6C-01 - Comparative Safety of Acetaminophen-Containing Hydrocodone versus Oxycodone among Adults with Cirrhosis

Background: Half of patients with cirrhosis live in chronic pain, which is a leading cause of disability. Patients with cirrhosis are prone to developing adverse drug reactions from pharmacotherapies for pain. Opioid use may confound the clinical picture of hepatic encephalopathy, and excessive acetaminophen can cause hepatotoxicity. Yet, opioid-acetaminophen combination therapy is commonly used for pain in this population.

Objectives: To compare hospitalization events after initiating acetaminophen-containing hydrocodone vs. oxycodone among community-dwelling older adults with cirrhosis.

Methods: We conducted a retrospective active comparator new user study of commercially insured U.S. adults using MarketScan (2012-2023) and Optum Clinformatics® Data Mart (CDM) Database (2004-May, 2024) with cirrhosis, who initiated acetaminophen-containing hydrocodone or oxycodone in an outpatient setting. Safety outcomes included all-cause mortality, all-cause hospitalization, decompensated cirrhosis, ascites, varices, hepatic encephalopathy, esophageal variceal hemorrhage, and fracture. Flu vaccination served as a negative control outcome. Propensity score matching was used to balance 141 covariates assessed in the 183-days preceding the dispensing date of hydrocodone or oxycodone (i.e., the cohort entry date [CED]). We used an intent-to-treat analysis with Cox proportional hazard regression to estimate hazard ratios (HR) and generalized linear models to estimate risk difference (RD) for safety outcomes. Database-specific estimates were pooled with inverse-variance weighting.

Results: We matched 34,613 pairs of hydrocodone and oxycodone initiators pooled from MarketScan and Optum CDM (Mean age [SD]=53.3 [17] years, women=56%). Compared to oxycodone initiators, hydrocodone initiation decreased the likelihood of decompensated cirrhosis (HR [95% CI], 0.91 [0.85, 0.98]; RD [95% CI] per 1,000 PYs, -16.1 [-28.9, -3.2]), ascites (0.88 [0.81, 0.96]; -14.1 [-24.4, -3.8]), and all-cause hospitalization (0.93 [0.90, 0.97]; 2.6 [-3.5, 8.7]). There was no significant difference in rates of mortality, varices, hepatic encephalopathy, esophageal variceal hemorrhage, and fracture between the two groups. The negative control outcome of flu vaccine was not different between groups (1.03 [0.97, 1.10]; 8.8 [-7.4, 24.9]).

Conclusions: Initiating acetaminophen-containing hydrocodone, compared to acetaminophen-containing oxycodone, was associated with a lower likelihood of liver decompensation among community-dwelling adults with cirrhosis. These findings will inform clinical practice guidelines to safely manage disabling pain in this high-risk population.