Impact of Evolving Evidence on Pharmacological Therapy for Alcohol‐Associated Hepatitis in Alberta: A Population‐Based Study of Practice Trends and MELD ‐Stratified Effectiveness of Corticosteroids

BACKGROUND AND AIMS: Randomised trials such as the STOPAH trial questioned the efficacy of pentoxifylline (PTX) and provided mixed evidence for corticosteroids (CS) in alcohol-associated hepatitis (AH). We aimed to assess temporal trends in the use of PTX and CS in Alberta and to evaluate the real-world effectiveness of CS using a target trial emulation approach. APPROACH AND RESULTS: We conducted a population-based cohort study using linked administrative, laboratory and pharmaceutical data from Alberta, Canada. Adults hospitalised with AH between 2012 and 2023 who met National Institute on Alcohol Abuse and Alcoholism (NIAAA) criteria were included. Joinpoint regression evaluated changes in medication use and mortality trends. A sequence of emulated target trials estimated the effect of initiating CS within 7 days on 30- and 90-day mortality. Among 2706 admissions (2138 patients), PTX use declined sharply after 2014 and was nearly absent after 2016. CS use increased modestly during the same period. Thirty-day mortality decreased gradually after 2016, while 90-day mortality remained stable. In 1365 patients with MELD ≥ 20, CS use was associated with a non-significant 2% absolute reduction in 30-day mortality (hazard ratio, 0.86; 95% CI, 0.73-1.03), with no effect at 90 days. No benefit was observed in patients with MELD > 35. CONCLUSIONS: RCT evidence led to a rapid reversal in PTX prescribing prior to formal guideline updates. Real-world data support a modest short-term benefit of CS and illustrate the potential of target trial emulation to assess treatment effectiveness in specific groups of patients.