Primary Care Physician Density per Capita and its Effects on Diabetic Complications

Introduction: The state of Texas has a severe shortage of primary care physicians that has worsened in recent years, with only 4 of 254 counties demonstrating a sufficient number of primary care physicians, based on federal guidelines. This number is more pronounced in rural areas. A striking 82% of Texas counties have less than one primary care physician per 3,500 patients (1). In addition, the state of Texas has a significant population of diabetic patients needing physician management. Over 2 million people in Texas have diagnosed diabetes, with over 500,000 that are undiagnosed (2). Without sufficient treatment and surveillance, diabetes can progress to a host of problems including coronary artery disease, chronic kidney disease, retinopathy, amongst other morbidities. Researchers have elucidated the benefits of preventative care for stroke, heart disease, and cancer, but its correlation to diabetic outcomes has not been well documented. Therefore, we wanted to assess the effects of an increased number of primary care physicians per capita on preventing adverse diabetic outcomes in Texas counties.

Methods: To further evaluate the effect of primary care physicians on preventing worse diabetic outcomes, we collected data from Texas Health Data on the admission rate for uncontrolled diabetic admissions in both adult diabetics and the lower extremity amputations for uncontrolled adult diabetics per county in 2019 (3). We then collected the ratio of patients to primary care physicians in each county in the state of Texas in 2019, using the county health rankings database (4). We compiled data from both sources and performed a statistical analysis to evaluate the correlation between primary care physicians per capita and diabetic admissions and amputations in each Texas county. Descriptive statistics were estimated to describe sample characteristics. We estimated the predicted ratio of admissions and amputation rates using bootstrapped ratio testing generated with 50 samples. All data were analyzed in Stata and the level of significance was set at p <.05.

Results: An average population of 295,998 individuals across the state of Texas were part of the final analytic sample. Results showed the mean risk-adjusted admissions per 100,000 were 52.3 (SD = 25.9). Bootstrapped ratio testing indicated a significant ratio of increased amputees and admissions (Bootstrapped Observed ratio: 0.00042, bootstrapped standard error: 0.00001, 95% CI 0.00040, 0.00043), X2= 99.9, p <.0001, given an increased ratio of patients to physicians.

Conclusion: These data further establish the importance of having a sufficient number of primary care providers in rural and underserved areas. The state of Texas and Texas medical schools should establish further methods to encourage physicians to practice in rural areas to decrease the amount of uncontrolled diabetes and amputations in our populations.

Given these findings, we hypothesize that there is an association between the number of physicians per capita and diabetic outcomes. We would like to further evaluate which counties specifically have the least access to primary care physicians, higher diabetic rates, and most unfavorable outcomes. As we perform further research, we would also like to evaluate other confounding variables that were not considered.