摘要:BACKGROUND AND OBJECTIVES: National workforce models fail to capture the regional effect of residency programs, despite local control over decisions to open or close training sites. In the last 5 years, 37 (nearly 8% of total) family medicine residency programs have closed. We report on a novel approach to measuring the regional effect of residency training programs closures using a combination of quantitative and spatial methods. METHODS: American Medical Association Physician Masterfile records and residency graduate registries for 22 of 37 family medicine residency programs that closed between 2000--2006 were analyzed to determine regional patterns of physician practice, as well as the effect of graduates from closed programs on areas that otherwise would be Health Professional Shortage Areas (HPSAs). Program graduate data from two sampled programs were mapped using geographic information system software to display the distribution "footprint" of graduates regionally. RESULTS: Of the 1,545 graduates of the 22 programs, 21% of graduates practice in rural locations, and 68% are in full-county or partial-county HPSAs. Without the graduates of these programs, there would have been 150 additional full HPSA counties in 15 states. The spatial distribution of the graduates of two closed programs demonstrates their effect across multiple counties and states. CONCLUSIONS: The effect of closing family medicine residency programs is likely to go undetected for many years. Decisions regarding the fate of family medicine programs are often made without benefit of a full assessment. Local and regional effects on physician access are often recognized only after the fact. Novel approaches to analysis and display of local effects of closures are essential for policy decisions concerning physician workforce training.