicated by the patients diagnosis. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
139 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 8. A patients cognitive status should only be assessed when the individual has a diagnosis of stroke, head injury, or Alzheimers disease. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 9. Cognitive status should only be assessed by a psychologist or other licensed mental health professional. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 10. Major impairment in cognitive function is part of normal aging. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 11. A knowledgeable therapist can eff ectively assess the cognitive status of a patient by conversing with the person. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 12. Occupational therapy education should include exte nsive training in assessing the cognitive status of adults and elderly patients. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
140 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 13. A persons ability to perform ADL s is a good measure of his/her cognitive status. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 14. Fluid intelligence deteriorates with age. (Choi ce from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 15. Fluid intelligence affects the ability to learn new skills in therapy. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 16. Non-compliance in older patients is generally caused by lack of cognitive capacity. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 17. Assessing cognitive status is useful for treatment. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
141 Appendix C Expert Panel Inst ructions and Worksheets (Cont. 18. Assessing cognitive status is a poor use of billabl e units. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 19. Assessing cognitive status is the responsibility of the occupational therapist. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 20. Therapists avoid assessing the cognitive status of patients, because they feel unsure of their assessment skills. (Choice from st rongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 21. Assessing the cognitive status of elderly patients is problematic, because it may l ead to costly referrals for further evaluation. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 22. Assessing cognitive status of elderly patients is problematic, because it is difficult to utilize the information in the treatment plan. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
142 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 23. Lack of motivation is the primar y reason why elderly patients are non-compliant. (Choice from strongly agree to strongly disagree). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 24. In my facility, assessment of cognitive function is done by the speech th erapist. (Choices are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 25. The company for which I work believes that assessing the cognitive status of all elderly patients is a poor way to use billable units. (Choices are yes, no.). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 26. If a rehab patient is identified as having a co gnitive impairment, my facility will be denied reimbursement for occupational therapy se rvices. (Choices are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support 27. My employer requires OTs use of cognitive assessm ents as part of our initial evaluation and for discharge planning. (Choi ces are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
143 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 28. My supervisor encourages comprehensive assessment of our elderly patients, including cognitive status, as part of our initial evaluation. (Choices are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 29. If a patient is not learning in occupational therapy, I am required to discharge him or her. (Choices are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 30. Because of productivity requirements, I lack th e time to use standardized cognitive screens and assessments. (Choices are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 31. In my facility, issues related to cognition are believe d to be social problems that we dont address. (Choices are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 32. My facility has sufficient resour ces (cognitive screening and assessment instruments, necessary supplies, or available funds) to support my use of standardized cognitive screens and assessments. (Choices are yes, no.) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
144 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 33. As you get ready to answer this question, think about how your current employer allows or expects you to deliver services to your patients. Then, on a 5 point scale from very dissatisfied to very satisfied, select your level of satisfaction with the treatment philosophy of the company for which your work. (Choice as described in question). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 34. On a scale from 1 (very poor) to 5 (very good), how would you rate the fit between your personal values and those of the company for which you work? (Choice as described in question). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 35. On a scale from 1 (very poor) to 5 (very good), how would you rate the fit between your professional values and those of the company for which you work? ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 36. How frequently do you utilize one particular theore tical perspective (frame of reference) in your practice? (Choice from almost always to almost never). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 37. If you utilize one particular theoretical perspective (frame of reference) in your practice, please indicate which one. (List of frames of references provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
