Dossa, Almas MS, PT, PhD, MPH; Capitman, John A. PhD
Meeting health and social service requirements of disabled elders still represents an increasing challenge for the U.S. due to increased numbers of elders in the coming years even though recent research reports disability rate decline among elders.1,2 Several randomized controlled trials and dissemination studies using disability prevention models such as the Chronic Disease Self-Management Programs and the Health Enhancement Program (HEP; now known as Enhance Wellness; www.projectenhance.org) for elders with chronic disease are reported to improve functional outcomes, improve exercise tolerance, decrease hospitalization, and improve quality of life.3–9 Translation of evidence-based research from efficacy studies into real-world settings is often difficult, and dissemination studies on the HEP have shown lower functional outcomes compared with the original trials.9,10 Despite the importance of these evidence-based disability prevention programs, little is known about the structure and process of care differences between original clinical trials and their replication in the real-world settings, and the implications of these differences for elder clients. In addition, studies on the influences of client factors and organizational site features on functional outcomes are minimal for these program replications.
In this study, we had 2 goals: first, we wanted to compare 12-month functional change differences and differences in structures and processes between the HEP replication (an ongoing dissemination of the clinical trial) and the original trial, and compare the structures and processes of care in both studies. Second, we wanted to examine the association of client characteristics (self-efficacy) and functional outcomes for the replication, and to explore the association of site features such as site size and location and 12-month functional outcomes for the replication. We expected to find lower functional changes, less resources, and less intense processes in the replication as compared with the original trial. We expected higher client baseline self-efficacy, smaller site size, and urban locations to lead to better functional outcomes for the replication.
The HEP, a component of the Senior Wellness Project, now known as Project Enhance, was developed by the Northshore Senior Center and researchers from the University of Washington, Seattle. Project Enhance aims to provide an evidence-based, cost-effective, high-quality, comprehensive program for elders with chronic disease, and is a model of community-academic collaboration of pragmatic research. On the basis of years of evidence-based research, Project Enhance has been implemented in more than 200 sites nationwide, including senior centers, assisted living centers, hospitals, and continuing retirement communities. Conducted by nurses and social workers in senior centers, the HEP program was designed to promote health and functioning of community elders with chronic disease who are at risk for functional decline. It relies on collaboration between community agencies and medical care providers to provide appropriate health promotion activities for elders with substantive health issues. It consists of comprehensive health reviews, functional assessments, creation of action plans, and use of support groups.3,8 Clients return within 2 weeks following initial assessment to review plans, then 6 and 12 months later for reassessments. The Health Mentor program, a component of the HEP, matches HEP clients to trained peer mentors who provide follow-up calls and support for the health action plan.11 In addition, clients are encouraged to enroll in other evidence-based program components: an evidence-based exercise class and the Chronic Disease Self-Management Program, a 6-week program including guidelines on goal setting, exercise, medications, coping, symptom management, and communication with other health professionals, and other support groups.3,6 Thus, any research on components of this project will be of value in improving the quality of life for community-living older adults.
Effective dissemination and implementation of evidence-based practices into community settings is challenging, and researchers struggle with the processes involved with translating and implementing these practices.12 Innovative adoption (the decision to use an innovation) differs from excellent innovation implementation (consistent, skillful, and committed use of an innovation).13 Even though a procedure may be adopted by communities and teams, it does not always mean that the procedure will be successfully implemented. Examples of types of barriers to implementation of evidence-based research include organizational constraints (ie, lack of time), financial constraints, resources such as staffing, and knowledge and attitudes of providers.14,15
Client factors, including high self-efficacy, are significantly associated with health outcomes.4,16 According to social cognitive theory, baseline self-efficacy appears essential for managing health regimens and setting disease management goals.17 High pain self-efficacy beliefs have been predictive of better physical functioning, less disability, and less depression. In addition, high self-efficacy predicts smoking cessation, diet guideline adherence, and exercise adherence. Most of the research, however, has focused on medication and exercise adherence, and not on self-management programs such as the HEP. The original HEP trial and dissemination studies3,8 did not focus on self-efficacy as a predictor of health outcomes, even though HEP sites were measuring self-efficacy at the time this study was conducted. Baseline client self-efficacy may be even more important for functional outcomes in the dissemination of clinical trials because of possible lack of resources such as time and staffing.
