Meta-Analysis of Drug Therapy in Patients With Severe Alcoholic Hepatitis

Statistical Analysis Plan for an Individual Patient Data Meta-analysis of Ten Clinical Trials Comparing Drug Therapy in Patients With Severe Alcoholic Hepatitis

In the specific setting of the evaluation of corticosteroids, pentoxifylline of their combination in severe alcoholic hepatitis, only meta-analysis combining individual data is able to provide detailed information from each individual with severe alcoholic hepatitis assessed by a DF ≥ 32. The need for such an approach is confirmed by the fact that in both univariate and multivariate analyses, truth survival is lower for conclusions from meta-analysis of the literature than for conclusions derived from non-meta-analyses. The present study is a meta-analysis of individual data from RCTs restricted to patients with a DF ≥ 32. The primary endpoint will be to compare 28-day survival of patients receiving either corticosteroids, or pentoxifylline or their combination to those of patients not receiving them adjusted on the independent prognostic factors at baseline. The secondary endpoints will be: a) assessment of response to the assigned treatment using the Lille model; b) analysis of 6-month survival according to allocated therapy.

Lille model evaluate to therapy and will be analyzed as dichotomous outcome (<0.45 vs. ≥0.45), ordinal outcome (≤0.16 vs. 0.16-0.56 vs. ≥ 0.56) and continuous outcome (after log-transformation) and will be restricted to patients with a dosage of bilirubin at day 7 after initiation of assignment treatment.

Inclusion Criteria: patients included in 10 randomized controlled trials testing either corticosteroids alone, or pentoxifylline alone or their combination have been published. Exclusion Criteria: -