Attrition and related trends in scientific rigor: a score card for ART adherence intervention research and recommendations for future directions

Five-year attrition, active enrollment, and predictors of level of participation in the Longitudinal Assessment of Bariatric Surgery (LABS-2) study

BACKGROUND

Reporting high-quality bariatric surgery outcomes depends on participant attrition and level of study participation among enrolled participants.

OBJECTIVES

Our aims are to report participant attrition, active enrollment, and level of participation, and to evaluate pre-surgery sociodemographic, physical health, and psychosocial factors as predictors of attrition and level of participation through 5 years.

SETTING

The Longitudinal Assessment of Bariatric Surgery-2 study which enrolled 2458 adults undergoing a first bariatric surgical procedure at 1 of 6 US cites from 2006 through 2009.

METHODS

In-person research assessments were conducted pre-surgery and annually for five years. Extensive retention strategies including offering remote assessments (telephone, email, mail, or a combination) were fully implemented in 2009. Among living participants, including those inactivated, annual follow-up assessments were categorized as in-person, remote or missed through 5 years.

RESULTS

By year 5, 1.7% of participants had died and 3.2% had withdrawn or were inactivated by the study staff; thus, attrition was 4.9% (n = 121). Controlling for site and calendar year, missed assessments increased from 14.7%-21.8% between years 1 and 2 and then stayed relatively stable (20.8%-19.6%) for years 3-5. Younger age, male sex, White race, lower body mass index, smoking, illicit drug use, and higher weight loss expectations preoperatively were independently associated with a higher likelihood of a missed versus in-person assessment across follow-up.

CONCLUSION

The LABS-2 participant attrition was low. The percentage of missed assessments did not increase after year 2, perhaps due to implementation of a comprehensive retention plan. Predictors of missed assessments highlight subgroups to target for focused retention efforts.

Effect of Adding Telephone-Based Brief Coaching to an mHealth App (Stay Strong) for Promoting Physical Activity Among Veterans: Randomized Controlled Trial

Background

Though maintaining physical conditioning and a healthy weight are requirements of active military duty, many US veterans lose conditioning and rapidly gain weight after discharge from active duty service. Mobile health (mHealth) interventions using wearable devices are appealing to users and can be effective especially with personalized coaching support. We developed Stay Strong, a mobile app tailored to US veterans, to promote physical activity using a wrist-worn physical activity tracker, a Bluetooth-enabled scale, and an app-based dashboard. We tested whether adding personalized coaching components (Stay Strong+Coaching) would improve physical activity compared to Stay Strong alone.

Objective

The goal of this study is to compare 12-month outcomes from Stay Strong alone versus Stay Strong+Coaching.

Methods

Participants (n=357) were recruited from a national random sample of US veterans of recent wars and randomly assigned to the Stay Strong app alone (n=179) or Stay Strong+Coaching (n=178); both programs lasted 12 months. Personalized coaching components for Stay Strong+Coaching comprised of automated in-app motivational messages (3 per week), telephone-based human health coaching (up to 3 calls), and personalized weekly goal setting. All aspects of the enrollment process and program delivery were accomplished virtually for both groups, except for the telephone-based coaching. The primary outcome was change in physical activity at 12 months postbaseline, measured by average weekly Active Minutes, captured by the Fitbit Charge 2 device. Secondary outcomes included changes in step counts, weight, and patient activation.

Results

The average age of participants was 39.8 (SD 8.7) years, and 25.2% (90/357) were female. Active Minutes decreased from baseline to 12 months for both groups (P<.001) with no between-group differences at 6 months (P=.82) or 12 months (P=.98). However, at 12 months, many participants in both groups did not record Active Minutes, leading to missing data in 67.0% (120/179) for Stay Strong and 61.8% (110/178) for Stay Strong+Coaching. Average baseline weight for participants in Stay Strong and Stay Strong+Coaching was 214 lbs and 198 lbs, respectively, with no difference at baseline (P=.54) or at 6 months (P=.28) or 12 months (P=.18) postbaseline based on administrative weights, which had lower rates of missing data. Changes in the number of steps recorded and patient activation also did not differ by arm.