Organizational predictors also influence health outcomes. Although older adults use community-based organizations (ie, senior centers) more than any other community-based elderly service,18 organizational analysis in the health care field primarily focuses on larger institutions such as hospitals and nursing homes. Hospital research shows that smaller site size and urban locations are associated with better health outcomes.19–21 The original clinical trial for the HEP was done in a large rural site. Little is known about the association of site size and urban versus rural locations and functional outcomes in the program dissemination.
We examined secondary longitudinal data from the full 2002–2004 client database from the replication dissemination, which was being conducted in 22 sites in the United States. This database is owned by the University of Washington, Seattle. In addition, using a cross-sectional design, we collected primary data on site features, structures, and protocol processes through telephone interviews between February and June 2005. Informants included 18 site managers, 20 nurses, and 23 social workers. We conducted this study with the full convenience sample of providers from all 22 HEP sites located in the United States at that time (14 sites in Washington State, 5 sites in Michigan, 1 site in Maine, and 2 sites in New York). We obtained Institutional Review Board approvals from Brandies University and the University of Washington, Seattle. We compared the pre-post data from this replication to the pre-post data from the clinical trial. The clinical trial took place in 1997 with 201 clients located in 2 independent health maintenance organizations and a large senior center.3 Patients aged 70 years and more were recruited through medical practices and the program was led by a geriatric nurse practitioner. The senior center site was also included in the replication study as 1 of the 22 sites. The replication study consisted of 719 clients. Patient characteristics including demographics, chronic conditions, and health status were similar for the clinical trial and replication.
Dependent Measures (Functional Outcomes)
Exercise “Readiness” or Exercise Stage at 12 Months
This consisted of the PACE (Physician-based Assessment and Counseling for Exercise) scale to evaluate physical activity, attitude, and exercise readiness.22,23 The PACE scale was developed and validated by the Centers for Disease Control and Prevention.24 This ordinal scale consists of questions that best describes the current level of exercise with possible answers ranging from 1 (“I do not exercise or walk regularly now and do not intend to start in the near future”) to 11 (“I do vigorous exercise 6 times per week”). Table 1 shows description of each score. Long et al22 suggest that a score of 4 or less corresponds to exercising no more than infrequently.22 On the basis of this, we used the scale to form a dichotomous variable: 0 (infrequent exercise: collapsing the first 4 categories) and 1 (active or frequent exercise: 5 and above).
Activities of Daily Living Sum at 12 Months
We measured activities of daily living (ADL) ability using the physical functions subscale of the Medical Outcomes Study Short Form-36 (SF-36). This continuous variable measures role limitations caused by difficulty with physical functioning. The SF-36 is widely used, and has been validated with older adults.25,26 Clients are asked how much difficulty they have for 9 usual activities: bathing and showering, dressing, using the toilet, moving in and out of bed, feeding, climbing one flight of stairs, climbing several flights of stairs, walking one block, and walking several blocks. The scale has 5 response options: from 0 (unable to do), 1 (did with a lot of difficulty), 2 (did with some difficulty), (3 did with a little difficulty), and 4 (did with no difficulty).
The main independent client variable was self-efficacy.27 Because self-efficacy has been shown to be significantly associated with health outcomes, our goal was to determine the association between self-efficacy and functional outcomes for the replication study.4,16 In addition, the clinical trial did not examine the association between self-efficacy and client outcomes. This validated self-efficacy scale consists of 6 questions, focusing on the client's confidence in doing certain tasks (Cronbach α = 0.86).28 Possible responses range from 0 (not at all confident) to 9 (extremely confident). Three questions focus on exercise (eg, “how confident are you that you can do an aerobic exercise such as walking, swimming, or bicycling 3 to 4 times a week?”) and three questions focus on asking the doctor questions or concerns (eg, “how confident are you that you can ask your doctor things about your illness that concern you?”).
