Skip to main content
NCBI home page
The Gerontologist logoLink to The Gerontologist
. 2011 Jun;51(Suppl 1):S21–S32. doi: 10.1093/geront/gnq083

Special Diabetes Program for Indians: Retention in Cardiovascular Risk Reduction

Spero M Manson 1,*, Luohua Jiang 1, Lijing Zhang 1, Janette Beals 1, Kelly J Acton 2, Yvette Roubideaux 3; the SDPI Healthy Heart Demonstration Project4
PMCID: PMC3092971  PMID: 21565816

Abstract

Purpose: This study examined the associations between participant and site characteristics and retention in a multisite cardiovascular disease risk reduction project. Design and Methods: Data were derived from the Special Diabetes Program for Indians Healthy Heart Demonstration Project, an intervention to reduce cardiovascular risk among American Indians and Alaska Natives with diabetes. In 2006, a total of 1,072 participants from 30 participating sites completed baseline questionnaires measuring demographics and sociobehavioral factors. They also underwent a medical examination at baseline and were reassessed annually after baseline. A Provider Annual Questionnaire was administered to staff members of each grantee site at the end of each year to assess site characteristics. Generalized estimating equation models were used to evaluate the relationships between participant and site characteristics and retention 1 year after baseline. Results: Among enrolled participants, 792 (74%) completed their first annual assessment. Participants who completed the first annual assessment tended to be older and had, at baseline, higher body mass index and higher level of physical activity. Site characteristics associated with retention included average age of staff, proportion of female staff members, and percentage of staff members having completed graduate or professional school. Implications: Understanding successful retention must reach beyond individual characteristics of participants to include features of the settings that house the interventions.

Keywords: Diabetes, Cardiovascular risk prevention, Case management, Participant retention, American Indians

The recruitment of individuals into intervention trials has garnered increasing attention as it has become clear that a number of factors, such as gender, age, race/ethnicity, and disease status, systematically as well as differentially affect individual decisions to participate in these efforts (Blanton et al., 2006; D. M. Brown et al., 2006; Coday et al., 2005). There is mounting evidence that biases of this nature compromise representativeness, estimates of effects or rates, and generalizability of results, thereby undermining the confidence one can place in the preventative or treatment approaches under study (Lachin & Foulkes, 1986; Matts, Launer, Nelson, Miller, & Dain, 1998). In response, March 1994, the National Institutes of Health introduced a specific policy mandating the recruitment of racial/ethnic minorities and women into publicly funded clinical research (http://grants.nih.gov/grants/funding/women_min/guidelines_update.htm; Hohmann & Parron, 1996). Although this attention to recruitment was long overdue, much less consideration has been given to the consequences of the differential retention of participants once engaged in a study or trial.

Retention is just as important as recruitment to the conclusions one may draw from intervention trials. Like recruitment, retention also affects statistical power, bias, and generalizability (Robinson, Dennison, Wayman, Pronovost, & Needham, 2007). Low retention rates affect the power to detect differences in outcomes (Lachin & Foulkes, 1986). Differential retention of subgroups—defined, for example, in terms of gender, age, or race/ethnicity—compromises comparison, especially if subgroup membership is related to the outcome of interest (Good & Schuler, 1997). Generalizability of results is diminished to the extent that the characteristics of those who dropped out of an intervention study are different from those who were retained (Curry & Jackson, 2003).

A wide range of factors appears to affect whether or not an individual remains in a trial. There is, for example, ample documentation that men are more likely than women to be lost to follow-up (Bailey, Bieniasz, Kmak, Brenner, & Ruffin, 2004; D. R. Brown, Fouad, Basen-Engquist, & Tortolero-Luna, 2000; Hessol et al., 2001; Neill & Chessa, 1998; Shumaker, Dugan, & Bowen, 2000); that older adults—barring functional impairments and issues of access—are likely to participate more fully than younger adults (Anderson, Ory, Cohen, & McBride, 2000; Cassidy, Baird, & Sheikh, 2001; Thornquist, Patrick, & Omenn, 1991); that individuals of higher rather than lower socioeconomic status are more likely to be retained (Warren-Findlow, Prohaska, & Freedman, 2003); that African Americans as well as American Indians are at greater risk of dropout than Whites or Asian Americans (Blumenthal, Sung, Coates, Williams, & Liff, 1995; D. R. Brown & Topcu, 2003; Dilworth-Anderson & Williams, 2004; Gorelick, Harris, Burnett, & Bonecutter, 1998; Noe et al., 2007; Shavers, Lynch, & Burmeister, 2002); that depression, anxiety, and alcohol involvement shorten exposure (Chang, Brown, & Nitzke, 2009; Gappoo et al., 2009; Hessol et al., 2001; Katon et al., 2004; Villcorta et al., 2007; Williams, Powers, Yun, & Foa, 2010); and that being single with multiple children contributes to limited availability and infrequent attendance (Gappoo et al., 2009). With respect to staff/provider characteristics, it has been shown that women generally are more effective than men in retaining participants of both sexes (Tansey, Matte, Needham, & Herridge, 2007); that older staff—to the extent that they are seen as mature, experienced, and respected—are more successful than their younger counterparts in sustaining involvement (Blanton et al., 2006; Blumenthal et al., 1995); that staff training, expertise, and professional education correlate highly with decisions to remain in a trial (Booth, Corsi, & Mikulich-Gilbertson, 2004; D. M. Brown et al., 2006); and that interpersonal skills such as personalized attention, empathy, and support result in higher rates of completion (Cassidy et al., 2001). Various features of the site at which an intervention is mounted also are known to relate to adherence. Well-described examples include settings that are welcoming, non-stigmatizing, able to ensure privacy, easily accessible, and located near relevant resources (Gappoo et al., 2009; Warren-Findlow et al., 2003; Williams et al., 2008).

Turning to retention strategies, Robinson and colleagues (2007) meta-analysis of the relevant literature identified 12 basic themes, including community involvement, study identity, reminders, contact and scheduling methods, and financial incentives. No optimal strategy emerged; rather, successful retention was associated with a combination of multiple strategies. Their review found that programs with higher retention rates adopted an average of 21 different strategies cutting across several themes.

Yet, in reality, these observations are not as consistent as this brief review may suggest. The truer picture likely reflects a complicated interaction of participant, provider/staff, study/program, and site characteristics as well as retention strategies. For example, the personal salience of the health condition at issue (Noe et al., 2007), and ethnic match of participant and staff (Parra-Medina et al., 2004; Shavers et al., 2002), as well as participants’ previous research or intervention experiences (Marcellus, 2004; Mazzuca et al., 2004) have been strongly associated with the quality and duration of engagement in intervention trials.

We sought to advance this line of inquiry by examining the relative contribution of participant, staff/provider, and site characteristics to retention in a nationwide intervention aimed at reducing the risk of cardiovascular disease (CVD) in American Indians and Alaska Natives (AI/ANs) with diabetes. The Special Diabetes Program for Indians Healthy Heart (SDPI-HH) Demonstration Project employed an intensive case-management approach to promote the Indian Health Service's (IHS) existing standards of care, which are generally recognized as state of the art in our country's public health system. SDPI-HH was subjected to a rigorous, multilevel evaluation that chronicled participant outcomes over time and provided the data to address the questions before us.

