Multilevel factors affect participant retention in community-based health studies

Background: Health research is often hampered by high rates of avoidable loss to follow-up of participants. The individuals and groups most vulnerable to social inequalities and poor health may also be more difficult to recruit and retain in studies. Loss to follow-up can bias results, leading to inaccurate findings. Practice guidelines and policies that are based on research that does not adequately include minorities and low-income populations risk perpetuating health disparities because findings may apply more to advantaged populations than those that experience social disadvantage. Methods: A social epidemiology framework was applied and six focus group interviews with 48 researchers and outpatient research subjects were conducted at three UCSF research centers to identify multilevel factors in participant retention. Data were coded, and a content analysis was performed. Results: While a portion of the data confirms information available in the literature, other factors have emerged or have been discussed in new ways, including factors at the contextual level (e.g., funding policies, bureaucratic and institutional barriers), the study level (e.g., hiring protocols, participant recruitment procedures, placement and integration of the study into the larger community, availability and use of resources), and participant-level (e.g., employment/retirement status, caregiver issues, perceived benefit, desire for access to information, and a sense of community and connection to the healthcare deliver system). Conclusions: The context in which research is conducted, and the characteristics of research participants and studies may interact to affect retention rates in longitudinal studies. Investigators should explore multilevel strategies to improve retention of diverse participants in health studies.