145 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 38. If you indicated that you utilize a frame of reference other than those listed in the previous question, please indicate which one by typing in the space provided ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 39. Based on your personal pref erence, please indicate your first, second, and third choice of CE programs to attend. ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 40. Which of the following statements best describes how you determine which performance in areas of occupation (formerly known as performance areas) to assess on initial evaluation? (Please see questionnaire choices). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 41. Which of the following statements best describes how you determine which performance skills (formerly known as performance components) to assess on initial evaluation? (Please see questionnaire choices). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
146 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 42. The next two questions are case examples. Please indicate what you are most likely to do in these situations. On initial evaluation your patient was oriented and could follow two-step commands. However in the course of therapy, you find th e patient to be non-co mpliant with treatment recommendations and safety precautions. Which of the following are you most likely to do? (Please see questionnaire for choices). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs Appendix C Expert Panel Instru ctions and Worksheets (Cont.) ____Resources ____Attitudes ____Support ____Values 43. Case example #2 Your patient does not seem to learn the new skills you anticipated in your initial treatment plan. Your supervisor prefers not using billable units to assess cognition. Which of the following are you most likely to do? (Please see questionnaire for choices). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 44. Please indicate the type of ownership of the ag ency or compan y where you are employed. (Forprofit, not-for-profit, VA, independent contractor). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 45. In which state do you practice? (Drop-down menu). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
147 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 46. In the space provided please indicate how many OTs are on staff in your facility. ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 47. How many PRN or agency OTs work in your facility? (Space provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 48. How many OTAs are on staff in your facility? (Space provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 49. How many PRN or agency OTAs work in your facility? (Space provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 50. What is the average caseload in your OT department? (Ranges provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
148 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 51. In which type of setting is your primary practice physically located? (Choice of settings). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 52. If your practice is in a setting different from the se lection offered in the previous question, please indicate which one in the space provided below ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 53. Which of the following describes your current pos ition? (Please see questionnaire for response choices). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy Appendix C Expert Panel Instru ctions and Worksheets (Cont.) ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 54. How many years have you worked at this facility? (Choice of ranges provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 55. What is your age? (Space provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
149 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 56. What is your gender? (Male, female). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 57. Did you receive your OT trai ning in the US? (Yes, no). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 58. What is your race/ethnicity? (Choices provided) ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 59. What is the highest degree you have attained to date? (Choices provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge Appendix C Expert Panel Instru ctions and Worksheets (Cont.) ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 60. Do you have any special occupationa l therapy certifica tion? (Yes, no). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values
150 Appendix C Expert Panel Instru ctions and Worksheets (Cont.) 61. If you have any special certificati on, please indicate which in the space provided below. ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 62. How many years have you worked in occupational therapy practice? (Range provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values 63. How many years have you worked in geri atric rehabilitation? (Range provided). ____Practice environment ____Personal context ____Setting ____Demographic ____Ownership ____Knowledge ____Patient factors ____Self-efficacy ____Autonomy ____Beliefs ____Resources ____Attitudes ____Support ____Values Now that youve had a chance to look at the questionnaire, would you take a few minutes to answer some questions? 1. Is the questionnaire appealing? 2. Are the questions such that therapists would want to answer them? 3. Is the language understandable? 4. Is there anything offensive about this questionnaire? 5. Is there something important I forgot to ask? _______________________________ __________________________________________________________________________ 6. Is there a question I should eliminate? ____________________________________ __________________________________________________________________________ 7. Are the response choices adequate? ______________________________________ __________________________________________________________________________ Thank you for your valuable time and feedback. I could not do this without you. Mirtha M. Whaley, Ph.D Candidate, MPH, OTRL