Research Design and Methods

SDPI-HH Demonstration Project

In recognition of the large disparities in diabetes and its complications among AI/ANs (Burrows, Geiss, Engelgau, & Acton, 2000; Howard et al., 1999), in 2002, the U.S. Congress directed the IHS to develop a competitive grant program to support demonstration projects implementing evidence-based diabetes and CVD prevention activities and to evaluate the effectiveness of these activities among AI/AN communities. SDPI-HH is one of the two projects funded and is designed to reduce cardiovascular risk among AI/ANs with diabetes through an intensive case-management program.

In 2004, 30 health care programs received funding to participate in SDPI-HH. Grantees represent a diverse mix of programs, serving 138 tribes in 13 states and each of the 12 IHS administrative areas. These programs include 7 IHS hospitals/clinics and 23 tribal health care programs or urban IHS-contracted programs.

Participating programs implement a case-management intervention designed to improve CVD risk factors. At baseline and annually for up to 3 years, participants undergo a medical examination, during which CVD risk is assessed and an individualized care plan is developed or refined. After completing the baseline assessment, participants attend individual case-management meetings at which progress toward CVD risk reduction is evaluated. During these visits and through structured group activities, participants receive education on diabetes self-management and CVD risk reduction.

Participants

Participants were required to be AI/AN, to be 18 years or older, and to have diabetes. Individuals were excluded if they were pregnant, receiving dialysis for end-stage renal disease, undergoing cancer treatment, having active alcohol/substance abuse problems, or having any other condition that would affect successful participation based on a provider's judgment. When authorized by their local institutional review boards (IRBs) or the National IHS IRB (for sites with no local review board), participating grant programs identified individuals with diabetes in local communities mainly through electronic medical records or local diabetes registries. Some grant programs also recruited participants through community events (e.g., diabetes screening and health fairs) and clinic-based activities (e.g., advertisements in local clinics and provider referrals). Enrollment began in January 2006 and is ongoing. However, in consideration of fiscal and workload issues, an abbreviated participant questionnaire, which does not include all of the items needed for this work, was implemented in August 2009. Of those who were identified to be eligible, 1,440 participants were recruited and consented in the first year of enrollment. The present analyses involved the 1,072 participants who completed the baseline questionnaire and started case management visit in year 2006.

When a participant was not compliant with attending case-management visits or completing assessments, grant programs were instructed to continue retention efforts unless the participant requested otherwise. Retention forms were completed on each participant describing the nature and extent of retention efforts and, if the participant dropped out of the intervention, the reason for dropout. Retention forms were submitted to the Coordinating Center either when a participant formally dropped out of the program or annually in January for all participants.

The protocol was approved by the IRB of the University of Colorado Denver and by the National IHS IRB. When required, grantees obtained approval from other entities charged with overseeing research in their programs (e.g., tribal review boards, tribal council, and regional IHS IRB). All participants provided written informed consent and Health Insurance Portability and Accountability Act authorization. SDPI-HH grantees reviewed and approved this article in advance of publication.

Measures

At baseline and annually for up to 3 years, participants underwent a medical examination, during which CVD incidence and risk were assessed. Additionally, at each assessment, participants completed a questionnaire that included questions regarding sociodemographics, health-related knowledge, attitude, and behaviors, as well as comorbidities. In this study, completion of the first annual medical examination was deemed as successful first-year retention and was used as the dependent variable for all analyses. Those who did not complete the first-year medical examination due to death or becoming pregnant (n = 9) were excluded from the analyses. We examined the association between retention and the following site- and participant-level characteristics.

Site Characteristics.—

Site-specific factors included type of grant program administration (IHS hospital or clinic vs. tribal health care program), the user population of the health facility of each grantee site (small [less than 5,000 users], medium [5,000–9,999], and large [10,000+]), and the number of participants accrued at each site (<20, 20–39, and 40+). The characteristics of staff members of each grantee site were obtained from Provider Annual Questionnaires completed by grantee staff members in December 2006. In this study, we examined the relationship between retention and average age of the staff members (<42, 42–47, and 48+ years), percent of female staff members (<60%, 60–79%, and 80+%), and percent of staff members who completed graduate/professional school (<50%, 50–74%, and 75+%). The cutoff values of each staff characteristic were chosen so that the grantee sites were divided into three groups, where the grantee sites in the lowest quartile and the highest quartile were compared with those in the middle half.

Participant Characteristics.—

Sociodemographics. 

Participants answered a limited number of questions related to their sociodemographic characteristics, including age, gender, educational attainment, employment status, marital status, and annual household income.

Clinical indicators. 

Baseline physical examination included measurements of height, weight, and sitting systolic blood pressure (SBP) and diastolic blood pressure (DBP). Body mass index (BMI) was calculated from height and weight (kg/m2). Blood pressure was dichotomized into a binary variable: normal (SBP ≤ 130 mm Hg and DBP ≤ 80 mm Hg) and high blood pressure (SBP > 130 mm Hg or DBP > 80 mm Hg). Blood was drawn after a 12-hr fast to measure triglycerides, high-density lipoprotein (HDL) cholesterol, low-density lipoprotein cholesterol, and hemoglobin A1c (HbA1c). In addition, self-report number of comorbid conditions was assessed with the Self-Administered Comorbidity Questionnaire (Sangha, Stucki, Liang, Fossel, & Katz, 2003).

Sociobehavioral factors. 

Participants were queried about a wide range of sociobehavioral factors that may be related to reducing CVD risk. These included smoking status, stages of change, family support, psychological distress, physical activity, pain, and health literacy/numeracy.

  1. Smoking status: History and current status of cigarette smoking were collected using items from the American Indian Service Utilization, Psychiatric Epidemiology, Risk and Protective Factors Project (Nez Henderson, Jacobsen, Beals, & AI-SUPERPFP Team, 2005).

  2. Stages of change: The transtheoretical model posits that behavior change is more likely among those who have actively contemplated and prepared for such change (Marcus, Banspach, et al., 1992; Marcus, Selby, Niaura, & Rossi, 1992; Velicer, Fava, et al., 1995; Velicer, Hughes, Fava, Prochaska, & DiClemente, 1995) and may be associated with participants’ willingness to stay in the project. Five stages of change are proposed: (a) precontemplation (no intention of changing behavior), (b) contemplation (intend to change in distant future), (c) preparation (intend to change in the near future), (d) Action (actively engaged in new behavior), and (e) maintenance (have engaged in the behavior for some time, such as 6 months or more; Prochaska, Johnson, & Lee, 1998). As part of the participant questionnaire, stage of change was captured for both reducing dietary fat intake and engaging in regular exercise each week (Parker, Baker, Williams, & Nurss, 1995). In this study, we dichotomized both of these variables into two categories: pre-action (including precontemplation, contemplation, and preparation stages), and action (including action and maintenance stages).

  3. Family support: The availability of a family support person was determined by having a family member complete a brief family questionnaire at baseline.

