Maximizing Retention in an Urban Prospective Cohort Study

• Main Authors: Murray, Elaina, Beatty, Kate E., Flick, Louise H., Elliot, Michael, John, Lisa V., Thompson-Sanders, Vetta, King, Allison, Bernaix, Laura W., Leduc, Candi, Lacy, Elizabeth, Helmkamp, Kristi
• Published: Digital Commons @ East Tennessee State University 2014
• Subjects: retention; urban; prospective cohort study; Health Services Management and Policy; Health Services Research; Public Health
• Online Access: https://dc.etsu.edu/etsu-works/6857; https://apha.confex.com/apha/142am/webprogram/Paper306068.html

BACKGROUND: Retaining participants in multi-year prospective cohort studies presents challenges, especially in urban settings. Early identification of participants at risk for attrition may enhance retention. We examine the validity of two risk for loss-to-follow-up assessments and early retention efforts in one Primary Sampling Unit during the National Children’s Study pilot. Our goal was to identify cases requiring additional attention. Retention challenges included high poverty, frequent moves, lack of spousal support, and mistrust of research. METHODS: Recruitment ended in 2012 and research activities shifted to retention. Data collectors (DC) completed subjective risk assignments (low, medium, high) based on knowledge of participants. Descriptive statistics compared risk assessments to socio-demographic characteristics, responses regarding participation, and missed appointments 11 months after risk assessment. RESULTS: We recruited approximately 100 participants. Higher perceived risk was associated with greater likelihood for mothers to be minorities, younger, and have lower education and income (X2=15.362, p<.01; X2=12.118, p<.05; X2=9.947. p<.01; and X2= 7.720, p<.05 respectively). Participants with income below federal poverty placed higher values on receiving incentives (X2= 6.011 p<.05). African American or “other” race participants placed a higher value on feeling comfortable with the interviewers than White respondents (X2=12.539 p<.01). Risk assignment and race were associated with number of missed appointments (X2=8.698 p<.01; X2 =4.307, p<.05). CONCLUSION: Results suggest DCs’ subjective assessment of risk predicts number of missed appointments. Future research might consider strategies to improve African American and “other” race participants’ comfort with interviewers. The ethics of dollar amounts for incentives among low-income participants remain a concern.