Association between physician characteristics and likelihood of caring for underserved patients with cancer.

1646 Background: Despite evidence that a diverse workforce may improve access to care for underserved patients, emerging policies threaten to limit diversity in the physician workforce. We assessed whether nonwhite and foreign-born physicians were more likely than White and US-born physicians to care for underserved (minority race, low-income, rural) Medicare beneficiaries with cancer. Methods: Using SEER-Medicare, we analyzed Medicare beneficiaries aged 18+ newly diagnosed with breast, colorectal, lung, or prostate cancer from 2015-2019. We linked patients to cancer physicians (medical, radiation, and surgical oncology) using claims in the 6 months after diagnosis. We assessed physician race/ethnicity (data from Association of American Medical Colleges) and country of birth (American Medical Association). For each physician race and foreign-born status, we calculated the standardized treatment ratio (STR) as the percent of that group's dyads with a given patient characteristic divided by the percent of all dyads with that patient characteristic. Thus, STR > 1 indicates physicians in that group are more likely than the average physician to treat patients with the given characteristic. Results: We linked 434,923 patients (30% breast, 16% colorectal, 29% lung, 25% prostate) to 35,257 physicians. The patient population (mean age 74.7) was 79% White, 8% Black, 7% Hispanic, 3% Asian, 17% Medicaid dual-eligible, and 16% rural. Black physicians were more likely than expected to care for Black patients (STR 2.55; 95% CI 2.48–2.63; Table) and dual-eligible patients (STR 1.26; 1.22–1.29). Hispanic physicians were more likely to care for Hispanic (1.97; 1.91–2.03) and dual-eligible (1.11; 1.08–1.14) patients. Asian physicians were more likely to care for Asian (1.93; 1.89–1.96) and Hispanic (1.18; 1.16–1.20) patients and less likely to care for rural (0.73; 0.72-0.74) patients. Foreign-born physicians (20.7% of physicians) were more likely to care for Asian (1.36; 1.33–1.39), Hispanic (1.25; 1.23–1.27), and dual-eligible (1.20; 1.18–1.21) patients and less likely to care for rural patients (0.90, 0.89-0.91). Conclusions: Nonwhite and foreign-born cancer physicians disproportionately care for nonwhite and low-income patients. Policies restricting entry of minority and foreign-born physicians into the oncology workforce may hinder access to care for underserved cancer patients. Standardized treatment ratios (STRs) by physician race, ethnicity and country of birth. Physician characteristic Patient Characteristic Asian Black Hispanic White Dual eligible Rural Race & ethnicity Asian 1.93 1.00* 1.17 0.95 1.09 0.73 Black 0.60 2.55 1.08 0.85 1.26 1.04 Hispanic 0.79 0.97* 1.97 0.94 1.11 1.01* White 0.68 0.92 0.81 1.04 0.88 1.09 Foreign born No 0.86 0.96 0.89 1.02 0.88 1.04 Yes 1.36 1.05 1.25 0.96 1.20 0.90 *Denotes STRs NOT significantly different from 1.0; all other values are significant (P<0.05).