Control baseline client variables that may affect functional outcomes (Table 2) included demographics, validated measures of health status,26 depression,29 functional status, walking status, the inability to shop/cook,22,26,30 number of prescription drugs (a proxy used for chronic disease status), marital status, and body mass index. Client data were recorded by nurses at baseline, and at 6- and 12-month reassessments. We were unable to include variables that reflected socioeconomic status such as education and income in our analysis because these variables had 40% or more missing data.
Site features included 2 dichotomous variables: site type (urban versus rural; rural category included suburban and rural sites) as defined by site managers and site size (large vs small = 0), based on senior average daily attendance, including those not in the HEP. We defined small size as 65 or less participants and large size as more than 65 participants.
Replication Structures and Processes
We asked respondents about methods of client recruitment, presence of program champions, percentage of primary care practitioner (PCP) referrals, health mentor program availability, and client follow-up processes. For program champion presence, we asked whether there was onsite presence of someone other than those conducting the program, who believed strongly in the program and promoted it to clients and other health personnel. To find out about recruitment, we asked about specific methods including mailings to community centers and to providers, public services announcements, community presentations, and so forth. For client follow-up processes, we asked whether clients were followed up to remind them about their appointments via phone or mail.
We obtained information on the functional outcomes changes from the original trial,3 and used the secondary database to determine functional changes for the replication. We used correlation coefficients between independent variables to test for multicollinearity. We conducted descriptive analyses of demographic and health status variables, tests of normality, and bivariate analyses using χ2 tests or logistic regressions to test independent variable and functional outcome association for the replication. We analyzed associations between baseline self-efficacy and ADL status at 12 months using multivariate models, and between baseline self-efficacy and exercise stage status at 12 months using logistic models. Self-efficacy for one behavior may not predict self-efficacy for another task.17 Therefore, we used the full version of this scale as well as the exercise self-efficacy subscale as independent variables in 2 separate models.
We controlled for the other independent client variables, and used baseline ADL and exercise stage status in the model. We examined the association between site features and functional outcomes by adding high site size and urban location one at a time leaving out low site size and rural location as reference. The final models used generalized estimating equations (GEE) regression models. Generalized estimating equations are methods of parameter estimation for correlated data.31 When data are collected on the same units across successive points in time, there exists a correlation between observations on individual clients. If this correlation is not taken into account, the standard errors of the parameter estimated will not be valid and hypothesis testing results will be nonreplicable. Generalized estimating equations models were developed to account for these correlated error issues. The Statistical Analysis Systems (SAS), version 9.1 (Cary, NC) was used for all analyses.
Descriptive Data and Preliminary Analyses
Ten sites with less than 10 clients participating were excluded from analysis. The final data set had 18 sites with 719 clients. Table 2 summarizes baseline characteristics of clients. The socioeconomic variables, income and education were not included in the models because of missing data. Fifty three percent of the sample were college graduates or had some college education, 12% went to graduate school, 31% were high school graduates or had some high school education, and about 4% had less than high school education. Sixty eight percent of the sample had low income, 16% had a moderate income, and 16% had a high income. We classified income (for a 1-person household) as follows: low income, less than $27 250; moderate income, $27 251 to $38 100; high income, more than $38 500. For multiperson households, categories were as follows: low income, less than $45 200; moderate income, $45 201 to $63 100; and high income, more than $63 100.
Forty two percent of sites were urban, and 56% of sites were small. Of the 719 clients who attended the initial session, 292 clients (41%) attended the 12-month session. All providers agreed to the primary data collection, except for 1 nurse and 1 site manager who were not interested. All providers except for 1 nurse and 2 site managers were women, with age ranging from 39 to 65 years. All were Caucasian, except for 1 nurse who was Asian. Correlation coefficients as measured by Pearson's correlation coefficient and φ coefficient between independent variables were all below r = 0.5.
