Background

Reporting bariatric surgery outcomes, including adverse events, depends on participant retention, data validity and completeness.

Methods

The Longitudinal Assessment of Bariatric Surgery-2 (LABS-2) study enrolled 2458 adults undergoing a first bariatric surgical procedure at one of 6 US sites. Research assessments were conducted pre-surgery and annually post-surgery. Multiple retention strategies including flexible scheduling, alternative research assessment sites, home visits, telephone interviews, increased reimbursement, and shorter assessment option for high-priority data were implemented to enhance retention and data completeness. A mixed-effect multinomial logistic regression model identified pre-surgery factors independently related to study participation (i.e. in-person, remote or missed/inactivated assessments), adjusting for site, calendar year, and follow-up time point.

Results

Table 1 shows follow-up assessment status by time point. The modeled percentage of missed assessments plus inactivations increased from 14.7% to 21.8% between Year 1 and 2, then stayed relatively stable through year 5 (20.8%-19.6% in Years 3-5). Most mental health-related measures were not independently associated with study participation. Younger participants, males, whites, smokers, illicit drug users, those reporting less than very good general health, and who reported higher weight loss expectations, independently had a higher likelihood of a missed/inactivated versus in-person assessment across follow-up.

Conclusions

The percentage of participants who missed an assessment or were inactivated did not increase after Year 2, perhaps due to implementing of an increasingly robust retention plan. The identified predictors of missed/inactivated assessments highlight sub-groups to target for focused retention efforts. Flexibility in data collection may minimize participant burden, and reduce missed assessments and incomplete data