Association of Physician Densities and Gynecologic Cancer Outcomes in the United States

Abstract

ABSTRACT Gynecologic cancers account for approximately 94,000 new cancer diagnosis yearly in the United States, with prognosis largely dependent on stage at diagnosis, sociodemographic, geographic, and other biological factors. Proximity to gynecologic oncologists has been shown to improve cancer outcomes; however, it remains unclear whether physician density of nonsubspecialists such as obstetrician-gynecologists (OBGYNs) or primary care physicians (PCPs) has a similar benefit. The aim of this study was to determine whether there is a correlation between the density of OBGYNs and PCPs with outcomes of patients with gynecologic cancer at the county level in the United States. County-level oncologic and sociodemographic data on cases between 2005 and 2018 were obtained from the Surveillance, Epidemiology, and End Results (SEER) Research Plus database. Female population by county was collected from the United States Census Bureau and physician density information from the Area Health Resources File. Backward stepwise linear regression was performed for both AJCC cancer stage at diagnosis and 5-year survival rate while accounting for county-level sociodemographic factors. A total of 113,938 patients were included in the analysis for AJCC stage at diagnosis and 98,573 for 5-year survival rate. Across all cancer types, the average density was 18 and 114 OBGYNs and PCPs per 100,000 females in metropolitan areas, respectively, whereas the average density was 8 and 89 OBGYNs and PCPs per 100,000 females in nonmetropolitan areas, respectively. Increasing PCP density was associated with an earlier stage at diagnosis for cervical cancer (95% CI, −6.27 to −0.05; P < 0.05), even after adjusting for sociodemographic factors. No correlation between OBGYN density and cervical cancer stage at diagnosis was found. After adjusting for county-level sociodemographic factors, increasing median household income was associated with an earlier stage at diagnosis for cervical cancer (P = 0.01) and increasing percentage of Black females was associated with a later stage at diagnosis for uterine cancer (P < 0.01). Increasing PCP density was associated with an increased 5-year survival rate for cervical cancer (95% CI, 0.03–0.09; P < 0.05). After adjustment for sociodemographic factors, no correlation between OBGYN density and cervical cancer survival was observed, nor was any correlation between PCP or OBGYN density and 5-year survival rate by stage for either endometrial or ovarian cancer. After adjustment for county-level sociodemographic factors, an increasing percentage of those with a bachelor’s degree was associated with an increased 5-year survival in cervical cancer (P < 0.01), an increasing percentage of Black females was associated with a decreased 5-year survival rate for uterine cancer (P < 0.01), and increasing median household income was associated with an increased 5-year survival rate for ovarian cancer (P < 0.01). The results of this study suggest that only cervical cancer outcomes are significantly correlated with PCP density at the county level, possibly because cervical cancer is the only gynecologic cancer with recommended routine screening guidelines.