Functional Change Differences in Clinical Trial Versus Replication Study
The average functional change over 12 months for the original HEP trial3 was greater than the change seen in the replication study (Table 3). Scores for the ADL sum for the original trial were calculated differently than that of the replication, but the scale used was the same. We did not have data on baseline values of exercise stage for the clinical trial or standard deviation for the change.
Comparison of Structures and Processes of Original Trial and Replication
In the original HEP trial, invitation letters were sent to clients by their physicians; in the replication, only a few physicians sent invitation letters. Additionally, in the replication, most of the older adults participants self-referred to the program, or were approached by a nurse or a social worker. In the original HEP trial, a nurse conducted an average of 3 follow-up visits in addition to the 6- and 12-month visits, and conducted a mean number of 9 follow-up phone calls. In the replication, follow-up visits varied and were inconsistent, and phone call follow-up was not part of the protocol. Baseline client variables including age and self-rated health were similar for clients in the original trial and the replication. In the replication, 71% of sites did not have a program champion, 17% of sites had only a nurse or social worker but not both, 61% of sites did not have consistent mentor programs, and 71% did not have PCPs involved in recruitment. This differed from the clinical trial, which was fully staffed, had a program champion, PCP involvement, and a mentor program.
Functional Outcomes and Self-efficacy
Baseline self-efficacy was significant for GEE models for 12-month ADL sum and 12-month exercise stage. The first GEE model (Table 4) shows associations between baseline client variables with the outcome being the 12-month ADL sum, after controlling for other client variables. With each 1-level increase in self-efficacy, the 12-month ADL sum estimate increased by 0.09 (P ≤ .001). For clients with arthritis, the estimate decreased by 1.41 compared with clients without arthritis (P ≤ .01). The second GEE model (Table 4) shows associations for baseline client variables with the outcome being exercise stage 12 months (1 = frequent exercise, 0 = infrequent exercise). Controlling for the effects of the other variables, the odds for being at a frequent exercise stage at 12 months was slightly higher with each 1-level increase in client self-efficacy (odds ratio [OR], 1.02; P ≤ .05). In addition, the odds for being at a frequent baseline exercise stage at 12 months was higher for those with better baseline health (OR, 1.84; P ≤ .01), and for clients at walk status greater than 1 block at baseline (OR, 3.0; P ≤ .01), and lower for those with baseline inability to shop and cook (OR, 0.33; P ≤ .001). Results were similar for models with full self-efficacy scale and the self-efficacy subscale.
Functional Outcomes and Site Variables
Site location and site size was significant for the 12-month ADL sum model (Table 5); model I shows that the estimate for 12-month ADL was 0.95 more for urban sites than for rural sites (P ≤ .05), and model II shows that the estimate for 12-month ADL sum was 1.18 less for larger sites than for smaller sites (P ≤ .01). Site location was also significant for 12-month exercise stage (Table 6); clients in urban sites were 14% more likely to be at a frequent exercise stage compared with clients in rural sites (P < .001). Postanalyses, we also conducted χ2 tests between size site and site location to determine if there were associations between these variables that could potentially lead to confounding results. Only 3 sites were both urban and small, and there were no significant associations between site size and site type.
Our results showed that functional changes achieved in the replication were less than those of the original clinical trial over a 12-month period. Replication sites had fewer available structures or resources, and processes were less intense as compared with those of the clinical trial. We found that baseline high self-efficacy was associated with higher ADL and exercise stage outcomes at 12 months. Clients in urban sites had better ADL outcomes and were at a more advanced exercise stage level at 12 months compared with those in rural sites. Clients in smaller sites had better ADL outcomes at 12 months than those in larger sites.