  4. Distress: The Kessler Distress Scale (Furukawa, Kessler, Slade, & Andrews, 2003; Kessler et al., 2003) is a general measure of psychological distress often used to screen for serious emotional problems and previously has been shown to be related to retention (Chang et al., 2009; Katzer et al., 2008; Yass-Reed, Barry, & Dacey, 1993); SDPI-HH used the six-item version (K6; α = 0.88).

  5. Physical activity: Rapid Assessment of Physical Activity (Topolski et al., 2006), a short, self-administered instrument assessing weekly physical activity, was included in the baseline questionnaire. It is a nine-item instrument with response options of “yes” or “no” to questions covering a range of physical activity levels from “sedentary” to “regular active,” as well as strength training and flexibility. The total score of the first seven items measures aerobic physical activities and is categorized into one of five levels of physical activity: “1 = sedentary, 2 = underactive, 3 = regular underactive (light activities), 4 = regular underactive, and 5 = regular active.”

  6. Pain: A visual analog pain scale (range 1–10) was used to assess each participant's feeling of general pain at each visit (Carlsson, 1983; Jerome & Gross, 1991).

  7. Health literacy/numeracy: Print literacy items assessed self-reported confidence in reading or completing medical forms and were adapted from the health literacy screening questions developed by Chew, Bradley, and Boyko (2004). Overall performance on the print literacy items was measured as the mean of a participant's responses to the three items (possible range of 1–5, with larger values indicating greater print literacy). The numeracy items were adapted from two sources to assess comprehension of proportions and percentages, the ability to identify the appropriate timing of a second dose of medication, and to determine whether a given blood glucose value is in the normal range (Lipkus, Samsa, & Rimer, 2001; Parker et al., 1995). In this study, overall numerical performance was measured as the percentage of items answered correctly.

Data Analysis

Successful first-year retention was used as a dichotomous variable in all the data analyses. Bivariate associations between retention and site characteristics and categorical participant-level factors were examined using χ2 tests. Bivariate associations between retention and continuous participant-level variables were evaluated using two sample t-tests. Factors with p values of <.25 in the bivariate analyses were entered into multivariate models in a block-wise manner. More specifically, all the participant-level sociodemographic factors with p values <.25 in the bivariate analyses were entered into the multivariate model first, which was then reduced by deleting the variables with p values >.2 at that step. We then entered all participant clinical indicators with bivariate p values <.25 into the previous model, and reduced that model by excluding the variables with p values >.2 at that time. Similar steps were performed for the sociobehavioral factors and site-level characteristics until we reached a final model. To account for within-site clustering, generalized estimation equation (GEE) models with a logit link and an exchangeable correlation matrix were utilized for all multivariate regression models. All data analyses were conducted using SAS 9.2 software (SAS Institute, 2008).

Missing data were uncommon for most of the site- and participant-level factors (<10%) except income and marital status (20% and 18%, respectively). However, the p values for bivariate analyses between retention and these two variables were greater than .25 and, hence, were not included in the final multivariate model. Furthermore, the insignificant associations did not change when we performed the bivariate analyses using data with multiply imputed sociodemographic variables. The multiple imputations were performed using IVEware developed by the University of Michigan Survey Methodology Center (Raghunathan, Solenberger, & Hoewyk, 2009). Because the results based on data with and without imputation were similar, and the final model without imputation lost few observations (n = 49, only 4.6% of the total sample), we report the results from data analysis without imputation in the following section.

Results

As of December 2006, 1,072 participants had enrolled in SDPI-HH and completed baseline participant questionnaires. Seventy-four percent (792 of 1,072) of these participants completed their first annual medical examination. Participants who were not retained because of death (n = 7) or becoming pregnant (n = 2) were excluded from subsequent analyses. Many grant programs struggled with scheduling the first annual medical examinations, which were to be obtained 12–13 months after the participant started case management. On average, the first annual assessments were obtained 12.7 months (range, 6.4–21.8 months) after initiating case management. The mean number of case-management visits attended by the participants in the first year of SDPI-HH was 9 (range, 1–15) for the participants who completed the first annual assessment.

The total column of Table 1 describes the sample characteristics of SDPI-HH participants enrolled in year 2006. The majority of participants were women (69.4%) and were married or living in a “marriage-like” relationship (59.9%). The average age of the participants was 55.4 years old with a standard deviation of 11.4. About half of the participants had attended some years of college, and 16.2% of the participants had graduated from college. Twenty percent of SDPI-HH first-year participants were retired, whereas 25.9% were unemployed or still going to school.

Table 1.

Baseline Participant Characteristics by Retention Status

First annual assessment completed
Characteristics Total Yes No p Value
Sociodemographics Mean (SD) Mean (SD) Mean (SD)
    Age (years) 55.4 (11.4) 55.9 (11.2) 53.8 (11.7) .0113
    Gender N (%) N (%) N (%)
        Men 325 (30.6) 238 (30.1) 87 (32.1) .5266
        Women 738 (69.4) 554 (69.9) 184 (67.9)
    Education status
        <Higher secondary 191 (18.5) 143 (18.6) 48 (18.0) .3571
        Higher secondary graduate 290 (28.0) 215 (28.0) 75 (28.2)
        Some college 385 (37.2) 277 (36.1) 108 (40.6)
        ≥College graduate 168 (16.2) 133 (17.3) 35 (13.2)
    Employment status
        Employed 510 (53.7) 377 (53.8) 133 (53.6) .9427
        Retired 193 (20.3) 144 (20.5) 49 (19.8)
        Unemployed/student 246 (25.9) 180 (25.7) 66 (26.6)
    Marital status
        Married or live together 538 (59.9) 394 (59.3) 144 (61.5) .6761
        Separated, divorced, or widowed 254 (28.3) 188 (28.3) 66 (28.2)
        Never married 106 (11.8) 82 (12.3) 24 (10.3)
    Annual household income
        <15k 221 (25.9) 163 (25.7) 58 (26.7) .8671
        15 to <30k 243 (28.5) 186 (29.3) 57 (26.3)
        30 to <50k 228 (26.8) 168 (26.5) 60 (27.6)
        ≥50k 160 (18.8) 118 (18.6) 42 (19.4)
Clinical indicators Mean (SD) Mean (SD) Mean (SD)
    HbA1C (%) 7.6 (1.8) 7.5 (1.7) 7.8 (1.9) .0600
    BMI 36.6 (7.8) 36.9 (7.8) 35.8 (7.6) .0416
    LDL (mg/dl) 98.5 (31.8) 97.4 (32.6) 101.1 (29.5) .1263
    HDL (mg/dl) 43.6 (11.9) 44.2 (12.3) 42.1 (10.4) .0122
    Triglycerides (mg/dl) 194.2 (143.6) 194.4 (131.2) 193.3 (175.3) .9151
    Self-administered comorbidity index 4.4 (3.6) 4.5 (3.6) 4.2 (3.5) .1895
N (%) N (%) N (%)
    High blood pressure 499 (46.9) 358 (45.2) 141 (52.0) .0519
Sociobehavioral factors N (%) N (%) N (%)
    Current smoker 169 (17.0) 112 (15.2) 57 (22.3) .0090
    Stages of change for avoiding fatty food .4717
        Pre-action 367 (39.4) 269 (38.7) 98 (41.4)
        Action 565 (60.6) 426 (61.3) 139 (58.6)
    Stage of change for exercising regularly .2092
        Pre-action 554 (56.6) 400 (55.4) 154 (59.9)
        Action 425 (43.4) 322 (44.6) 103 (40.1)
    Presence of family support person 694 (65.3) 525 (66.3) 169 (62.4) .2412
Mean (SD) Mean (SD) Mean (SD)
    Kessler Distress Scale 1.81 (0.82) 1.82 (0.81) 1.78 (0.83) .4159
    Rapid Assessment of Physical Activity (aerobic) 3.92 (1.07) 3.96 (1.04) 3.83 (1.17) .1068
    Pain 3.82 (2.51) 3.85 (2.50) 3.75 (2.56) .6111
    Print literacy 3.42 (0.84) 3.42 (0.85) 3.40 (0.81) .7458
    Numeracy 0.70 (0.28) 0.70 (0.28) 0.70 (0.27) .9936
Open in a new tab

