Occurrence, Characteristics, and Outcome of Hypoxic Liver Injury among Patients Aged ≥90 Years Admitted to the Intensive Care Unit: A Retrospective Cohort Study

Introduction: Hypoxic liver injury (HLI) is a frequent and life-threatening complication occurring in up to 10% of critically ill patients. Heart failure and age were previously identified as risk factors for occurrence of HLI. However, there is lack of data on incidence of HLI and its clinical implications on outcome in very old (>= 90 years) patients. The aim of this study was to investigate occurrence, clinical characteristics, and outcome of HLI in critically ill patients >= 90 years. Methods: This is a retrospective analysis of all consecutive critically ill patients >= 90 years admitted to the intensive care unit (ICU) of a tertiary care university hospital in Hamburg, Germany. Clinical course and laboratory data were analyzed for all patients. HLI was defined according to established criteria as elevation of aminotransferase levels (>20-fold upper limit of normal). Predictors of HLI occurrence, clinical course, and outcome were assessed and compared to those of patients without HLI. Results: In total, 1,065 critically ill patients >= 90 years were included. During the ICU stay, 3% (n = 35) developed HLI. Main causes of HLI were cardiogenic shock (51%, n = 18), septic shock (23%, n = 8), and cardiac arrest (20%, n = 7). Presenting characteristics including age, gender, and BMI were comparable between patients with and without HLI. The admission cause was primary medical (HLI: 49% vs. No-HLI: 34%, p = 0.07), surgical - planned (9% vs. 38%, p < 0.001), and surgical - emergency (43% vs. 28%, p = 0.06). The median Charlson Comorbidity Index (CCI) and the median updated CCI were 2 (1-3) and 2 (1-2) points in patients with HLI and 1 (0-2) and 1 (0-2) in patients without HLI (p < 0.01 and p = 0.08). Patients with HLI presented with higher SAPS II (55 vs. 36 points p < 0.001) score on admission and required mechanical ventilation (66% vs. 34%, p < 0.001), vasopressor therapy (91% vs. 40%, p < 0.001), renal replacement therapy (20% vs. 2%, p < 0.001), and parenteral nutrition (29% vs. 7%, p < 0.001). The ICU mortality and hospital mortality in patients with HLI were 66% (n = 23) and 83% (n = 29) compared with 17% (n = 170) and 28% (n = 292) in patients without HLI, respectively (both p < 0.001). Regression analysis identified SAPS II (OR 1.05, 95% CI: [1.02-1.07]; p < 0.001) and vasopressor therapy (OR 9.21, 95% CI: [2.58-32.86]; p < 0.01) as factors significantly associated with new onset of HLI. Occurrence of HLI was independently associated with mortality (HR 2.23, 95% CI: [1.50-3.30]; p < 0.001). Conclusion: HLI is an uncommon but not rare condition in critically ill patients aged >= 90 years. Occurrence of HLI is associated with high mortality and is mainly caused by cardiogenic or septic shock. HLI may serve as early prognostic marker in critically ill patients aged >= 90 years.