Introduction: Common indications for admission to a liver intensive care unit (LICU) are often due to complications of an underlying chronic liver disease or to an acute insult to the liver either by bacterial or viral infections, alcohol or drugs. Secondary or acquired liver injury occurs in patients with systemic disease in patients who are admitted to a non-LICU setting. These patients are often hemodynamically unstable with multiorgan failure. The aim of the study was to understand the pattern of liver injury in patients admitted to a non-LICU for issues unrelated directly to the liver. Materials and Methods: Patients admitted to medical, neuro and cardiac ICU of the institution were included. All patients had liver biochemistry at admission. R factor (RF) criteria (alanine aminotransferase: S. Alkaline phosphatase ratio) was used for stratifying the liver injury as hepatocellular (HC) (RF-HC; ≥5), cholestatic (RF-C; ≤2) and mixed type (RF-HC+C; 2–5). Patients were followed up in the ICU until discharge. Statistical Analysis: SPSS v26 (IBM Corp.) was used for analysis. Chi-square test was done for categorical variables and Fisher’s exact test, when the expected frequency in the contingency table was <5 in more than 25% of the cells. P < 0.05 was considered as statistically significant. Results: 108 (43.9%) of 246 admissions had liver injury (43.9%). The elevation in liver enzymes were <2 times the upper limit of normal range. There were no significant differences in enzyme elevations in the 3 ICUs. The most common type of liver injury was RF-C (67; 62%) followed by RF-HC+C (34; 31.5%) and RF-HC (7; 6.48%). There was no difference in the disease specific liver injury. Patients with RF-C needed significantly more inotrope support (P = 0.001); 76% survived (P = 0.023). Those with RF-HC required hemodialysis (P = 0.001) with poor survival outcome (42.9%). Conclusions: Liver enzymes are not significantly elevated in non-LICU patients. RF criteria is useful in not only predicting the type of liver injury but also predicting the likely outcome.
Read full abstract