Note: HbA1c = hemoglobin A1c; BMI = body mass index; LDL = low-density lipoprotein; HDL = high-density lipoprotein.

Table 1 also compares baseline participant characteristics of those who completed their first annual assessment with those who did not complete their first annual assessment. Participants with a year-one assessment were significantly older (55.9 vs. 53.8 years, p = .0113) and had, at baseline, marginally lower HbA1c levels (7.5% vs. 7.8%, p = .0600), significantly higher BMI (36.9 vs. 35.8, p = .0416), and significantly higher HDL levels (44.2 vs. 42.1 mg/dl, p = .0122). They also were less likely to have high blood pressure and were less likely to be current smokers at baseline than those who did not finish their first annual medical examination.

Retention rates by different site-specific characteristics are shown in Table 2. Medium-sized sites in terms of user population (5,000–10,000 patients) had significantly higher retention rates than small or large-sized grantee sites (78.7% vs. 74.8% and 71.1%, respectively, p = .0418). Further, the sites that recruited 40 or more participants in year 2006 had significantly higher retention rates than the sites that recruited fewer participants in the same year (p = .0305). Regarding the characteristics of staff members in each grantee site, those sites with an average staff member age between 42 and 47 years exhibited significantly lower retention rates than did other sites (p = .0002). Sites with more than 80% female staff members had marginally lower retention rates than did sites with fewer female staff (p = .0697). Finally, the retention rates of sites at which more staff members had completed graduate or professional schools were significantly higher than those of sites with less-educated staff (p < .0001).

Table 2.

SDPI-HH Retention Rates by Site Characteristics

Characteristics No. of sites No. of participants % Completed Year 1 annual assessment p Value
Organization type
    IHS 7 243 77.4 .2441
    Tribal 23 820 73.7
User population size
    Small 8 218 74.8 .0418
    Medium 9 367 78.7
    Large 13 478 71.1
Total accrual number in 2006
    <20 7 104 70.2 .0305
    20–39 12 372 70.7
    40+ 11 587 77.7
Average age (years) of staff members
    <42 8 299 78.3 .0002
    42–47 15 532 69.2
    48+ 7 232 81.9
Proportion of female staff
    <60% 8 312 77.9 .0697
    60%–79% 13 419 75.4
    80%+ 9 332 70.2
Proportion of staff completing graduate/professional school
    <50% 6 206 57.3 <.0001
    50%–74% 17 611 76.9
    75%+ 7 246 82.9
Open in a new tab

Note: SDPI-HH = Special Diabetes Program for Indians Healthy Heart; IHS = Indian Health Service.

The final multivariate GEE model for year-one retention success incorporating both participant- and site-level variables is presented in Table 3. In the final model, baseline age was strongly associated with the odds of successful retention 1 year after baseline with an odds ratio of 1.26 for every 10 years of increase in age. In addition, participants with larger BMI and higher level of physical activity at baseline were significantly more likely to complete the first annual medical examination (OR = 1.02, p = .018; OR = 1.17, p = .0316, respectively). In terms of site characteristics, the average age of staff, proportion of female staff members, and proportion of staff completing graduate or professional school were marginally associated with retention. More specifically, sites with an average staff member age between 42 and 47 years were marginally less likely to have participants complete the first annual assessment than sites with an average staff member age of 48 years or greater (OR = 0.56, p = .067). Further, sites with 60–80% female staff were significantly more likely to retain participants one year after baseline than sites with 80% or more female staff (OR = 1.99, p = .045). Finally, sites where 75% or more staff members had completed graduate or professional school had significantly higher odds of retention success than did sites where less than 50% of staff members had finished graduate or professional school (OR = 3.12, p = .052).

Table 3.

Final Generalized Estimation Equation Model for First-Year Retention Success

Characteristics Odds ratio 95% confidence limits p Value
Participant level
    Age (10 years) 1.26 1.10 1.43 .0006
    BMI 1.02 1.00 1.05 .0180
    Rapid Assessment of Physical Activity (aerobic) 1.17 1.01 1.35 .0316
    High blood pressure 0.86 0.70 1.06 .1572
Site level
    Average age (years) of staff members
        <42 1.18 0.46 3.03 .7285
        42 to <47 0.56 0.30 1.04 .0671
        48+ (reference) 1.00
    Proportion of female staff
        <60% 1.06 0.47 2.41 .8802
        60%–79% 1.99 1.02 3.91 .0446
        80%+ (reference) 1.00
    Proportion of staff completing graduate/professional school
        <50% (reference) 1.00
        50%–74% 2.19 0.87 5.50 .0944
        75%+ 3.12 0.99 9.81 .0520
Open in a new tab

Note: BMI = body mass index.

Discussion

The most startling finding of this study is the minimal role that participant characteristics played in predicting retention. This is especially true with respect to the results of the bivariate analyses, which tend to overestimate the importance of individual factors independent of the possibly simultaneous contribution of others. Among sociodemographic features, only participant age, specifically being older, was associated with remaining in the program. Age likely operated here as a marker of perceived risk and relevance of participation to longer-term well-being (Anderson et al., 2000; Arean, Alvidrez, Nery, Estes, & Linkins, 2003). However, many other sociodemographic factors widely assumed to influence retention simply did not: gender, educational level, marital status, and income were prominent by virtue of their absent effect. As indices of tangible resources that can facilitate an individual's investment of time and effort in interventions of this nature, their lack of importance here is particularly striking.

Nor did commonly noteworthy sociobehavioral factors predict retention in SDPI-HH. Participant status in regard to readiness to change diet or extent of exercise, health literacy and numeracy, psychological distress, and pain, as well as availability of family support, bore no relationship to whether or not participants continued in the intervention at the close of the year. These factors relate to individual motivation to change, to assimilating knowledge specific to desired health outcomes, to cognitively reframing and acquiring skills to change related behavior, to coping with the physical demands of such change, and to encouraging as well as routinizing subsequent gains (Conwell et al., 2007; Davis, Broome, & Cox, 2002; De Bruyn, Hudgens, Sullivan, & Duerr, 2005). All have surfaced to varying degrees in the literature as being linked to retention, but not here. At the bivariate level, only being a current smoker upon entry into the program proved to be strongly associated with later leaving it—at first glance, a logical observation for an intervention focused on cardiovascular risk reduction and the challenge of smoking cessation.