Conclusions

Adding personalized health coaching comprised of in-app automated messages, up to 3 coaching calls, plus automated weekly personalized goals, did not improve levels of physical activity compared to using a smartphone app alone. Physical activity in both groups decreased over time. Sustaining long-term adherence and engagement in this mHealth intervention proved difficult; approximately two-thirds of the trial’s 357 participants failed to sync their Fitbit device at 12 months and, thus, were lost to follow-up.

Trial Registration

ClinicalTrials.gov NCT02360293; https://clinicaltrials.gov/ct2/show/NCT02360293

International Registered Report Identifier (IRRID)

RR2-10.2196/12526

Understanding the Effect of Adding Automated and Human Coaching to a Mobile Health Physical Activity App for Afghanistan and Iraq Veterans: Protocol for a Randomized Controlled Trial of the Stay Strong Intervention

Background

Although maintaining a healthy weight and physical conditioning are requirements of active military duty, many US veterans rapidly gain weight and lose conditioning when they separate from active-duty service. Mobile health (mHealth) interventions that incorporate wearables for activity monitoring have become common, but it is unclear how to optimize engagement over time. Personalized health coaching, either through tailored automated messaging or by individual health coaches, has the potential to increase the efficacy of mHealth programs. In an attempt to preserve conditioning and ward off weight gain, we developed Stay Strong, a mobile app that is tailored to veterans of recent conflicts and tracks physical activity monitored by Fitbit Charge 2 devices and weight measured on a Bluetooth-enabled scale.

Objective

The goal of this study is to determine the effect of activity monitoring plus health coaching compared with activity monitoring alone.

Methods

In this randomized controlled trial, with Stay Strong, a mobile app designed specifically for veterans, we plan to enroll 350 veterans to engage in an mHealth lifestyle intervention that combines the use of a wearable physical activity tracker and a Bluetooth-enabled weight scale. The Stay Strong app displays physical activity and weight data trends over time. Enrolled participants are randomized to receive the Stay Strong app (active comparator arm) or Stay Strong + Coaching, an enhanced version of the program that adds coaching features (automated tailored messaging with weekly physical activity goals and up to 3 telephone calls with a health coach—intervention arm) for 1 year. Our primary outcome is change in physical activity at 12 months, with weight, pain, patient activation, and depression serving as secondary outcome measures. All processes related to recruitment, eligibility screening, informed consent, Health Insurance Portability and Accountability Act authorization, baseline assessment, randomization, the bulk of intervention delivery, and outcome assessment will be accomplished via the internet or smartphone app.

Results

The study recruitment began in September 2017, and data collection is expected to conclude in 2019. A total of 465 participants consented to participate and 357 (357/465, 77%) provided baseline levels of physical activity and were randomized to 1 of the 2 interventions.

Conclusions

This novel randomized controlled trial will provide much-needed findings about whether the addition of telephone-based human coaching and other automated supportive-coaching features will improve physical activity compared with using a smartphone app linked to a wearable device alone.

Trial Registration

ClinicalTrials.gov NCT02360293; https://clinicaltrials.gov/ct2/show/NCT02360293 (Archived by WebCite at http://www.webcitation.org/75KQeIFwh)

International Registered Report Identifier (IRRID)

DERR1-10.2196/12526

Reporting missing participant data in randomised trials: systematic survey of the methodological literature and a proposed guide

OBJECTIVES

We conducted a systematic survey of the methodological literature to identify recommended approaches for how and what randomised clinical trial (RCT) authors should report on missing participant data and, on the basis of these approaches, to propose guidance for RCT authors.

METHODS

We defined missing participant data (MPD) as missing outcome data for trial participants. We considered both categorical and continuous outcome data. We searched MEDLINE and the Cochrane Methodology Register for articles in which authors proposed approaches to reporting MPD from RCTs. We selected eligible articles independently and in duplicate and extracted data in duplicate. Using an iterative process of discussion and revisions, we used the findings to develop guidance.

