Hypoxic hepatitis (HH), historically termed ischemic hepatitis or “shock liver,” is an acute liver injury syndrome observed in critically ill patients. It is characterized by a rapid and marked elevation in serum aminotransferases occurring in the setting of circulatory, cardiac, or respiratory failure, after exclusion of other causes of acute hepatic injury. Clinical interest centers on how frequently HH occurs in intensive care units and whether it independently predicts worse outcomes.
The authors conducted a systematic review and meta-analysis following PRISMA guidance and registered the protocol in PROSPERO (CRD420251088120). A comprehensive search of multiple bibliographic databases (PubMed, Scopus, Web of Science, EMBASE, Cochrane, LILACS, CNKI) and Google Scholar was performed through July 4, 2025. Eligible studies included randomized trials, cohort studies, and case-control studies of critically ill adult patients reporting incidence, mortality, or relevant secondary outcomes for HH. Primary outcomes prespecified were HH incidence and all-cause mortality; secondary outcomes included hospital length of stay, use of organ support, and direct comparisons with ICU patients without HH. Random-effects meta-analyses were used and heterogeneity was quantified with I2 statistics.
Across 25 included studies comprising 150,928 critically ill patients, the pooled prevalence of hypoxic hepatitis in ICU populations was 8.3%. This pooled estimate integrates heterogeneous study settings and designs reported in the source literature.
The pooled all-cause mortality among patients with HH was reported as 64.0%. When compared with critically ill patients who did not develop HH, the diagnosis was associated with a substantially higher risk of death (relative risk [RR] 2.94). The authors emphasize that, while the association is consistent across studies, the pooled data do not by themselves establish a causal role for HH in producing the excess mortality.
Patients with HH had greater reported use of organ support. Pooled relative risks indicated higher use of mechanical ventilation (RR 2.44) and renal replacement therapy (RR 3.80) among HH patients, although these pooled associations were described as elevated but not statistically significant in the analyses presented. Secondary outcomes such as length of stay and other organ support metrics were considered but specific pooled estimates beyond organ support RRs were not detailed in the abstract.
The leading reported precipitants of HH across included studies were cardiogenic shock, cardiac arrest, and sepsis—all conditions that produce systemic hypoperfusion or severe cardiopulmonary compromise. By definition, HH occurred in the context of circulatory, cardiac, or respiratory failure after alternative hepatic causes were excluded.
The authors note important limitations in the existing literature. Heterogeneity across studies was high, and residual confounding cannot be excluded. Possible publication bias may have influenced pooled estimates. Diagnostic criteria for HH varied across reports, complicating synthesis. Because included studies were observational in large part, the evidence chiefly supports HH as an associated prognostic marker rather than a proven causal determinant of increased mortality in ICU populations.
This systematic review and meta-analysis found that HH affects a meaningful proportion of critically ill patients (pooled prevalence 8.3%) and is consistently associated with adverse outcomes, including high all-cause mortality (64.0%) and an almost threefold higher risk of death compared with non-HH ICU patients (RR 2.94). The findings suggest that HH identifies a high-risk subgroup of ICU patients who more frequently require organ support.
However, interpretation should be cautious: heterogeneity, potential confounding, and publication bias limit causal inference. The authors call for standardized diagnostic criteria for hypoxic hepatitis and prospective studies that evaluate whether HH provides incremental prognostic information beyond established ICU severity scores. Such research would help determine whether HH should be used as an independent prognostic indicator or primarily as a marker of severe underlying illness.
(Details about individual study characteristics, exact heterogeneity metrics, subgroup analyses, and formal risk-of-bias appraisals were not reported in the abstract and therefore are not summarized here.)