151 Appendix D IRB Exemption Certificate
152 Appendix D IRB Exemption Certificate (Cont.)
153 Appendix E. Pilot Study Reliability Summary
154 Appendix E. Pilot Study Reliability Summary (Cont.)
155 Appendix E. Pilot Study Reliability Summary (Cont.)
156 Appendix E. Pilot Study Reliability Summary (Cont.)
157 Appendix E. Pilot Study Reliability Summary (Cont.)
158 Appendix E. Pilot Study Reliability Summary (Cont)
159 Appendix E. Pilot Study Reliability Summary (Cont.)
160 Appendix E. Pilot Study Reliability Summary (Cont.)
161 Appendix E. Pilot Study Reliability Summary (Cont.)
162 Appendix E. Pilot Study Reliability Summary (Cont.)
163 Appendix E. Pilot Study Reliability Summary (Cont.)
164 Appendix E. Pilot Study Reliability Summary (Cont.)
165 Appendix E. Pilot Study Reliability Summary (Cont.)
166 Appendix E. Pilot Study Reliability Summary (Cont.)
167 Appendix E. Pilot Study Reliability Summary (Cont.)
168 Appendix E. Pilot Study Reliability Summary (Cont.)
169 Appendix E. Pilot Study Reliability Summary (Cont.)
170 Appendix F IRB Certificate of Approval for URL
171 Appendix G Electronic a nd Postal Notifications
172 Appendix G Electronic and Po stal Notifications (Cont.)
173 Appendix G Electronic and Po stal Notifications (Cont.)
174 Appendix G Electronic and Po stal Notifications (Cont.)
175 Appendix G Electronic and Po stal Notifications (Cont.)
176 Appendix G Electronic and Po stal Notifications (Cont.)
177 Appendix G Electronic and Po stal Notifications (Cont.)
178 Appendix G Electronic and Po stal Notifications (Cont.)
179 Appendix G Electronic and Po stal Notifications (Cont.)
180 Appendix G Electronic and Po stal Notifications (Cont.)
181 Appendix G Electronic and Po stal Notifications (Cont.)
182 Appendix G Electronic and Po stal Notifications (Cont.)
183 Appendix G Electronic and Po stal Notifications (Cont.)
184 Appendix H IRB Certificate of Appr oval for Incentive and Yahoo Account
185 Appendix I Distribution of Re spondents by SIS Membership Special Interest Section Frequency Percent Gerontological 208 64.6 Home and Community Health 95 29.5 *Physical Disabilities 10 3.1 *Developmental Disabilities 1 0.3 *Education 1 0.3 *Administration and Management 1 0.3 *Technology 1 0.3 *Work Programs 1 0.3 *Missing 4 1.2 Total 322 100.0 *Denotes participants who did not meet the inclusio n criteria and were therefore excluded from analyses
186 Appendix J Bivariate Analysis Chi Square Variable Chi Square Df P value Gender 0.706 1 .401 Education 0.085 1 .770 Setting 2.995 2 .224 Ownership 1.891 2 .388 Special Interest Section 6.277 1 .012* Responders 0.427 1 .514 Percentage of Caseload > 65 years of age 0.908 1 341 Years OT Practice 8.435 3 .038* Years Geriatric Rehab 8.975 3 .030* Years with Company 4.088 2 130 significant at p < .05
187 Appendix K Summary of Logistic Regressions Summary of Logistic Regression Analyses To Dete rmine Associations Between Predictor Variables and Outcome Adjusted for Age, Education, and Gender Variable Odds Ratio P value 95% Confidence Interval Lower Upper Autonomy to determine areas and skills to assess 1.466 .173 .845 2.542 Lack time to assess because of productivity 1.510 .143 .870 2.620 Employer requires assessment 3.537 .000** 1.947 6.426 Support from supervisor .714 .359 .348 1.467 Sufficient resources in facility 2.735 .000** 1.579 4.738 Company believes assessment is poor use of units 1.942 .311 .538 7.010 Knowledge of decline in fluid intelligence 1.720 .057 .984 3.006 Impairment in cognitive function normal aging 1.124 .680 .646 1.954 Knowledgeable therapists can assess conversing 1.429 .181 .847 2.409 Ability to perform ADLs good measure 1.137 .644 .660 1.957 Fluid intelligence affects ability to learn .855 .608 .471 1.555 Knowledge of a variety of assessment instruments 1.254 .002** 1.084 1.451 Lack of motivation primary reason for noncompliance 1.288 .470 .648 2.561 Assessment can lead to costly referrals .812 .735 .244 2.705 Therapist believes is poor use of Tx units 3.201 .135 .697 14.695 Assessment if indicated by diagnosis 1.759 .489 .355 8.708 Assessment should only address physical status .542 .436 .116 2.528 Difficult to incorporate information in treatment plan 1.853 .082 .924 3.718 If cognitive impairment facility denied reimbursement .391 .510 .024 6.386 OT education should include training in use of assess. 