Our findings of lower functional changes in the replication study than in the original trial is supported by other HEP dissemination studies that have also shown lower exercise stage changes compared with the original trial.8–10 This HEP replication project is an excellent example of how a “real life” situation can differ from rigorous clinical trials. Many differences existed in process of care, such as client recruitment and client follow-up between the original HEP 1-year randomized controlled trial involving 201 older adults3 and the replication study. This could account for better functional outcomes in the original trial than in the replication. Intensity of these processes of care was lower in the replication. Similarly, structures such as program champions, staffing, PCP involvement in recruitment, and health mentor program availability were lower and less consistent for the replication as compared with the clinical trial. Although baseline client variables for both studies were similar, we did not have information about baseline self-efficacy for the original trial. Possibly, lower levels of self-efficacy in clients in the original trial could have resulted in greater room for improvement and more meaningful functional changes; we do not know whether clients in this study had the same or better baseline self-efficacy.
Translation of evidence-based research into practice can be unpredictable and is a slow process.32 Many of the providers in the replication study discussed the difficulties of funding, resources, and time. Any new protocol's success depends on how well it is implemented, and how consistently the protocol is followed during implementation as well as on organizational attributes, such as availability of resources (ie, funding, human resources, and internal processes).33 The intensity of interventions followed in a clinical trial can be a barrier to translation of efficacious evidence-based programs; few practice settings have the resources required.15 Glasgow and Emmons15 suggest that the extensive approaches for recruitment, assessment, and program management be replaced with lower cost strategies, such as mail, phone, or computer-based approaches. Because the HEP replication study had fewer resources compared with the clinical trial, with fewer one-on-one follow-up visits, strategies such as follow-up phone calls during the HEP replication would have required less time, and possibly lead to better functional outcomes. Health mentor program/volunteer availability and reasons that client discontinued participation needs to be further examined by program directors. In the Urban Institute survey of US nonprofit organizations, the most frequently cited challenges for replication of programs included recruiting volunteers and lack of funds to pay staff to train and supervise volunteers.34 Researchers in the Urban Institute study learned that designation of a staff member or a dedicated volunteer to manage the volunteer program resulted in greater net benefits to the organization. This is an important consideration for community programs that are inadequately funded; whether the dedicated person was paid or unpaid does not matter.
The finding of higher baseline self-efficacy association with functional outcomes in the HEP replication is supported by other studies, and adds onto the sparse literature on the ability of community-based programs to improve functional outcomes. Much of the application of self-efficacy theory for health promotion programs and health outcomes has been focused on arthritis for chronic disease self-management programs where baseline higher self-efficacy significantly reduced health care utilization at 1 year and increased self-management activities such as exercise and function.4,16 Other studies suggest that higher self-efficacy may reduce functional limitations and improve exercise performance, and that low self-efficacy is related to disability and functional decline.35,36 This finding has important implications for clients in the HEP replication, which was not as successful as the original trial in promoting functional improvement. High baseline self-efficacy may become more important for clients in the HEP dissemination because other structures and processes are less intense. Self-efficacy itself, however, is also affected by other factors such as income, literacy, and education.37–39 We did not have information about these client variables and future research on these variables for community-based health promotion programs is needed.
Our findings that lower baseline activity levels and that clients with arthritis were associated with lower functional outcomes differ from the Lorig studies,5 in which clients at a higher baseline activity level showed less improvement, and clients with arthritis showed more improvement. Lorig's programs differ from the HEP as clients are seen more intensely for 7 weeks whereas the HEP involved participation over a one year period.
Clients in smaller sites and those in urban locations did better with functional outcomes at 12 months versus those in larger sites and rural locations. Most studies on site size and locations have been done on hospitals and nursing homes. In support of our findings, improved mortality outcomes are associated with decreased hospital size in surgical patients.19 Possibly, older adults in smaller sites form strong relationships with providers and thus do better functionally in the HEP. Hospital studies also show lower quality of care for cardiac patients and higher 30-day post–cardiac death rates in rural hospital versus urban hospitals.20,21 Our findings add onto the literature on site size and site location importance in community-based organizations.
