Primary Care Physician Density and Mortality in the United States

Abstract

BackgroundGeographic physician availability differences are associated with healthcare outcomes. However, the association between primary care physician (PCP) density and mortality outcomes is less well-established. MethodsThe study analyzed 2019 county-level nonfederal PCP data from the Health Resources and Services Administration Area Health Resource File and mortality data using the CDC WONDER (Wide-ranging Online Data for Epidemiologic Research). All-cause and cardiovascular disease (CVD)- related age-adjusted mortality rates (AAMR) per 100,000 population stratified by the number of PCPs per 100,000 quartiles were extracted. Using AAMRs as continuous variables, linear regression was performed to determine the association of AAMRs with PCPs per 100,000 (reference, first quartile), adjusting for the social vulnerability index (SVI). ResultsA total of 3142 counties were included in the analysis. Among counties stratified by PCPs per 100,000 quartiles, all-cause AAMRs were 828 (95% CI, 824-832) in the first quartile, 798 (95% CI, 796-801) in the second quartile, 737 (95% CI, 735-739) in the third quartile, and 679 (95% CI, 678-680) in the fourth quartile. Similar trends were seen in CVD-related AAMRs, which were 446 (95% CI, 443-449),439(95% CI, 437-441), 403 (95% CI, 402-404), and 365 (95% CI, 364-366), respectively. Counties without PCP (221, included in first quartile) had all-cause and CVD-related AAMR of 797 (95%CI, 783-812) and 430 (95%CI, 419-440), respectively. Compared with the first quartile, SVI-adjusted analyses showed β-coefficient (95%CI) of all-cause mortality for the second, third, and fourth quartiles of -4.11 (95% CI, -18.31, 10.08), -35.37 (95% CI, -49.57, -21.17) and -85.79 (95% CI, -100.10, -71.48). Similar results were observed for CVD-related AAMR. ConclusionHigher PCP per 100,000 is generally associated with better all-cause and CVD-associated mortality outcomes, however complex factors likely play a role in determining these outcomes in counties with lower PCP per 100,000, which warrant further investigation.