Of the relevant clinical indicators, higher BMI at baseline was related to successful retention in both bivariate and multivariate analyses. The need to improve—and the desirability of improving—one's status with respect to these indicators clearly is consistent with the goals of this intervention. Hence, its emergence in distinguishing between participants retained and participants lost to follow-up was not surprising. In the same vein, however, it was unexpected that other comorbid physical health problems, several of which are often the subject of CVD prevention, did not correlate with retention (Dudley et al., 1995; Froelicher et al., 2003).

In sharp contrast, the bivariate analyses repeatedly underscored the contribution of site-specific characteristics to retention. Sites successfully retaining their participants served medium-sized user populations; accrued large numbers of participants; and were characterized by a mix of younger and older staff, a balance of male and female providers, and higher levels of professional preparation. These attributes speak to issues of access, to optimal availability of resources (e.g., time, equipment, expertise, and services), to program presence, to varied experience, and to creditability as well as to authority (Good & Schuler, 1997; Morse, Simon, Besch, & Walker, 1995; Newberry et al., 2010; Orr, Blackhurst, & Hawkins, 1992; Williams et al., 2008).

Subsequent multivariate modeling sharpened our initial findings. With respect to participant characteristics, increasing age and higher BMI remained strong predictors of retention, unexpectedly joined by higher levels of physical activity at baseline, which previously had not been suggested. Smoking status, HDL level, and high blood pressure faded in importance. The contribution of several site characteristics was attenuated, notably user population size and number of intervention participants. Staff age, gender, and professional training remained important to retention, although not as robustly as before.

Several limitations to this study may have constrained our ability to identify other elements critical to retention. First, we did not directly measure transportation as a possible barrier. The intensive case-management activities emanated from largely community-based clinics, the long-standing seat of most medical care offered in AI/AN communities. Participants can live at substantial distance from such sites, thereby compromising their ability to keep scheduled appointments. However, lack of transportation is not likely to explain dropout more than a year-long intervention. Other measures of instrumental resources—for example, income, marital status, and family support—were unrelated to continued participation. In addition, staff went to great lengths to schedule alternate appointments to accommodate the exigencies of life in these communities. For example, telephone follow-up and even home visits frequently were used to maintain contact and to retain patient involvement. This latter point introduces the second limitation of the present study. Namely, we did not consider here the specific retention strategies that program staff used to engage participants. As noted previously, the evaluation systematically gathered this information, but we have not yet had the opportunity to assemble, categorize, and quantify such activities. Clearly, outreach and engagement efforts of the type just mentioned can affect retention. Grantees proved remarkably creative in developing a wide range of incentives as well as events to sustain interest in the program. Indeed, it may be that these retention strategies are linked to the site characteristics that predicted retention. Moreover, the power of such strategies may explain why the individual characteristics of participants contributed so little to identifying the forces at work in sustaining their involvement. We intend to examine these possibilities in future reports. Finally, SDPI-HH is comprised of 30 grantee sites, which may have limited statistical power in identifying significant relationships between site characteristics and retention. This may partially explain why several significant effects of site characteristics to retention in bivariate analysis disappeared in the multivariate GEE model when we took the multilevel data nature of SDPI-HH into consideration.

These findings suggest that we may have a tendency to overemphasize the contribution of participant attributes and dispositions to retention in preventative and treatment trials. Although they are undoubtedly important, recognizing the important role of other factors, notably environmental ones, promises to lead us away from “blaming the victim,” as is too frequently the case, especially when examining such matters among women, older adults, and racial/ethnic minorities. In addition to recruitment activities, our future work in this regard will consider how staff turnover, organizational culture, agency status (e.g., private, tribal, or federal), and rurality may contribute to successfully retaining participants in this large-scale intervention. The present findings argue that, despite individual predilections to the contrary (even if determined to be true), we can alter process and setting to promote engagement through supports and rewards. This is good news. We need to more aggressively pursue this line of inquiry and, as importantly, discover how to translate the resulting knowledge into everyday practice.

Funding

Support for this manuscript was provided by the Indian Health Service (HHSI242200400049C) and the National Institute on Aging, Resource Centers for Minority Minority Aging Research (P30AG15297) to Spero M. Manson.

Acknowledgments

We wish to express our gratitude to the IHS as well as tribal and urban Indian health programs and participants involved in the SDPI-HH Demonstration Project.

Appendix.

Absentee Shawnee Tribe of Oklahoma; Albuquerque Service Unit; Bad River Band of Lake Superior Chippewa; Blackfeet Tribe; Choctaw Nation of Oklahoma; Confederated Salish and Kootenai Tribes; Montana/Wyoming Tribal Consortium with Assiniboine and Gros-Ventre, Chippewa Cree Tribe, Crow Nation; Hualapai Tribe; Indian Health Care Resource Center of Tulsa, Inc., in consortium with Northeastern Tribal Health System Miami Service Unit; Indian Health Council, Inc.; Leech Lake Reservation Tribal Council; Mille Lacs Band of Ojibwe in consortium with St. Croix Chippewa Indians of Wisconsin; Muscogee Creek Nation Health System; Navajo Area Indian Health Service with Northern Navajo Medical Center and Inscription House Clinic; Northwest Washington Indian Health Board with Lummi Indian Nation, Nooksack Tribe of Indians, Swinomish Tribal Community, and Upper Skagit Indian Tribe; Ramah Navajo School Board, Inc.; Redding Rancheria Indian Health Clinic; Hoopa Valley Tribe; Riverside-San Bernardino County Indian Health, Inc.; Santo Domingo Tribe; Sault Ste Marie Tribe Chippewa; Seattle Indian Health Board; St. Regis Mohawk Health Services; Taos-Picuris Service Unit; Tohono O’Odham Healthy Heart Demonstration Project; Toiyabe Indian Health Project, Inc.; Uintah and Ouray IHS Clinic; Wagner Health Care Center IHS; Whiteriver IHS Service Unit; Yakama Indian Health Center-IHS/DHHS; Yukon-Kuskokwim Health Corporation.