Study Limitations and Future Research
We used the 2002–2004 client data set and interviewed the providers in 2005; however, the HEP was essentially unchanged and providers followed the same protocol. Because we used secondary data at the client level, a lack of control existed on the quality and reliability of data collected. We used GEE to deal with the correlated error data issues. These models, however, tend to be less sensitive to the specification of the correlation structure. Although self-efficacy was significantly associated with improved functional outcomes, its effects appear to be weak. The probability for an improved exercise stage increased by 2% for each self-efficacy level increase. This impact could be meaningful for someone with a 10% increase in self-efficacy, which would then improve exercise stage by 22%. In the replication, there was a wide range in client self-efficacy (from 0 to 54): this could have a strong influence on exercise stage. Because the self-efficacy scale was self-reported, there may have been a tendency for clients to overreport on their ability to carry out tasks. Because there was missing data for the socioeconomic variables, we were unable to control for these variables in our analyses, which could have biased our final results. Both minority status and lower socioeconomic status may also play a role in functional outcomes. It is possible that clients were not comfortable with providing information on education and income. We cannot generalize this study to organizations other than community sites. Future research should include implementation research, the scientific study of methods to improve translation of evidence-based practice into real world settings, and exploring systemic strategies for using available resources to improve research translation. In addition, we need research on using self-efficacy enhancement techniques to improve functional outcomes in community-based programs.
CONCLUSIONS AND IMPLICATIONS FOR POLICY AND PRACTICE
Our findings highlight the challenges of dissemination of community evidence-based research, and the need to develop low-cost innovative strategies to achieve the same outcomes as for the clinical trial. Our findings also have implications for the design and implementation of health promotion and self-management programs such as the HEP. Baseline higher client self-efficacy seems to be an important factor to obtain similar functional outcomes as original trials, due to the lower intensity of resources in replications.
Our findings reinforce the need to get a dedicated trained volunteer to manage the health mentor program to minimize the challenges experienced by our respondents. In addition, program directors need to find solutions to ensuring program champions for the sites, adequate staffing, and PCP involvement. To improve translation of evidence-based services, researchers and community sites need to improve their collaboration and communication to help community providers and managers understand the structures and processes involved in program implementation and help them make realistic decisions about whether their site can sustain such a program. Closely related geriatric field clinicians such as physical therapists would benefit from the information presented in this study. Physical therapists could have an important role in promoting health self-management and increasing self-efficacy in their patients as patients return to their independent or supervised living conditions after participating in skilled rehabilitation interventions during home health, outpatients, skilled nursing facilities, or inpatient rehabilitation, These strategies may help obtain optimum outcomes for vulnerable populations such as elders with chronic disease at risk for disability. In addition, this may assist in the visionary plan focused on wellness and prevention put forth by the National Prevention Health Promotions and Public Health Council40 and supported by the American Physical Therapy Association.
The authors thank Susan Snyder, Vice President, Senior Services, Seattle, WA, for her guidance and help in administering this project, recruiting the participants, and providing information about the project; the University of Washington, Seattle, researchers: Elizabeth Phelan, MD, who provided project guidance, instrument validity, and administrative assistance, and Barbara Williams, PhD, who provided the database for the project and guidance in its usage; Terence Tivnan, PhD, Harvard University for his assistance with the statistical analysis; and the nurses, social workers, and site managers who participated in the study. The authors would also like to thank the other dissertation committee members, Suzanne Leveille, PhD, Walter Leutz, PhD, and Sarita Bhalotra, MD, for their participation.
1. Stuck AE, Walthert JM, Nikolaus T. Risk factors for functional status decline in community-living elderly people: a systematic literature review. Soc Sci Med. 1999;48:445–469.