RESULTS

Of 10,501 unique citations identified, 13 articles reporting on 10 approaches proved eligible. The identified approaches recommend reporting the following aspects (from most to least frequently recommended): number of participants with MPD (n=10), reasons for MPD (n=7), methods used to handle MPD in the analysis (n=4), flow of participants (n=3), pattern of missingness (eg, whether at random) (n=3), differences in rates of MPD between trial arms (n=2), differences between participants with and without MPD (n=2), results of any sensitivity analyses (n=2), implication of MPD on interpreting the results (n=2) and methods used to prevent missing data (n=1). We propose a guide with nine items related to reporting the number, reasons, patterns, analytical methods and interpretation of MPD.

CONCLUSIONS

Most identified approaches invite trial authors to report the extent of MPD and the underlying reasons. Fewer approaches focus on reporting missingness patterns, methods for handling MPD and implications of MPD on results. Our proposed guidance could help RCT authors to better report, and readers to better identify participants with missing data.

Patterns and predictors of attrition in a trial of a housing intervention for homeless people with mental illness

PURPOSE

Participant retention is an important challenge in longitudinal research on homeless people. High attrition can threaten validity, and may represent lost opportunities to deliver interventions. In this article, we report on attrition in the At Home/Chez Soi study, a multi-site randomized controlled trial of a housing intervention for homeless people with mental illness.

METHODS

We first calculate life tables, and then use clustered logistic regression to implement a discrete-time survival model. We use splines and indicator variables to capture non-linear and group-specific variation over time in the hazard function. As potential predictors, we consider study group, site, date of recruitment, age, sex, baseline substance dependence, baseline psychotic disorder, time homeless in life, community functioning, and education.

RESULTS

The study recruited 2,148 homeless people with mental illness. Of these, 1,158 were randomized to the housing first intervention (HF), and 990 to treatment as usual (TAU). Excluding 79 people known to have died, attrition was 14%. This proportion was higher in TAU than in HF (21 vs. 8%, p < 0.01). Attrition was significantly lower in one site than elsewhere, and was also lower among those with substance dependence (13 vs. 18%, p < 0.01) and among those enrolled earlier in the study. The hazard varied over time in complex ways.

CONCLUSIONS

Results imply that study factors are more important than participant characteristics as determinants of retention, and that the immediate period after randomization is a crucial one. The high overall retention demonstrates the effectiveness of existing techniques for retaining participants.

Weight and weight change following breast cancer: evidence from a prospective, population-based, breast cancer cohort study

BACKGROUND

While weight gain following breast cancer is considered common, results supporting these findings are dated. This work describes changes in body weight following breast cancer over 72 months, compares weight with normative data and explores whether weight changes over time are associated with personal, diagnostic, treatment or behavioral characteristics.

METHODS

A population-based sample of 287 Australian women diagnosed with early-stage invasive breast cancer was assessed prospectively at six, 12, 18 and 72 months post-surgery. Weight was clinically measured and linear mixed models were used to explore associations between weight and participant characteristics (collected via self-administered questionnaire). Those with BMI changes of one or more units were considered to have experienced clinically significant changes in weight.

RESULTS

More than half (57%) of participants were overweight or obese at 6 months post-surgery, and by 72 months post-surgery 68% of women were overweight or obese. Among those who gained more weight than age-matched norms, clinically significant weight gain between 6 and 18 months and 6 and 72 months post-surgery was observed in 24% and 39% of participants, respectively (median [range] weight gain: 3.9 kg [2.0-11.3 kg] and 5.2 kg [0.6-28.7], respectively). Clinically-significant weight losses were observed in up to 24% of the sample (median [range] weight loss between 6 and 72 months post-surgery: -6.4 kg [-1.9--24.6 kg]). More extensive lymph node removal, being treated on the non-dominant side, receiving radiation therapy and lower physical activity levels at 6 months was associated with higher body weights post-breast cancer (group differences >3 kg; all p < 0.05).

