Variations of a group coaching intervention to support early-career biomedical researchers in Grant proposal development: a pragmatic, four-arm, group-randomized trial

BACKGROUND: Funded grant proposals provide biomedical researchers with the resources needed to build their research programs, support trainees, and advance public health. Studies using National Institutes of Health (NIH) data have found that investigators from underrepresented groups in the biomedical workforce are awarded NIH research grants at disproportionately lower rates. Grant writing training initiatives are available, but there is a dearth of rigorous research to determine the effectiveness of such interventions and to discern their essential features. METHODS: This 2 x 2, unblinded, group-randomized study compares the effectiveness of variations of an NIH-focused, grant writing, group coaching intervention for biomedical postdoctoral fellows and early-career faculty. The key study outcomes are proposal submission rates and funding rates. Participants, drawn from across the United States, are enrolled as dyads with a self-selected scientific advisor in their content area, then placed into coaching groups led by senior NIH-funded investigators who are trained in the intervention's coaching practices. Target enrollment is 72 coaching groups of 4-5 dyads each. Groups are randomized to one of four intervention arms that differ on two factors: [1] duration of coaching support (regular dose = 5 months of group coaching, versus extended dose = regular dose plus an additional 18 months of one-on-one coaching); and [2] mode of engaging scientific advisors with the regular dose group coaching process (unstructured versus structured engagement). Intervention variations were informed by programs previously offered by the NIH National Research Mentoring Network. Participant data are collected via written surveys (baseline and 6, 12, 18, and 24 months after start of the regular dose) and semi-structured interviews (end of regular dose and 24 months). Quantitative analyses will be intention-to-treat, using a 2-sided test of equality of the effects of each factor. An inductive, constant comparison analysis of interview transcripts will be used to identify contextual factors -- associated with individual participants, their engagement with the coaching intervention, and their institutional setting - that influence intervention effectiveness. DISCUSSION: Results of this study will provide an empirical basis for a readily translatable coaching approach to supporting the essential grant writing activities of faculty, fellows, and other research trainees, including those from underrepresented groups.