Systematic review with meta-analysis: high mortality in patients with non-severe alcoholic hepatitis

Alcoholic hepatitis is a serious complication of alcohol misuse. Severe alcoholic hepatitis with its high mortality, has been investigated in detail but 'nonsevere alcoholic hepatitis' is poorly characterised. Survival of this group of patients is unknown.

To conduct a systematic review and meta-analysis to determine 28-day, 90-day and 1-year mortality of patients with nonsevere alcoholic hepatitis.

The protocol was registered on the PROSPERO database (CRD42018107451). Embase, Medline and Cochrane Central databases were searched until July 2018. All study designs reporting mortality rates in patients with nonsevere alcoholic hepatitis were eligible. Mortality data were extracted and meta-analysis performed using a random effects model. Risk of bias was assessed by Cochrane risk of bias or National Institutes of Health quality assessment tool for case series studies.

Twenty-five studies (n = 1372 patients; 12 prospective) met criteria. Nonsevere was variably defined based on bilirubin, prothrombin time, and creatinine. Twenty-eight day mortality (17 studies; n = 993) was 6% (95% CI 3%-9%; I²  = 67.3%; P < 0.001), 90-day mortality (15 studies; n = 755) was 7% (4%-11%, I²  = 64.2%; P < 0.001) and 1-year mortality (five studies; n = 234) was 13% (4%-24%; I²  = 72%; P = 0.006). Subgroup analyses by method of diagnosis (histological vs clinical) or study design (prospective vs retrospective) did not reveal differences in mortality.

Nonsevere alcoholic hepatitis is not benign with 6% and 13% 28-day and 1-year mortality, respectively. This systematic review demonstrates the paucity of high quality studies in patients with nonsevere alcoholic hepatitis. Our analysis suggests that patients who do not meet criteria for severe alcoholic hepatitis are an important and hitherto overlooked clinical group. Full characterisation of clinical outcome and development of treatment strategies to reduce mortality in this group is a priority.