Association of Primary Care Physicians Per Capita With COVID-19 Vaccination Rates Among US Counties

COVID-19 vaccines have helped slow the spread of SARS-CoV-2 in the US. However, maximal population uptake of vaccines has been hindered by vaccine hesitancy. Greater participation of primary care physicians (PCPs) in vaccine distribution has been proposed as a strategy to combat vaccine hesitancy.^1,2 Survey data suggest that a substantial portion of the unvaccinated population would be willing to get vaccinated if they had greater access to accurate information and receive encouragement from a trusted source.³ Primary care physicians can reach such individuals through direct engagement or alliances with community health workers, community centers, and mass-vaccination sites.⁴ Therefore, we sought to examine the association of the number of PCPs per capita with COVID-19 vaccination rates among US counties.

We conducted a cross-sectional study of 2739 counties and county equivalents (herein termed counties) of the 3142 total counties in the US (87.2%) as of August 23, 2021. Eight California counties with populations less than 20 000 and 6 states (Alaska, Georgia, Hawaii, Vermont, Virginia, and West Virginia) were excluded because of insufficient vaccination data and missing variables. We linked the number of PCPs (including general family medicine, general practice, general internal medicine, and general pediatrics physicians) per 100 000 population for each county and other key variables with data on the percentage of the population who were fully vaccinated against SARS-CoV-2 obtained from the US Centers for Disease Control and Prevention and 3 state health departments (Colorado, Massachusetts, and Texas) (eTable in the Supplement). This study was deemed exempt by the Massachusetts General Hospital Institutional Review Board because it involved analysis of only publicly available aggregated data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.

We examined the associations between the number of PCPs per 100 000 population and COVID-19 vaccination rates by using generalized estimating equation models with robust SEs after accounting for clustering within states and county population weights. The multivariable models were adjusted for demographic factors, urbanicity, socioeconomic status, and political leaning.⁵ We also conducted stratified analyses for metropolitan vs rural counties and Democratic vs Republican states (eMethods in the Supplement). All statistical analyses were performed using RStudio, version 1.4.1717. Two-sided P < .05 was considered statistically significant.

Among the 2739 US counties included in this study, PCPs were primarily concentrated in the Northeast, Florida, and many counties in the Midwest and West (Figure, A). This finding roughly corresponded to the distribution of counties with higher COVID-19 vaccination rates (Figure, B). After adjustment for potential confounders, counties in the highest decile of the number of PCPs per 100 000 population were associated with a 5.5% higher vaccination rate compared with those in the lowest decile (95% CI, 2.6%-8.4%) (Table). Every 10 additional PCPs per 100 000 population was associated with a 0.3% higher vaccination rate (95% CI, 0.2%-0.4%). In stratified analyses, we observed similar positive associations between the number of PCPs per 100 000 population and vaccination rates in rural areas or those with less than 2500 urban population (0.5% higher rate; 95% CI, 0.3%-0.7%) and in the 10 states with the highest Republican vote share (0.4% higher rate; 95% CI, 0.2%-0.6%).

In this cross-sectional study, we found that the number of PCPs per 100 000 population was independently associated with higher COVID-19 vaccination rates in the US. Our findings suggest that PCPs play a critical role in ensuring vaccine acceptance, especially in resource-limited and vaccine-hesitant regions, potentially through counseling and building local community trust and partnerships before they had access to vaccines.⁶ Limitations of the study include the accuracy of COVID-19 vaccine administration data, potential unmeasured confounders associated with the number of PCPs per capita and vaccination rates (eg, vaccine mandate), and the difficulty of disentangling our primary exposure from access to vaccines. Nevertheless, this study’s results provide support for expanding COVID-19 vaccine distribution to PCPs. Although fully incorporating PCPs into vaccination campaigns poses many challenges,⁴ PCPs may leverage their role as trusted messengers of scientific knowledge and educate communities about the importance of vaccination.

Accepted for Publication: December 21, 2021.

Published: February 10, 2022. doi:10.1001/jamanetworkopen.2021.47920

Open Access: This is an open access article distributed under the terms of the CC-BY License. © 2022 Lo CH et al. JAMA Network Open.

Corresponding Author: Andrew T. Chan, MD, MPH, Massachusetts General Hospital, 100 Cambridge St, 15th Floor, Boston, MA 02114 (achan@mgh.harvard.edu).

Author Contributions: Dr Lo had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.

Concept and design: Lo, Chiu, Qian, Duval, Chan.

Acquisition, analysis, or interpretation of data: All authors.

Drafting of the manuscript: All authors.

Critical revision of the manuscript for important intellectual content: Lo, Chiu, Qian, Alhassan, Duval, Chan.

Statistical analysis: Lo, Alhassan, Chan.

Obtained funding: Chan.

Administrative, technical, or material support: Lo, Chiu, Qian, Khan, Duval, Chan.

Supervision: Duval, Chan.

Conflict of Interest Disclosures: Dr Chan reported receiving grants from Pfizer Inc and Zoe, Ltd, and personal fees from Pfizer Inc, Bayer Pharma AG, and Boehringer Ingelheim outside the submitted work. No other disclosures were reported.

Funding/Support: This study was supported in part by grants from the Massachusetts Consortium on Pathogen Readiness and the Mark and Lisa Schwartz Foundation.

Role of the Funder/Sponsor: The funders had no role in the design and conduct of the study; collection, management, analysis, and interpretation of the data; preparation, review, or approval of the manuscript; and decision to submit the manuscript for publication.

Disclaimer: The content is solely the responsibility of the authors and does not necessarily represent the official views of the funders.

Additional Contributions: We thank Hang Lee, PhD, and Douglas Hayden, PhD (members of the Harvard Catalyst Biostatistical Consulting Program), for their statistical assistance. These contributors were not compensated for their assistance.