References

  1. Anderson RT, Ory M, Cohen S, McBride JS. Issues of aging and adherence to health interventions. Controlled Clinical Trials. 2000;21:171S–183S. doi: 10.1016/s0197-2456(00)00075-1. doi:10.1016/S0197-2456(00)00075-1. [DOI] [PubMed] [Google Scholar]
  2. Arean PA, Alvidrez J, Nery R, Estes C, Linkins K. Recruitment and retention of older minorities in mental health services research. The Gerontologist. 2003;43:36–44. doi: 10.1093/geront/43.1.36. doi:10.1093/geront/43.1.36. [DOI] [PubMed] [Google Scholar]
  3. Bailey JM, Bieniasz ME, Kmak D, Brenner DE, Ruffin MT. Recruitment and retention of economically underserved women to a cervical cancer prevention trial. Applied Nursing Research. 2004;17:55–60. doi: 10.1016/j.apnr.2003.12.002. doi:10.1016/j.apnr.2003.12.002. [DOI] [PubMed] [Google Scholar]
  4. Blanton S, Morris DM, Prettyman MG, McCulloch K, Redmond S, Light KE, et al. Lessons learned in participant recruitment and retention: The EXCITE Trial. Physical Therapy. 2006;86:1520–1533. doi: 10.2522/ptj.20060091. doi:10.2522/ptj.20060091. [DOI] [PubMed] [Google Scholar]
  5. Blumenthal D, Sung J, Coates R, Williams J, Liff J. Recruitment and retention of subjects for a longitudinal cancer prevention study in an inner-city black community. Health Services Research. 1995;30:197–205. [PMC free article] [PubMed] [Google Scholar]
  6. Booth RE, Corsi KF, Mikulich-Gilbertson SK. Factors associated with methadone maintenance treatment retention among street-recruited injection drug users. Drug and Alcohol Dependence. 2004;74:177–185. doi: 10.1016/j.drugalcdep.2003.12.009. doi:10.1016/j.drugalcdep.2003.12.009. [DOI] [PubMed] [Google Scholar]
  7. Brown DM, Thorne JE, Foster GL, Duncan JL, Brune LM, Muñana A, et al. Factors affecting attrition in a longitudinal study of patients with AIDS. AIDS Care. 2006;18:821–829. doi: 10.1080/09540120500466747. doi:10.1080/09540120500466747. [DOI] [PubMed] [Google Scholar]
  8. Brown DR, Fouad MN, Basen-Engquist K, Tortolero-Luna G. Recruitment and retention of minority women in cancer screening, prevention, and treatment trials. Annals of Epidemiology. 2000;18:S13–S21. doi: 10.1016/s1047-2797(00)00197-6. doi:10.1016/S1047-2797(00)00197-6. [DOI] [PubMed] [Google Scholar]
  9. Brown DR, Topcu M. Willingness to participate in clinical treatment research among older African Americans and whites. The Gerontologist. 2003;43:62–72. doi: 10.1093/geront/43.1.62. doi:10.1093/geront/43.1.62. [DOI] [PubMed] [Google Scholar]
  10. Burrows NR, Geiss LS, Engelgau MM, Acton KJ. Prevalence of diabetes among Native Americans and Alaska Natives, 1990-1997: An increasing burden. Diabetes Care. 2000;23:1786–1790. doi: 10.2337/diacare.23.12.1786. [DOI] [PubMed] [Google Scholar]
  11. Carlsson AM. Assessment of chronic pain. I. Aspects of the reliability and validity of the visual analogue scale. Pain. 1983;16:87–101. doi: 10.1016/0304-3959(83)90088-X. [DOI] [PubMed] [Google Scholar]
  12. Cassidy EL, Baird E, Sheikh JI. Recruitment and retention of elderly patients in clinical trials: Issues and strategies. American Journal of Geriatric Psychiatry. 2001;9:136–140. [PubMed] [Google Scholar]
  13. Chang MW, Brown R, Nitzke S. Participant recruitment and retention in a pilot program to prevent weight gain in low-income overweight and obese mothers. BMC Public Health. 2009;9:424. doi: 10.1186/1471-2458-9-424. doi:10.1186/1471-2458-9-424. [DOI] [PMC free article] [PubMed] [Google Scholar]
  14. Chew LD, Bradley KA, Boyko EJ. Brief questions to identify patients with inadequate health literacy. Family Medicine. 2004;36:588–594. [PubMed] [Google Scholar]
  15. Coday M, Boutin-Foster C, Goldman Sher T, Tennant J, Greaney ML, Saunders SD, et al. Strategies for retaining study participants in behavioral intervention trials: Retention experiences of the NIH behavior change consortium. Annals of Behavioral Medicine. 2005;29:55–65. doi: 10.1207/s15324796abm2902s_9. doi:10.1207/s15324796abm2902s_9. [DOI] [PubMed] [Google Scholar]
  16. Conwell DS, Mosher A, Khan A, Tapy J, Sandman L, Vernon A, et al. Factors associated with loss to follow-up in a large tuberculosis treatment trial (TBTC Study 22) Contemporary Clinical Trials. 2007;28:288–294. doi: 10.1016/j.cct.2006.09.003. doi:10.1016/j.cct.2006.09.003. [DOI] [PubMed] [Google Scholar]
  17. Curry L, Jackson J. Recruitment and retention of diverse ethnic and racial groups in health research: An evolving science. In: Curry L, Jackson J, editors. The science of inclusion: Recruiting and retaining racial and ethnic elders in health research. Washington, DC: Gerontological Society of America; 2003. pp. 1–7. [Google Scholar]
  18. Davis LL, Broome ME, Cox RP. Maximizing retention in community-based clinical trials. Journal of Nursing Scholarship. 2002;34:47–53. doi: 10.1111/j.1547-5069.2002.00047.x. doi:10.1111/j.1547-5069.2002.00047.x. [DOI] [PubMed] [Google Scholar]
  19. De Bruyn G, Hudgens MG, Sullivan PS, Duerr AC. Participant retention in clinical trials of HIV vaccines. Journal of Acquired Immune Deficiency Syndrome. 2005;39:499–501. doi: 10.1097/01.qai.0000148532.12329.df. [DOI] [PubMed] [Google Scholar]
  20. Dilworth-Anderson P, Williams SW. Recruitment and retention strategies for longitudinal African American caregiving research: The Family Caregiving Project. Journal on Aging and Health. 2004;16:137S–156S. doi: 10.1177/0898264304269725. doi:10.1177/0898264304269725. [DOI] [PubMed] [Google Scholar]
  21. Dudley J, Jin S, Hoover D, Metz S, Thackeray R, Chmiel J. The Multicenter AIDS Cohort Study: Retention after 9 1/2 years. American Journal of Epidemilogy. 1995;142:323–330. doi: 10.1093/oxfordjournals.aje.a117638. [DOI] [PubMed] [Google Scholar]
  22. Froelicher ES, Miller NH, Buzaitis A, Pfenninger P, Misuraco A, Jordan S, et al. The Enhancing Recovery in Coronary Heart Disease Trail (ENRICHD): Strategies and techniques for enhancing retention of patients with acute myocardial infarction and depression or social isolation. Journal of Cardiopulmonary Rehabilitation. 2003;23:269–280. doi: 10.1097/00008483-200307000-00004. [DOI] [PubMed] [Google Scholar]