2. Manton KG, XiLiang G, Lamb VL. Change in chronic disability from 1982 to 2004/2005 as measured by long-term changes in function and health in the U.S. elderly population. Proc Natl Acad Sci U S A. 2006;103:18374–18379.
3. Leveille SG, Wagner EH, Davis C, Davios C, Grothaus LC, Wallace JI. Preventing disability and managing chronic illness in frail older adults: a randomized trial of a community-based partnership with primary care. J Amn Geriatr Soc. 1998;46:1191–1198.
4. Lorig K, Ritter PL, Stewart AL, et al. Chronic disease self-management program: 2-year health status and health care utilization outcomes. Med Care. 2001;39:1217–1223.
5. Lorig K, Sobel DS, Stewart A, Brown WB, Bandura A, Ritter P. Evidence suggesting that a chronic disease self-management program can improve health status while reducing hospitalization. Med Care. 1999;37:5–14.
6. Lorig K, Sobel DS, Ritter PL, Laurent D, Hobbs M. Effect of a self-management program on patients with chronic disease. Eff Clin Pract. 2001;4:256–262.
7. Phelan EA, Williams B, Pennix BW, LoGerfo JP, Leveille SG. Activities of daily living function and disability in older adults in a randomized trial of the Health Enhancement Program. J Gerontol A Biol Sci Med Sci. 2004;59:838–843.
8. Phelan EA, Williams B, Leveille SG, Snyder S, Wagner EH, LoGerfo JP. Outcomes of a community-based dissemination of the health enhancement program. J Am Geriatr Soc. 2002;50:1519–1524.
9. Phelan EA, Williams B, Snyder S, Fitts SS, LoGerfo JP. A five state dissemination of a community-based disability prevention program for older adults. Clin Intervent Aging. 2006;1:267–274.
10. Fitts SS, Won CC, Williams B, et al. What is the optimal duration of participation in a community-based health promotion program for older adults. J Appl Gerontol. 2008;27:201–214.
11. Davis C, Leveille SG, Favaro S. Benefits to volunteers in a community-based health promotion and chronic illness self-management program for the elderly. J Gerontol Nurs. 1998;24:16–23.
12. Mendel P, Meredith LS, Schoenbaum M, Sherbourne CD, Wells KB. Interventions in organizational and community context: a framework for building evidence on dissemination and implementation in health services research. Adm Policy Mental Health. 2008;35:21–37.
13. Klein KJ, Knight AP. Innovation implementation: overcoming the challenge. Curr Dir Psychol Sci. 2005;14:243–246.
14. Grol R, Grimshaw J. From best evidence to best practice: effective implementation of change in patients' care. Lancet. 2003;362:1225–1230.
15. Glasgow RE, Emmons KM. How can we increase translation of research into practice? Types of Evidence needed. Ann Rev Public Health. 2007;28:413–433.
16. Marks R, Allegrante JP, Lorig K. A review and synthesis of research evidence for self-efficacy enhancing interventions for reducing chronic disability: implications for health education practice (Part II). Health Promot Pract. 2005;6:148–156.
17. Marks R, Allegrante JP, Lorig K. A review and synthesis of research evidence for self-efficacy enhancing interventions for reducing chronic disability: implications for health education practice (Part I). Health Promot Pract. 2005;6:37–43.
18. Krout J. Senior Centers and At-Risk Older Persons: A National Research Agenda. Washington, DC: The National Council on Aging, Inc, National Eldercare Institute on Multipurpose Senior Centers and Community Focal Points; 1993.
19. Carbonell AM, Lincourt AE, Kercher KW, et al. Do patient or hospital demographics predict cholecystectomy outcomes? A nationwide study of 93,578 patients. Surg Endosc. 2005;19:767–773.
20. Baldwin LM, MacLehose RF, Hart LG, Beaver SK. Quality of care for acute myocardial infarction in rural and urban US hospitals. J Rural Health. 2004;20:99–108.
