| Location choices of health care providers have significant implications for public policies designed to alleviate geographic maldistribution of providers. The physician literature identifies certain affinities of the incumbents that significantly influence location choices. A predictive model based on affinity was developed in this research for initial and continued location of physician assistants (PAs) in rural and other underserved areas, as PAs are believed to improve access to care in those areas.;Data from the Physician Assistant Career Study, a longitudinal study of the class of 1977 at 45 PA programs in the United States, were used to scrutinize shifts in location choices across the careers of 1,126 participants since they matriculated in 1975. A binary logistic regression model of PA location choice was tested with gender, race/ethnicity, location of upbringing, PA program type, early commitment to primary care, specialty choice, and hospital job as independent variables, across the career---at matriculation, graduation, first job, 1989 and 2000. Career patterns were mapped with respect to location changes, "changers" and "non-changers" were compared, and work in underserved areas was analyzed.;The model successfully predicts PA location initially and also later in the career. Male gender, rural or small town background, early commitment to primary care, and primary care choice of specialty were predictive of rural location. Non-white race predicted urban/inner-city preference at matriculation, but attrition of blacks precluded testing of this variable later. Medex graduates seemed more likely than the others to prefer underserved areas, but this could not be fully explored because Medex graduates were underrepresented in the study. Career pattern analysis showed that earlier location preferences significantly influenced later location preferences. PAs with the above-mentioned affinities were more likely than the others to stay with their preferences for rural areas and to serve in medically underserved areas. Overall, this research demonstrated the value of the affinity model in predicting PA location. Policies based on this model for recruitment and retention of PAs in rural and underserved areas are likely to be more effective, as they would be better aligned with the individual inclinations and the free-market values of American society. |