  23. Furukawa TA, Kessler RC, Slade T, Andrews G. The performance of the K6 and K10 screening scales for psychological distress in the Australian National Survey of Mental Health and Well-Being. Psychological Medicine. 2003;33:357–362. doi: 10.1017/s0033291702006700. doi:10.1017/S0033291702006700. [DOI] [PubMed] [Google Scholar]
  24. Gappoo S, Montogomery E, Gerdts C, Naidoo S, Chidanyika A, Nkala B, et al. Novel strategies implemented to ensure high participant retention rates in a community based HIV prevention effectiveness trial in South Africa and Zimbabwe. Contemporary Clinical Trials. 2009;30:411–418. doi: 10.1016/j.cct.2009.05.002. doi:10.1016/j.cct.2009.05.002. [DOI] [PubMed] [Google Scholar]
  25. Good M, Schuler L. Subject retention in a controlled clinical trial. Journal of Advanced Nursing. 1997;26:351–355. doi: 10.1046/j.1365-2648.1997.1997026351.x. doi:10.1046/j.1365-2648.1997.1997026351.x. [DOI] [PubMed] [Google Scholar]
  26. Gorelick PB, Harris Y, Burnett B, Bonecutter FJ. The recruitment triangle: Reasons why African Americans enroll, refuse to enroll, or voluntarily withdraw from a clinical trial. Journal of the National Medical Association. 1998;90:141–145. [PMC free article] [PubMed] [Google Scholar]
  27. Hessol NA, Schneider M, Greenblatt RM, Bacon M, Barranday Y, Holman S, et al. Retention of women enrolled in a prospective study of human immunodeficiency virus infection: Impact of race, unstable housing, and use of human immunodeficiency virus therapy. American Journal of Epidemiology. 2001;154:563–573. doi: 10.1093/aje/154.6.563. doi:10.1093/aje/154.6.563. [DOI] [PubMed] [Google Scholar]
  28. Hohmann AA, Parron D. How the new guidelines on inclusion of women and minorities apply: Efficacy trials, effectiveness trials, and validity. Journal of Consulting and Clinical Psychology. 1996;64:851–855. doi: 10.1037//0022-006x.64.5.851. doi:10.1037/0022-006X.64.5.851. [DOI] [PubMed] [Google Scholar]
  29. Howard BV, Lee ET, Cowan LD, Devereux RB, Galloway JM, Go OT, et al. Rising tide of cardiovascular disease in American Indians. The Strong Heart Study. Circulation. 1999;99:2389–2395. doi: 10.1161/01.cir.99.18.2389. [DOI] [PubMed] [Google Scholar]
  30. Jerome A, Gross RT. Pain disability index: Construct and discriminant validity. Archives of Physical Medicine Rehabilitation. 1991;72:920–922. doi: 10.1016/0003-9993(91)90012-8. [DOI] [PubMed] [Google Scholar]
  31. Katon WJ, Simon G, Russo J, Von Korff M, Lin EH, Ludman E, et al. Quality of depression care in a population-based sample of patients with diabetes and major depression. Medical Care. 2004;42:1222–1229. doi: 10.1097/00005650-200412000-00009. doi:10.1097/00005650-200412000-00009. [DOI] [PubMed] [Google Scholar]
  32. Katzer L, Bradshaw AJ, Horwath CC, Gray AR, O’Brien S, Joyce J. Evaluation of a “nondieting” stress reduction program for overweight women: A randomized trial. American Journal of Health Promotion. 2008;22:264–274. doi: 10.4278/060728113R1.1. doi:10.4278/060728113R1.1. [DOI] [PubMed] [Google Scholar]
  33. Kessler RC, Barker PR, Colpe LJ, Epstein JF, Gfroerer JC, Hiripi E, et al. Screening for serious mental illness in the general population. Archives of General Psychiatry. 2003;60:184–189. doi: 10.1001/archpsyc.60.2.184. doi:10.1001/archpsyc.60.2.184. [DOI] [PubMed] [Google Scholar]
  34. Lachin JM, Foulkes MA. Evaluation of sample size and power for analyses of survival with allowance for nonuniform patient entry, losses to follow-up and non-compliance. Biometrics. 1986;42:507–519. doi:10.2307/2531201. [PubMed] [Google Scholar]
  35. Lipkus IM, Samsa G, Rimer BK. General performance on a numeracy scale among highly educated samples. Medical Decision Making. 2001;21:37–44. doi: 10.1177/0272989X0102100105. doi:10.1177/0272989X0102100105. [DOI] [PubMed] [Google Scholar]
  36. Marcellus L. Are we missing anything? Pursuing research on attrition. Canadian Journal of Nursing Research. 2004;36:82–98. [PubMed] [Google Scholar]
  37. Marcus BH, Banspach SW, Lefebvre RC, Rossi JS, Carleton RA, Abrams DB. Using the stages of change model to increase the adoption of physical activity among community participants. American Journal of Health Promotion. 1992;6:424–429. doi: 10.4278/0890-1171-6.6.424. [DOI] [PubMed] [Google Scholar]
  38. Marcus BH, Selby VC, Niaura RS, Rossi JS. Self-efficacy and the stages of exercise behavior change. Research Quarterly for Exercise and Sport. 1992;63:60–66. doi: 10.1080/02701367.1992.10607557. [DOI] [PubMed] [Google Scholar]
  39. Matts JP, Launer CA, Nelson ET, Miller C, Dain B. A graphical assessment of the potential impact of losses to follow-up on the validity of study results. Statistical Medicine. 1998;16:1943–1954. doi: 10.1002/(sici)1097-0258(19970915)16:17<1943::aid-sim631>3.0.co;2-r. doi:10.1002/(SICI)1097-0258(19970915)16:17<1943::AID-SIM631>3.0.CO;2-R. [DOI] [PubMed] [Google Scholar]
  40. Mazzuca SA, Brandt KD, Katz BP, Lane KA, Bradley JD, Heck LW, et al. Subject retention and adherence in a randomized placebo-controlled trial of a disease-modifying osteoarthritis drug. Arthritis and Rheumatology. 2004;51:933–940. doi: 10.1002/art.20831. doi:10.1002/art.20831. [DOI] [PubMed] [Google Scholar]
  41. Morse EV, Simon PM, Besch CL, Walker J. Issues of recruitment, retention, and compliance in community-based clinical trials with traditionally underserved populations. Applied Nursing Research. 1995;8:8–14. doi: 10.1016/s0897-1897(95)80240-1. doi:10.1016/S0897-1897(95)80240-1. [DOI] [PubMed] [Google Scholar]
  42. Neill KM, Chessa F. Recruitment and retention of women in nontherapeutic clinical trials. Applied Nursing Research. 1998;3:148–151. doi: 10.1016/s0897-1897(98)80152-3. doi:10.1016/S0897-1897(98)80152-3. [DOI] [PubMed] [Google Scholar]
  43. Newberry A, Sherwood P, Hricik A, Kuo J, Crago E, Hoffman LA, et al. Understanding recruitment and retention in neurological research. Journal of Neuroscience. 2010;42:47–57. doi: 10.1097/jnn.0b013e3181c1fdd9. [DOI] [PMC free article] [PubMed] [Google Scholar]
  44. Nez Henderson P, Jacobsen C, Beals J AI-SUPERPFP Team. Correlates of cigarette smoking among selected Southwest and Northern plains tribal groups: The AI-SUPERPFP study. American Journal of Public Health. 2005;95:867–872. doi: 10.2105/AJPH.2004.050096. doi:10.2105/AJPH.2004.050096. [DOI] [PMC free article] [PubMed] [Google Scholar]