21. Shiekh K, Bullock C. Urban-rural differences in the quality of care for Medicare patients with acute myocardial infarction. Arch Int Med. 2001;161:737–743.
22. Long BJ, Calfas KJ, Wooten W. A multisite field test of the acceptability of physical activity counseling in primary care: Project PACE. Am J Prevent Med. 1996;12:73–81.
23. Prochaska JO, DiClemente CC. The Transtheoretical Approach: Towards a Systematic Eclectic Framework. New York, NY: Brunner-Mazel; 1986:163–299.
24. Project PACE. Physician Manual. Atlanta, GA: Cardiovascular Health Branch, Centers for Disease Control. 1992.
25. McHorney C, Ware J, Raczek A. The MOS 36 item short-form health survey (SF-36): psychometric and clincal tests of validity in measuring physical and mental health constructs. Med Care. 1993;31:247–263.
26. Ware JE, Sherbourne CD. The MOS 36-item short-form health survey (SF-36). I. Conceptual framework and item selection. Med Care. 1992;30:473–483.
27. Bandura A. Self-efficacy: toward a unifying theory of behavioral change. Psychol Rev. 1977;84:191–215.
28. Lorig K, Stewart A, Ritter P, Gonzales V, Laurent D, Lynch J. Otucomes Measures for Health Education and Other Health Care Interventions. Thousand Oak, CA: Sage; 1996:24–45.
29. Yesavage JA, Brink TL, Rose TL. Development and validation of a geriatric depression screening scale: a preliminary report. J Psychiatr Res. 1982;17:37–49.
30. Ainsworth BE, Youmans CP. Tools for physical activity counseling in medical practice. Obesity Res. 2002;10:69S–75S.
31. Horton N, Lipstiz S. Review of software to fit Generalized Estimating Equation Regression Models. Am Statist. 1999;53:160–169.
32. Eccles MP, Johnston M, Hrisos S, et al. Translating clinicians' beliefs into implementation interventions (TRACII): a protocol for an intervention modeling experiment to change clinicians' interventions to implement evidence-based practice. Implementation Science. 2007. http://www.implementationscience.com/content/2/1/27
. Accessed April 4, 2011.
33. Kimberly J, Cook JM. Organizational measurement and the implementation of innovations in mental health services. Adm Policy Ment Health. 2008;35:11–20.
34. Hager MA, Brudney JL. Balancing Act: The Challengers and Benefits of Volunteers. Washington, DC: The Urban Institute; 2004.
35. Seeman TE, Unger JB, McAvay G, Mendes de Leon CF. Self-efficacy beliefs and perceived declines in functional ability: MacArthur studies of successful aging. J Gerontol B Psychol Sci Soc Sci. 1999;54: P214–P222.
36. McAuley E, Konopack JF, Morris KS, et al. Physical activity and functional limitations in older women: influence of self-efficacy. J Gerontol B Psychol Sci Soc Sci. 2006;61:P270–P277.
37. Ross C, Wu C. The links between education and health. Am Sociolog Rev. 1995;60:719–745.
38. Siahpush M, McNeill A, Borland R, Fong G. Socioeconomic variations in nicotine dependence, selfefficacy, and intention to quit across four countries: findings from the International Tobacco Control (ITC) Four Country Survey. Tobacco Control. 2006;15. www.tobaccocontrol.bmj.com
. Accessed August 3, 2011.
39. Francis L, Weiss B, Senf J, Heist K, Hargraves R. Does literacy education improve symptoms of depression and self-efficacy in individuals with low literacy and depressive symptoms? A preliminary investigation. J Am Board Fam Med. 2007;20:23–27.
40. National Prevention Council. The National Prevention Strategy: America's Plan for Better Health and Wellness. Washington, DC: US Department of Health and Human Services, Office of the Surgeon General; 2011. www.healthcare.gov/prevention/nphpphc/strategy/report.pdf
. Accessed December 1, 2011.
community programs; disability prevention; implementation challenges; functional outcomes; self-efficacy