1.462 .577 .384 5.563 Assessment is responsibility of OT 1.738 .217 .723 4.177 Assessment done by other licensed professional 1.639 .166 .815 3.295 Social Desirability Scale 1.139 .259 .909 1.428 *p < .05. **p < .01
188 Appendix L Summary of Res ponses to Belief Questions Beliefs Frequency % Strongly or Moderately Agreed % Strongly or Moderately Disagreed Non-compliance in older patients is caused by lack of cognitive capacity 301 26.9 73.1 Lack of motivation is the primary reason for noncompliance in older adults 303 20.2 79.8 Assessing cognition is probl ematic because it may lead to costly referrals for further evaluation 302 4.6 95.4 Assessing cognition is problematic because it is difficult to translate the information into the treatment plan 302 20.5 79.5 Assessing cognitive status of elderly patients on every initial evaluation is a poor use of billable units 303 5.3 94.7 Cognitive status need only be assessed on initial evaluation if the patient carries a diagnosis of stroke, Alzheimer's disease, or head injury 303 3.4 96.6 Initial evaluation only address a patient's physical function 303 2.4 97.6 OT education should include extensive training in assessing the cognitive status of older patients 303 95.7 4.3 Assessing cognitive status is the responsibility of the occupational therapist 303 87.0 13.0 Cognitive status should onl y be addressed by a psychologist or other licensed mental health professional 303 20.8 79.2
189 Appendix M Temporal Vari ables of Participants Characteristic Frequency Percent Age 24-29 48 16.1 30-35 44 14.8 36-41 46 15.4 42-47 49 16.4 48-53 55 18.5 54-59 39 13.1 >65 17 5.7 Years in OT Practice <1 3 1.0 2-5 61 20.2 6-10 64 21.2 11-15 31 10.2 >15 143 47.4 Years in Geriatric Practice <1 9 3.0 2-5 77 25.6 6-10 68 22.6 11-15 55 18.3 >15 92 30.6 Years with Company <1 47 15.6 2-5 135 44.7 6-9 60 19.9 >10 60 19.9 Spearman Correlations Between Ther apists Age and Temporal Values Temporal variable Spearman rho Sig. N Years Occupational Therapy practice .746 .000** .302 Years geriatric rehabilitation .648 .000** .301 Years with company .421 .000** 302 ** Correlation is significant at the 0.01 level (2 tailed)
190 Appendix N Summary of Logistic Regressions Model 1 Full Model of Odds Ratio Estimates for Covariates (13 predictor variables) of Use of Cognitive Assessment Instruments Variable Odds Ratio 95% CI P-value Age 0.016 1.02 0.98 1.06 0.437 Gender 0.288 1.33 0.47 3.76 0.586 Education (Post baccalaureate) -0.128 0.88 0.46 1.69 0.699 Case example 1 (charge assessment) -0.051 0.95 0.34 2.65 0.923 Case example 2 (charge treatment) -0.646 0.52 0.18 1.57 0.249 Knowledge of assessment instruments 0.174 1.19 1.00 1.42 0.052 Employer requires assessment 1.145 3.14 1.56 6.32 0.001 Knowledge of age decline in fluid intelligence 0.712 2.04 1.05 3.95 0.035 Assessment not difficult to incorporate 0.411 1.51 0.67 3.39 0.321 Facility has sufficient resources 0.854 2.35 1.20 4.59 0.012 Gerontological SIS 0.666 1.95 0.92 4.12 0.082 Years OT experience 0.123 1.13 0.67 1.95 0.657 Years geriatric experience 0.272 1.31 0.82 2.11 0.263 Model 2 Odds Ratio Estimates for Covariates (12*) of Use of Cognitive Assessment Instruments Variable Odds Ratio 95% CI P-value Age 0.016 1.02 0.98 1.06 0.801 Gender 0.288 1.33 0.47 3.76 0.586 Education (Post baccalaureate) -0.127 0.88 0. 46 1.69 0.701 Case example 2 (charge for treatment) -0.603 0.55 0.28 1.08 0.800 Knowledge of assessment inst ruments 0.173 1.19 1.00 1.42 0.052 Employer requires assessment 1.145 3.14 1.56 6.32 0.001 Knowledge of age decline in fluid intelligence 0.711 2.04 1.05 3.95 0.035 Assessment not difficult to incorporate 0.413 1.51 0.67 3.40 0.318 Facility has sufficient resour ces 0.854 2.35 1. 20 4.59 0.012 Gerontological SIS 0.666 1.95 0.92 4.12 0.082 Years OT experience 0.124 1.13 0.66 1.95 0.655 Years geriatric experience 0.271 1.31 0.82 -2.11 0.264 Predictor variable Case Example charge for cogn itive assessment ( p value 0.923) deleted from analysis
