Intrinsic motivation of GPs was not related to recruitment success, whereas interest in the study topic was

Limitations of this study are that we retrospectively surveyed physicians, using a convenience sample when selecting trials, not performing a multilevel analysis because of the limited sample size, the presence of responder bias as participation in this study predicted successful recruitment, and that the UWES-9 and the other questions were not specifically designed to measure intrinsic motivation for recruitment. The goal should be to build and validate a prediction model for recruitment success that could help optimize recruitment to clinical trials in primary care.CRediT authorship contribution statement Henk van der Worp: Formal analysis, Writing - original draft, Writing - review & editing. The authors thank Dr Robert Sykes (www.doctored.org.uk) for providing editorial services.Supplementary data Data Profile Measures of motivation, interest, and relevance Univariable analyses Adjusted analyses Nagelkerke R2 Hosmer–Lemeshow test AUC Measures of motivation, interest, and relevance UWES GP 0.96 (0.59–1.56) 0.93 (0.55–1.55) 0.07 0.64 0.59 UWES recruiter 1.32 (0.95–1.83) 1.26 (0.86–1.84) 0.09 0.39 0.62 Motivated to participate 1.49 (0.83–2.66) 1.50 (0.79–2.86) 0.09 0.41 0.89 Participate actively when invited 1.19 (0.72–1.97) 1.15 (0.67–1.98) 0.08 0.67 0.61 Interest in the topic 1.99 (1.15–3.43) 2.11 (1.18–3.78) 0.15 0.49 0.67 Relevance of the topic 1.69 (0.98–2.92) 1.60 (0.89–2.85) 0.10 0.92 0.64 Potential confounders GP active researcher (ref = no) 0.98 (0.28–3.40) Obtained a PhD (ref = no) 4.05 (1.03–15.84) Table 1 Predictors of recruitment success