Defining Primary Care Shortage Areas: Do GIS-based Measures Yield Different Results?

Purpose To examine whether geographic information systems (GIS)-based physician-to-population ratios (PPRs) yield determinations of geographic primary care shortage areas that differ from those based on bounded-area PPRs like those used in the Health Professional Shortage Area (HPSA) designation process. Methods We used geocoded data on primary care physician (PCP) locations and census block population counts from 1 US state to construct 2 shortage area indicators. The first is a bounded-area shortage indicator defined without GIS methods; the second is a GIS-based measure that measures the populations' spatial proximity to PCP locations. We examined agreement and disagreement between bounded shortage areas and GIS-based shortage areas. Findings Bounded shortage area indicators and GIS-based shortage area indicators agree for the census blocks where the vast majority of our study populations reside. Specifically, 95% and 98% of the populations in our full and urban samples, respectively, reside in census blocks where the 2 indicators agree. Although agreement is generally high in rural areas (ie, 87% of the rural population reside in census blocks where the 2 indicators agree), agreement is significantly lower compared to urban areas. One source of disagreement suggests that bounded-area measures may overlook some shortages in rural areas; however, other aspects of the HPSA designation process likely mitigate this concern. Another source of disagreement arises from the border-crossing problem, and it is more prevalent. Conclusions The GIS-based PPRs we employed would yield shortage area determinations that are similar to those based on bounded-area PPRs defined for Primary Care Service Areas. Disagreement rates were lower than previous studies have found.