Association between insurance status, rurality, and quality of care in breast, colorectal, and lung cancer: Results from a statewide database in North Carolina.

Abstract

215 Background: Despite advances in cancer treatment, disparities in cancer care quality persist. Some prior work has shown disparities by insurance type and geography, but these findings have been mixed. We analyzed whether people with three commonly diagnosed cancers (breast, colorectal, lung) received guideline-concordant care and explored differences by insurance status, rural residence, and the interaction between insurance and rurality. Methods: We used the Cancer Information and Population Health Resource to link North Carolina state cancer registry data to multi-payer claims data for cancer patients diagnosed 2015-2019. Indicators, drawn from national guidelines, included: 1) recommended therapy based on HER2 status (breast cancer); 2) recommended therapy for ERPR+ patients (breast cancer); 3) receipt of chemo within 120 days of diagnosis (Stage III colon cancer); 4) receipt of chemotherapy within 270 days of diagnosis (Stage II-III rectal cancer); 5) no bevacizumab < 3 months of diagnosis (metastatic non-small-cell lung cancer [NSCLC]); 6) receipt of chemotherapy within 60 days of surgery (Stage II-IIIA NSCLC). For indicators with lower (<95%) adherence, we used adjusted odds ratios (aORs) to examine differences by insurance status, rural residence, and the interaction between the two. Covariates included age, race/ethnicity, and sex (lung/colorectal cancer only). Results: Sample sizes for these indicators ranged from 454 patients (colon cancer) to 3,460 patients (breast cancer). Two indicators showed high adherence: no bevacizumab for metastatic NSCLC (99%) and recommended therapy for breast cancer based on HER2 status (97%). Models for indicators with lower adherence showed two significant differences by insurance status. Compared to patients with Medicaid, receipt of appropriate care was more likely in ERPR+ breast cancer patients with private insurance (aOR=2.1, 95%CI [1.5, 2.8]) or Medicare (aOR=2.3, 95%CI [1.5, 3.4]). Compared to patients with Medicaid, rectal cancer patients with private insurance were more likely to receive chemotherapy within 270 days of diagnosis (aOR=3.3, 95% CI [1.7, 6.4]). There were no significant differences by rurality. The interaction between rurality and insurance was nonsignificant in all models. Conclusions: Medicaid-insured patients with breast and rectal cancer were significantly less likely to receive guideline-concordant cancer compared to people with private insurance, pointing to critical gaps. Policy changes such as better reimbursement for oncology services may be important in achieving equitable cancer care.