Incidence and Predictors of Diabetes Mellitus after a Diagnosis of Early-Stage Breast Cancer in the Elderly Using Real-World Data

Purpose The incidence and predictors of diabetes (DM) in patients with breast cancer (BC) were evaluated. We compared DM incidence and physician access in BC patients to matched controls. Methods We identified women with stage I-III BC diagnosed from 2005 to 2013 in the SEER-Medicare database, with >= 2 years of follow-up after diagnosis, without previous DM claims. Incident DM was determined by >= 1 DM claims after BC diagnosis. Multivariable analysis was used to identify factors associated with incident DM. Age- and race-matched non-cancer controls were obtained from a 5% random sample and assigned an index date. Physician and PCP visits per-patient-per-year were compared between cases and controls in the two-year period prior to and after the index date. Results Among 14,506 eligible BC patients, 3234 (22.3%) developed DM versus 16.5% of controls. Among BC patients, factors associated with incident DM included race (Black OR 1.63 95% CI 1.39-1.93, Hispanic OR 3.03 95% CI 1.92-4.81; vs. Caucasians), SES (Quintile 0 vs. Quintile 4 OR 1.55 95% CI 1.33-1.78), and receipt of chemotherapy (vs. none OR 1.19 95% CI 1.08-1.31). Among cases and controls, respectively, median physician visits per-patient-per-year were 19 and 17 prior to the index date, and 46 and 19 after the index date; median PCP visits were 2 for both groups in both periods. Conclusion About 22% of BC patients developed DM, more than controls in the same period. While there were differences in healthcare access, there weren't differences in PCP access between groups. This represents an opportunity for better comorbidity management in BC patients.