Overdose Risk Associated with Opioid Use upon Hospital Discharge in Veterans Health Administration Surgical Patients.

OBJECTIVE: To determine an association between opioid use upon hospital discharge (ongoing and newly started) in surgical patients and risks of opioid overdose and delirium for the first year. DESIGN: Retrospective, cohort study. SETTING: Population-level study of Veterans Health Administration patients. SUBJECTS: All Veterans Health Administration patients (N = 64,391) who underwent surgery in 2011, discharged after one or more days, and without a diagnosis of opioid overdose or delirium from 90 days before admission through 30 days postdischarge (to account for additional opioid dosing in the context of chronic use). METHODS: Patients' opioid use was categorized as 1) no opioids, 2) tramadol only, 3) short-acting only, 4) long-acting only, 5) short- and long-acting. We calculated unadjusted incidence rates and the incidence rate ratio (IRR) for opioid overdose and drug delirium for two time intervals: postdischarge days 0-30 and days 31-365. We then modeled outcomes of opioid overdose and delirium for postdischarge days 31-365 using a multivariable extended Cox regression model. Sensitivity analysis examined risk factors for overdose for postdischarge days 0-30. RESULTS: Incidence of overdose was 11-fold greater from postdischarge days 0-30 than days 31-365: 26.3 events/person-year (N = 68) vs 2.4 events/person-year (N = 476; IRR = 10.80, 95% confidence interval [CI] = 8.37-13.92). Higher-intensity opioid use was associated with increasing risk of overdose for the year after surgery, with the highest risk for the short- and long-acting group (hazard ratio = 4.84, 95% CI = 3.28-7.14). Delirium (IRR = 10.66, 95% CI = 7.96-14.29) was also associated with higher opioid intensity. CONCLUSIONS: Surgical patients should be treated with the lowest effective intensity of opioids and be monitored to prevent opioid-related adverse events.