A spatial analysis of physician assistant programs

Purpose The Accreditation Review Commission on Education for the Physician Assistant projects a total of 273 accredited programs by the summer of 2020. Over the past 10 years, the number of Central Application Service for Physician Assistants (CASPA) applicants per seat has increased by 53%. However, no studies have addressed the current geographic relationship of applicants to programs and program growth. The purpose of this study was to describe the geospatial patterns and relationships of physician assistant (PA) programs and CASPA applicants. Methods Program directory information for established (n = 159), satellite (n = 18), and new PA programs (n = 95) was mapped using ArcGIS software. Permanent US ZIP codes for PA applicants (n = 22,603) from the 2014 to 2015 CASPA admissions cycle were also mapped. Point data were used to calculate the nearest neighbor by program type. Correlation was used to measure the association between PA applicants, program class size, and state population metrics. Results Most of the 95 new PA programs were geographically close to established programs. The median distance of new programs to the nearest neighboring established program was 25.6 miles (mean 39, standard deviation 38). Both established and new PA programs were found to be highly clustered (Moran’s I z score < 2.58, p =.01). The geographic distribution of the CASPA applicant pool was related to distribution of the US population, certified PAs, and practicing physicians. Conclusions PA program growth has exceeded projections. The close proximity of new programs to established programs will likely result in continued competition for quality applicants, PA faculty members, and clinical training sites.