191 Appendix N Summary of Logi stic Regressions (Cont.) Model 3 Odds Ratio Estimates for Covariates (11*) of Use of Cognitive Assessment Instruments Variable Odds Ratio 95% CI P-value Age 0.015 1.02 0.98 1.06 0.466 Gender 0.289 1.34 0.47 3.76 0.584 Case example 2 (charge for treatment) -0.611 0.54 0.28 1.07 0.076 Knowledge of assessment inst ruments 0.168 1.18 1.00 1.41 0.056 Employer requires assessment 1.153 3.17 1.58 6.36 0.001 Knowledge of age decline in fluid intelligence 0.722 2.06 1.07 3.98 0.032 Assessment not difficult to incorporate 0.432 1.54 0.69 3.44 0.293 Facility has sufficient resour ces 0.846 2.33 1. 19 4.55 0.013 Gerontological SIS 0.645 1.91 0.91 4.00 0.088 Years OT experience 0.133 1.14 0.67 1.96 0.630 Years geriatric experience 0.267 1.31 0.81 210 0.270 *Predictor variable post-baccalaureate educati on (p value .701) deleted from the analysis Model 4 Odds Ratio Estimates for Covariates (10*) of Use of Cognitive Assessment Instruments Variable Odds Ratio 95% CI P-value Age 0.02 1.02 0.99 1.06 0.27 Gender 0.253 1.29 0.46 3.58 0.628 Case example 2 (charge for treatment) -0.614 0.54 0.28 1.06 0.075 Knowledge of assessment inst ruments 0.163 1.18 0.99 1.40 0.062 Employer requires assessment 1.169 3.22 1.61 6.44 0.001 Knowledge of age decline in fluid intelligence 0.722 2.06 1.07 3.98 0.032 Assessment not difficult to incorporate 0.445 1.56 0.70 3.49 0.277 Facility has sufficient resour ces 0.852 2.34 1. 20 4.57 0.012 Gerontological SIS 0.651 1.92 0.91 4.03 0.085 Years geriatric experience 0.35 1.42 1.01 -1.99 0.042 Predictor variable years of occupational therapy exper ience (p value .630) deleted from the analysis.
192 Appendix N Summary of Logi stic Regressions (Cont.) Model 5 Odds Ratio Estimates for Covariates (9*) of Use of Cognitive Assessment Instruments Variable Odds Ratio 95% CI P-value Age 0.02 1.02 0.99 1.06 0.263 Case example 2 (charge for treatment) -0.622 0.54 0.27 1.05 0.071 Knowledge of assessment inst ruments 0.159 1.17 0.99 1.39 0.067 Employer requires assessment 1.17 3.22 1.61 6.45 0.001 Knowledge of age decline in fluid intelligence 0.712 2.04 1.06 3.93 0.034 Assessment not difficult to incorporate 0.456 1.58 0.71 3.52 0.265 Facility has sufficient resour ces 0.877 2.4 1. 24 4.65 0.009 Gerontological SIS 0.65 1.92 0.91 4.02 0.086 Years geriatric experience 0.346 1.41 1.01 1.98 0.043 *Predictor variable gender (p va lue .628) deleted from the analysis Model 6 Odds Ratio Estimates for Covariates (8) of Use of Cognitive Assessment instruments Variable Odds Ratio 95% CI Pvalu e C ase example 2 (charge for treatment) -0.611 0.54 0.28 1.06 0.075 K nowledge of assessment instruments 0.144 1.16 0.98 1.36 0.091 E mployer requires assessmen t 1.174 3.23 1.62 6.46 0.001 K nowledge of age decline in fluid intelligence 0.682 1.98 1.03 3.80 0.040 A ssessment not difficult to incorporate 0.485 1.63 0.73 3.61 0.233 F acility has sufficient resour ces 0.896 2.45 1. 27 4.72 0.008 G erontological SIS 0.62 1.86 0.89 3.89 0.100 Y ears geriatric experience 0.453 1.57 1.19 2.09 0.000 *Predictor variable age (p value .263) deleted from the analysis
