Who stays engaged in long term follow up studies? A study of adolescent retention across 7 years.

Abstract

Abstract Background Mental health and substance use disorders typically emerge early in life with impacts extending into early adulthood. Longitudinal studies are well-placed to investigate the progression of mental health and substance use disorders in adolescents. However, a significant challenge of longitudinal studies is sample attrition, or failure to retain participants. The impact of attrition may be reduced by understanding factors related to drop out, which can then be mitigated through optimised research design. To date, there remains a gap in the literature in determining characteristics of adolescent participants that are associated retention in school-based studies. Methods This paper aimed to; (1) investigate the baseline characteristics of adolescent participants who are retained in long-term follow-up research, and (2) retrospectively explore the effectiveness of methods used to retain adolescents in long-term follow up. Data from the Climate Schools Combined (CSC) study was used to understand which baseline participant characteristics predicted retention at either 60- and 72-month long term follow-up occasions. Of the 5,417 eligible participants, 1877 (34.65%; 61.40% female) were retained at either the 5- or 6-year post-baseline survey occasion. On average, participants were aged 18.5 years (SD = 0.60) at the 5-year follow-up and 19.9 years (SD = 0.50) at the 6-year follow-up. Results Being female (OR = 1.22, [1.03–1.45]), reporting grades between 80–89% (OR = 2.01, [1.21–3.32]) and 90–100% (OR = 2.41, [1.46–3.98]), higher self-reported depression (PHQ) scores (OR = 1.03, [1.02–1.05]) and higher mental health knowledge scores (OR = 1.08, [1.05–1.11]) at baseline were significant predictors of participant retention in long term follow-up. Conclusions Longitudinal clinical trial research is key to better understanding trajectories of mental health and substance use and the durability of prevention programs. Researchers need to be aware of the constantly changing technological landscape, including social media and shifting privacy policies, that may impact the ability to contact study participants. Researchers should consider these factors when designing retention strategies for groups that are more likely to drop out if they are to minimise the impact of missing data on results. This paper has outlined several practical strategies that can be implemented in future research trials.