BACKGROUND: Many older adults with cancer are coprescribed opioids/benzodiazepines
evidence on harms is lacking. METHODS: Using SEER-Medicare (2012-2019), we identified patients with breast, colorectal, or lung cancer. Cox proportional hazards models estimated the adjusted hazard ratios (HR) of coprescribing (measured from claims, at the day level) with the immediate risk of overdose, fall/fracture, and all-cause hospitalization. In a secondary analysis, models were stratified by ≥ 90 days of continuous medication supply. RESULTS: In our cohort of 107,288 patients, compared to those prescribed neither medication, those prescribed benzodiazepines had an increase in the immediate risk of falls/fractures (HR: 1.17, 95% CI: 1.02-1.34) and all-cause hospitalizations (HR: 1.08, 95% CI: 1.04-1.12). Patients prescribed opioids had an increase in the immediate risk of overdose (HR: 5.62, 95% CI: 4.86, 5.62), falls/fractures (HR: 1.56, 95% CI: 1.41, 1.73), and all-cause hospitalizations (HR: 1.26, 95% CI: 1.23, 1.30). HRs were similar for patients coprescribed opioids/benzodiazepines. For patients without continuous exposure to one or both medications, effects were larger for coprescribed opioids/benzodiazepines vs opioids for overdose (HR: 15.22, 95% CI: 8.79-26.35 vs HR: 6.85, 95% CI: 6.85- 26.35) and all-cause hospitalization (HR: 3.21, 95% CI: 2.60-3.96 vs HR: 1.98, 95% CI: 1.87- 2.10). CONCLUSIONS: Overall, compared to no prescribing, benzodiazepine prescribing and opioid prescribing were each associated with an increase in the immediate risk of adverse events. Effects were similar for those prescribed opioids and those coprescribed opioids/benzodiazepines. For patients with intermittent vs long-term medication exposure, coprescribing additionally increased the risk of overdose and all-cause hospitalization.