CONCLUSIONS

While average weight gain among breast cancer survivors in the long-term is small, subgroups of women experience greater gains linked with adverse health and above that experienced by age-matched counterparts. Weight change post-breast cancer is a contemporary public health issue and the integration of healthy weight education and support into standard breast cancer care has potential to significantly improve the length and quality of cancer survivorship.

Treatment adherence and health outcomes in MSM with HIV/AIDS: patients enrolled in "one-stop" and standard care clinics in Wuhan China

BACKGROUND

Conducted in Wuhan China, this study examined follow-up and health markers in HIV patients receiving care in two treatment settings. Participants, all men who have sex with men, were followed for 18-24 months.

METHOD

Patients in a "one-stop" service (ACC; N = 89) vs those in standard care clinics (CDC; N = 243) were compared on HIV treatment and retention in care outcomes.

RESULTS

Among patients with CD4 cell count ≦350 cells/µL, the proportion receiving cART did not differ across clinic groups. The ACC was favored across five other indicators: proportion receiving tests for CD4 cell count at the six-month interval (98.2% vs. 79.4%, 95% CI 13.3-24.3, p = 0.000), proportion with HIV suppression for patients receiving cART for 6 months (86.5% vs. 57.1%, 95% CI 14.1-44.7, p = 0.000), proportion with CD4 cell recovery for patients receiving cART for 12 months (55.8% vs. 22.2%, 95% CI 18.5-48.6, p = 0.000), median time from HIV confirmation to first test for CD4 cell count (7 days, 95% CI 4-8 vs. 10 days, 95% CI 9-12, log-rank p = 0.000) and median time from first CD4 cell count ≦350 cells/µL to cART initiation (26 days, 95% CI 16-37 vs. 41.5 days, 95% CI 35-46, log-rank p = 0.031). Clinic groups did not differ on any biomedical indicator at baseline, and no baseline biomedical or demographic variables remained significant in the multivariate analysis. Nonetheless, post-hoc analyses suggest the possibility of self-selection bias.

CONCLUSIONS

Study findings lend preliminary support to a one-stop patient-centered care model that may be useful across various HIV care settings.

Longitudinal Assessment of Bariatric Surgery (LABS): retention strategy and results at 24 months

BACKGROUND

Retaining participants in observational longitudinal studies after bariatric surgery is difficult yet critical because the retention rate affects interpretation and generalizability of results. Strategies for keeping participants involved in such studies are not commonly published. The objective of this study was to review LABS retention strategies and present the 24-month retention data.

METHODS

The LABS Consortium monitors an observational cohort study of 2458 adults enrolled before bariatric surgery at 10 centers within the United States (LABS-2). To maximize data completeness, the investigators developed retention strategies, including flexible scheduling, a call protocol, reminder letters, abbreviated visit options, honoraria, travel reimbursement, providing research progress reports, laboratory results, newsletters, study website, and retention surveys. Strategies for locating participants included frequent updates of contact information, sending registered letters, and searching medical and public records.

RESULTS

At 12 and 24 months, 2426 and 2405 participants remained active, with vital status known for 98.7% and 97.3% and weight obtained for 95.2% and 92.2%, respectively. There were 148 missed visits (6.2%) at 24 months primarily because of inability to contact the participant. Only 15 (0.6%) active participants at 24 months missed all follow-up visits. Although 42 participants could not be located or contacted at 6 months, data were obtained for 23 (54.7%) of them at 12 months, and of the 52 participants who could not be located or contacted at 12 months, data were obtained for 18 (34.6%) at 24 months.

CONCLUSION

Longitudinal studies provide the ability to evaluate long-term effects of bariatric surgical procedures. The retention achieved in LABS is superior to that of many published reports but requires extensive effort and resources. This report identifies useful retention strategies. Further research is needed to identify the efficacy and cost-effectiveness of specific retention strategies.

