Primary care provider exit and cancer stage at diagnosis among Medicare beneficiaries.

Abstract

147 Background: Primary care clinicians play a central role in cancer screening, symptom evaluation, and timely referral. As the primary care workforce ages and turns over, patients may experience disruptions in continuity before cancer diagnosis. Evidence on the downstream consequences of these disruptions for cancer detection remains limited. Methods: We used linked SEER–Medicare data for beneficiaries diagnosed with breast or colorectal cancer during 2010–2018. Patients were attributed to a primary care provider (PCP) based on the plurality of office-based evaluation and management visits in the 24 months before diagnosis. PCP exit was defined as the attributed PCP exiting within 24 months before diagnosis through durable relocation or sustained cessation of Medicare billing. We estimated linear probability models adjusting for patient demographics, comorbidity, neighborhood socioeconomic status, SEER registry indicators, diagnosis year, area fixed effects, and practice context. Results: PCP exit increased the probability of late-stage diagnosis by 0.9 percentage points, a 2.4% relative increase from the 36.8% baseline among unexposed patients. Estimated effects were present for both cancer sites: 0.7 percentage points for breast cancer and 1.4 percentage points for colorectal cancer. Effects were larger when the PCP exit occurred within six months of diagnosis. Mechanism analyses suggested reductions in pre-diagnosis detection opportunities. Among breast cancer patients, PCP exit was associated with a 1.9 percentage-point decline in screening mammography and a 0.8 percentage-point decline in diagnostic mammography. Among colorectal cancer patients, PCP exit was associated with a 0.6 percentage-point decline in preventive screening and a 0.6 percentage-point decline in claims-based diagnostic evaluation for gastrointestinal or rectal bleeding. Conclusions: PCP exit before cancer diagnosis was associated with reduced use of selected preventive and diagnostic services and a higher probability of late-stage diagnosis among older adults. These findings suggest that primary care workforce disruptions may have downstream consequences for cancer detection and highlight the importance of continuity, care transitions, and referral systems that preserve timely diagnostic evaluation when patients lose an established PCP.