193 Appendix N Summary of Logi stic Regressions (Cont.) Model 7 Odds Ratio Estimates for Covariates ( 7*) of Use of Cognitive Assessment instruments Variable Odds Ratio 95% CI P-value Case example 2 (charge for treatment) -0.702 0.50 0.26 0.96 0.036 Knowledge of assessment inst ruments 0.153 1.17 0.99 1.38 0.071 Employer requires assessment 1.168 3.22 1.62 6.39 0.001 Knowledge of age decline in fluid intelligence 0.644 1.91 1.00 3.63 0.050 Facility has sufficient resour ces 0.913 2.49 1. 30 4.79 0.006 Gerontological SIS 0.586 1.80 0.86 3.74 0.118 Years geriatric experience 0.463 1.59 1.20 2.10 0.001 *Predictor variable assessment not difficult to incorporate into treatment plan (p value .233) deleted from the analysis Model 8 Odds Ratio Estimates for Covariates ( 6*) of Use of Cognitive Assessment instruments Variable Odds Ratio 95% CI P-value Case example 2 (charge for treatment) -0.696 0.5 0.26 0.96 0.036 Knowledge of assessment instruments 0.17 1.19 1.01 1.40 0.041 Employer requires assessment 1.222 3.39 1.71 6.73 0 Knowledge of age decline in fluid intelligence 0.691 2 1.05 3.78 0.34 Facility has sufficient resources 0.961 2.62 1.37 5.00 0.004 Years geriatric experience 0.413 1.51 1.15 1.98 0.003 *Predictor variable SIS (p value 0.118) deleted from the analysis
194 Appendix O Summary of Regression Diagnostics Case # Pred Prob Analog Cooks Stat Leverage Value Logit Res Student Res S Standard Res Deviance Value DFB Const DFB Case1 DFB Case2 DFB Fluint age DFB Employ req FB acsuff DFB Years gerrehab 51 07687 .11245 .00928 13.00941 .27583 3 .46546 .26525 07458 .06977 -.00790 -.04161 .02604 .03845 -.00223 77 12612 .11926 .01692 7.92886 .05239 2 .63227 2.03495 10575 .02299 .00945 -.05238 .03691 .04707 -.02597 88 12355 .09409 .01309 8.09382 .05856 2 .66342 2.04504 13859 .03419 -.01604 -.03821 .02282 .02650 -.00506 89 10856 .11212 .01347 9.21134 .12168 2 .86554 2.10734 13814 .02577 .00334 -.05251 .03753 .04283 -.02908 145 12298 .13580 .01869 8.13170 .06673 2 .67052 2.04732 04819 .06382 -.01449 .05585 .02778 .01093 .00049 182 12355 .09409 .01309 8.09382 .05856 2 .66342 2.04504 13859 .03419 -.01604 -.03821 .02282 .02650 -.00506 275 11176 .11617 .01441 8.04787 .10877 2 .81920 2.09352 01241 06779 -.00496 -.03491 .01803 .03929 .01513 285 11176 .11617 .01441 8.04787 .10877 2 .81920 2.09352 01241 06779 -.00496 -.03491 .01803 .03929 .01513 Expected value of leverage = k+1/N (k=no. of predictors, N=sample size). For this an alysis, expected value of leverage = 6+1/2 71 = 0.026. Range of values for these cases is from 0.009 to 0.02; leverage close to 0 indicates no undue influence by any case.; 95% of cases shoul d have studentized residuals, standardiz ed residuals, and deviance values that lie within + or 2; 99% of cases should have values that lie between + or 2.5. All cases above (~3% of the sample) have studentized residuals ranging f rom 2.05 to 2.28 and deviance ranging from 2.03 to 2.26, so they are below 2.5 and are not cause for concer n. Although, all cases have standardized residuals >2.5 and case #51 exceeds 3.0, al l DFBetas and Cooks values ar e <1, indicating no case has undue influence on the model.
About the Author Mirtha Montejo Whaley received a Bachelors Degree in Occupational Therapy from the University of Flor ida in 1968 and for over 30 year s engaged in the psychosocial rehabilitation of chronically ill adult individual s. During the last ten years of her clinical practice, she returned to physical rehabil itation and became intere sted in the cognitive function of older medical patients. In 1993, Mrs. Whaley obtained a Masters Degree of Public Health from the University of South Fl orida, and later returned to obtain a Ph.D in Public Health, which she completed in 2007. In addition to her rehabilitation experi ence, Mrs. Whaley ha s provided clinical training to occupational therapy students, and instructed undergraduate students at USF. She is a certified Allen Cogn itive Advisor, and her research interests include cognition and function, caregiver educa tion and preparedness, cultural competence, successful aging, preventing excess disability in nursing homes, aging in place, and quality in long term care.