Risk factors for non-adherence and loss to follow-up in a three-year clinical trial in Botswana

Background

Participant non-adherence and loss to follow-up can compromise the validity of clinical trial results. An assessment of these issues was made in a 3-year tuberculosis prevention trial among HIV-infected adults in Botswana.

Methods and Findings

Between 11/2004–07/2006, 1995 participants were enrolled at eight public health clinics. They returned monthly to receive bottles of medication and were expected to take daily tablets of isoniazid or placebo for three years. Non-adherence was defined as refusing tablet ingestion but agreeing to quarterly physical examinations. Loss to follow-up was defined as not having returned for appointments in ≥60 days. Between 10/2008–04/2009, survey interviews were conducted with 83 participants identified as lost to follow-up and 127 identified as non-adherent. As a comparison, 252 randomly selected adherent participants were also surveyed. Multivariate logistic regression analysis was used to identify associations with selected risk factors. Men had higher odds of being non-adherent (adjusted odds ratio (AOR), 2.24; 95% confidence interval [95%CI] 1.24–4.04) and lost to follow-up (AOR 3.08; 95%CI 1.50–6.33). Non-adherent participants had higher odds of reporting difficulties taking the regimen or not knowing if they had difficulties (AOR 3.40; 95%CI 1.75–6.60) and lower odds associated with each year of age (AOR 0.95; 95%CI 0.91–0.98), but other variables such as employment, distance from clinic, alcohol use, and understanding study requirements were not significantly different than controls. Among participants who were non-adherent or lost to follow-up, 40/210 (19.0%) reported that they stopped the medication because of work commitments and 33/210 (15.7%) said they thought they had completed the study.

Conclusions

Men had higher odds of non-adherence and loss to follow-up than women. Potential interventions that might improve adherence in trial participants may include:targeting health education for men, reducing barriers, clarifying study expectations, educating employers about HIV/AIDS to help reduce stigma in the workplace, and encouraging employers to support employee health.

Trial Registration

ClinicalTrials.gov NCT00164281

A multisite randomized trial of a motivational intervention targeting multiple risks in youth living with HIV: initial effects on motivation, self-efficacy, and depression

PURPOSE

Interventions targeting multiple risk behaviors are needed for youth living with HIV (YLH). A randomized clinical trial compared Healthy Choices, a four session motivational intervention targeting two of the three risk behaviors (HIV medication adherence, sexual risk behavior and substance use) to multidisciplinary specialty care alone. This article presents intermediary outcomes available at 3-month follow-up, variables proposed to be precursors to behavior change (motivation, self-efficacy, and depression).

METHODS

YLH (N=186) with at least one of the three problem behaviors were recruited from four sites in the Adolescent Trials Network and one non-Adolescent Trials Network site, and were assessed at baseline and 3 months.

RESULTS

Of the 94 youth randomly assigned to the treatment condition, 84% received at least one session, 67% received at least two sessions, 56% received at least three sessions, and 49% completed all four sessions. In intent-to-treat analysis, only depression was significantly improved in the treatment group as compared with controls. However, in per-protocol analysis, youth receiving at least two sessions of the intervention also showed significant improvements in motivational readiness to change as compared with youth in the control condition.

CONCLUSION

Results suggest the potential benefits of clinic-based motivational interventions for YLH who access these interventions. Delivering interventions in the community using an outreach model may improve access. Analysis of subsequent time points will determine effects on actual behavior change.

Percent total attrition: a poor metric for study rigor in hosted intervention designs

Health behavior interventions delivered at point of service include those that yoke an intervention protocol with existing systems of care (e.g., clinical care, social work, or case management). Though beneficial in a number of ways, such "hosted" intervention studies may be unable to retain participants that specifically discontinue their use of the hosting service. In light of recent practices that use percent total attrition as indicative of methodological flaws, hosted interventions targeting hard-to-reach populations may be excluded from consideration in effective intervention compendiums or research synthesis because of high attrition rates that may in fact be secondary to the natural flow of service use or unrelated to differential attrition or internal design flaws. Better methods to characterize rigor are needed.