Research support infrastructure and productivity in US family practice residency programs

Purpose. To examine the associations between 11 research support infrastructural characteristics and measures of research productivity. Method. A questionnaire was mailed to 462 directors of non-military family practice residency programs in the United States. A total of I I research support infrastructural characteristics and six research productivity measures were coded. Initial analyses indicated a skewness in responses given by larger versus smaller programs. Respondents were divided into those from programs with eight or fewer full-time faculty and those from programs with nine or more. Separate analyses were run for each. Logistic regression was employed to determine which research support characteristics would best predict productivity in the top quartile. Results. Of the 461 recipients of deliverable questionnaires, 351 (76.1%) responded. A large proportion of programs reported no research productivity for the preceding 12 months. Separate stepwise logistic regression analyses were run for small and large programs; the ability of the I I characteristics to predict research-productive programs varied with size. Employment of full-time research professionals was the only characteristic positively associated with research productivity for both groups. For small programs, research productivity was positively associated with the requirement of faculty members to do research. For large programs, it was positively associated with both the presence of fellowship programs and the presence of a specific, written research strategic plan. Conclusion. This study demonstrated a positive association between several elements of research infrastructure and research productivity but that such infrastructure is inconsistent across programs and seemingly insufficient to develop the necessary research culture and socialization.