Geographic distribution of clinical trials may lead to inequities in access

Background: We sought to describe the geographic distribution of clinical trial sites across the continental USA and to identify drivers of trial site location. Methods/Results: Locations of 174,503 clinical trial sites were collected from 2002–2007 from the US FDA’s Bioresearch Monitoring Information System and geo-coded for spatial analysis. Predictors examined included urban population percentage (2000 US Census) and number of healthcare/social service and educational establishments (2002 Economic Census) per zip code. Extensive clustering of trial sites was detected. Urban composition and healthcare/social service facilities were strong predictors of the number of trial sites per zip code (p < 0.0001; R2= 0.69), but not their location (only 27% of clusters explained by these covariates).Conclusion: US clinical trial sites are highly clustered around urban areas with healthcare/social service facilities, which may partly explain why rural communities are underrepresented in clinical research.