  45. Noe TD, Manson SM, Croy C, McGough H, Henderson J, Buchwald DS. The influence of community-based participatory research principles on the likelihood of participation in health research in American Indian communities. Ethnicity & Disease. 2007;17(Suppl. 1):S6–S14. [PubMed] [Google Scholar]
  46. Orr PR, Blackhurst DW, Hawkins BS. Patient and clinic factors predictive of missed visits and inactive status in a multi-center clinical trial. Controlled Clinical Trials. 1992;13:40–49. doi: 10.1016/0197-2456(92)90028-x. doi:10.1016/0197-2456(92)90028-X. [DOI] [PubMed] [Google Scholar]
  47. Parker RM, Baker DW, Williams MV, Nurss JR. The test of functional health literacy in adults: A new instrument for measuring patients’ literacy skills. Journal of General Internal Medicine. 1995;10:537–541. doi: 10.1007/BF02640361. [DOI] [PubMed] [Google Scholar]
  48. Parra-Medina D, D’antonio A, Smith SM, Levin S, Kirkner G, Mayer-Davis E. Successful recruitment and retention strategies for a randomized weight management trial for people with diabetes living in rural, medically underserved counties of South Carolina: The POWER study. Journal of the American Dietetic Association. 2004;104:70–75. doi: 10.1016/j.jada.2003.10.014. doi:10.1016/j.jada.2003.10.014. [DOI] [PubMed] [Google Scholar]
  49. Prochaska JO, Johnson S, Lee P. The transtheoretical model of behavior change. In: Shumaker SA, Schron EB, Ockene JK, McBee WL, editors. The handbook of health behavior change, second edition. New York: Springer Publishing Company Book section; 1998. pp. 59–84. [Google Scholar]
  50. Raghunathan TE, Solenberger PW, Hoewyk JV. IVEware: Imputation and variance estimation software. Ann Arbor, MI: University of Michigan Survey Research Center Software; 2009. [Google Scholar]
  51. Robinson KA, Dennison CR, Wayman DW, Pronovost PJ, Needham DM. Systematic review identifies number of strategies important for retaining study participants. Journal of Clinical Epidemiology. 2007;60:757–765. doi: 10.1016/j.jclinepi.2006.11.023. doi:10.1016/j.clinepi.2006.11.023. [DOI] [PMC free article] [PubMed] [Google Scholar]
  52. Sangha O, Stucki G, Liang MH, Fossel AH, Katz JN. The Self-Administered Comorbidity Questionnaire: A new method to assess comorbidity for clinical and health services research. Arthritis Rheum. 2003;49:156–163. doi: 10.1002/art.10993. doi:10.1002/art.10993. [DOI] [PubMed] [Google Scholar]
  53. SAS Institute. SAS (version 9.2) Cary, NC: SAS Institute Software; 2008. [Google Scholar]
  54. Shavers V, Lynch C, Burmeister L. Racial differences in factors that influence the willingness to participate in medical research studies. Annals of Epidemiology. 2002;12:249–256. doi: 10.1016/s1047-2797(01)00265-4. doi:10.1016/S1047-2797(01)00265-4. [DOI] [PubMed] [Google Scholar]
  55. Shumaker SA, Dugan E, Bowen DJ. Enhancing adherence in randomized controlled trials. Controlled Clinical Trials. 2000;21:226S–232S. doi: 10.1016/s0197-2456(00)00083-0. doi:10.1016/S0197-2456(00)00083-0. [DOI] [PubMed] [Google Scholar]
  56. Tansey C, Matte A, Needham DM, Herridge M. Review of retention strategies in longitudinal studies and application to follow-up of ICU survivors. Intensive Care Medicine. 2007;33:2051–2057. doi: 10.1007/s00134-007-0817-6. doi:10.1007/s00134-007-0817-6. [DOI] [PubMed] [Google Scholar]
  57. Thornquist M, Patrick D, Omenn G. Participation and adherence among older men and women recruited to the Beta-Carotene and Retinol Efficacy trial (CARET) The Gerontologist. 1991;31:593–597. doi: 10.1093/geront/31.5.593. [DOI] [PubMed] [Google Scholar]
  58. Topolski TD, LoGerfo J, Patrick DL, Williams B, Walwick J, Patrick MB. The Rapid Assessment of Physical Activity (RAPA) among older adults. Preventing Chronic Disease. 2006;3:A118. [PMC free article] [PubMed] [Google Scholar]
  59. Velicer WF, Fava JL, Prochaska JO, Abrams DB, Emmons KM, Pierce JP. Distribution of smokers by stage in three representative samples. Preventive Medicine. 1995;24:401–411. doi: 10.1006/pmed.1995.1065. doi:10.1006/pmed.1995.1065. [DOI] [PubMed] [Google Scholar]
  60. Velicer WF, Hughes SL, Fava JL, Prochaska JO, DiClemente CC. An empirical typology of subjects within stage of change. Addictive Behavior. 1995;20:299–320. doi: 10.1016/0306-4603(94)00069-b. doi:10.1016/0306-4603(94)00069-B. [DOI] [PubMed] [Google Scholar]
  61. Villcorta V, Kegeles S, Galea J, Konda K, Cuba J, Palacios C, et al. Innovative approaches to cohort retention in a community-based HIV/STI prevention trial for socially marginalized Peruvian young adults. Clinical Trials. 2007;4:32–41. doi: 10.1177/1740774506075869. doi:10.1177/1740774506075869. [DOI] [PMC free article] [PubMed] [Google Scholar]
  62. Warren-Findlow J, Prohaska T, Freedman D. Challenges and opportunities in recruiting and retaining underrepresented populations into health promotion research. The Gerontologist. 2003;43:37–46. doi: 10.1093/geront/43.suppl_1.37. doi:10.1093/geront/43.suppl_1.37. [DOI] [PMC free article] [PubMed] [Google Scholar]
  63. Williams M, Powers M, Yun YD, Foa E. Minority participation in randomized controlled trials for obsessive-compulsive disorder. Journal of Anxiety Disorders. 2010;24:171–177. doi: 10.1016/j.janxdis.2009.11.004. doi:10.1093/aje/kwn072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  64. Williams P, Van Dyke R, Eagle M, Smith D, Vincent C, Ciupak G, et al. Association of site-specific and participant-specific factors with retention of children in a long-term pediatric cohort study. American Journal of Epidemiology. 2008;167:1375–1386. doi: 10.1093/aje/kwn072. doi:10.1093/aje/kwn072. [DOI] [PMC free article] [PubMed] [Google Scholar]
  65. Yass-Reed EM, Barry NJ, Dacey CM. Examination of pretreatment predictors of attrition in a VLCD and behavior therapy weight-loss program. Addictive Behaviors. 1993;18:431–435. doi: 10.1016/0306-4603(93)90060-m. doi:10.1016/0306-4603(93)90060-M. [DOI] [PubMed] [Google Scholar]

Articles from The Gerontologist are provided here courtesy of Oxford University Press

RESOURCES