Abstract
Older adults with cirrhosis commonly experience chronic noncancer pain managed with chronic opioid therapy. Current guidelines recommend opioid deprescribing for high-risk populations, including in cirrhosis, but data on tapering and discontinuation are scarce. We described opioid discontinuation rates and identified predictors of tapering or discontinuation. This retrospective cohort study of Medicare fee-for-service beneficiaries (N = 800,763) ≥ 65 years with continuous opioid use for ≥ 90 days. The primary outcome was opioid discontinuation (i.e., a gap in refills > 30 days). The secondary outcome included opioid tapering (i.e., a 35% decrease in average daily morphine milligram equivalents; [yes/no]). The primary exposure was diagnosed cirrhosis severity (i.e., none, compensated, decompensated). We estimated discontinuation rates using the Kaplan-Meier method, time to opioid discontinuation using proportional hazards regression, and predictors of tapering with logistic regression. After 1 year, 37% (95% CI = 37-37%) of individuals without cirrhosis discontinued opioids, similar to those with compensated (36% (34-37%)) and decompensated (37% (35-39%)) cirrhosis. Age did not modify the adjusted association between cirrhosis status and discontinuation (Wald χ2(4) = 7.72, p = 0.10) but calendar year (pre/post COVID-19) did (Wald χ2(1) = 26.57, p < 0.001). This finding indicated higher deprescribing rates prior to the pandemic, especially for those without cirrhosis (HR, 1.32; 95% CI, 1.31-1.33) compared with those with cirrhosis (HR, 1.13; 95% CI, 1.06-1.20). Non-opioid analgesic use, fall history, and frailty significantly increased the odds of tapering. In conclusion, these findings may reflect a lack of safe analgesic alternatives or higher pain burden in cirrhosis.