RESUMEN
BACKGROUND AND OBJECTIVES: Despite a growing need for family practice to contribute to the national primary care research agenda, the specialty is ill-equipped to assume a more active role. Information about residency programs that are successful in research is a valuable resource for increasing family medicine's research capacity. METHODS: A three-stage investigation was completed in May 1996, consisting of 1) a telephone survey of family practice residency program directors, 2) a mail survey of recent graduates from relatively successful programs identified in stage 1, and 3) in-depth interviews with the program directors or research directors identified by combining data from the first two stages. RESULTS: Most residents in the programs included in stage 2 completed a research project (68.7%) and currently have an interest in practice-based research (57.2%). Residents from programs selected for the study's final stage were more likely to have published a research article (32% versus 20.3%) and to have completed a project while a resident (81% versus 60.1%) than those from the programs not selected. Virtually unanimous characteristics of successful programs include program director support of research, time for research, faculty involvement, a research curriculum, professional support, and opportunities for presenting research. CONCLUSIONS: Individual family practice residencies can be considered to be at one of three levels with respect to their level of research activity: 1) relatively undeveloped, 2) developing, or 3) relatively developed. Programs can expect successful results if they make research a priority, and means are needed for communicating successful strategies between programs.