Outcome predictors of cirrhosis patients admitted to the intensive care unit

OBJECTIVE: To evaluate outcome predictors of patients with cirrhosis admitted to an intensive care unit (ICU). METHODS: One hundred and twenty-nine consecutive patients with cirrhosis admitted to the ICU at a tertiary care transplant centre in Saudi Arabia between March 1999 and December 2000 were entered prospectively in an ICU database. Liver transplantation patients and readmissions to the ICU were excluded. The following data were documented: demographic features, severity of illness measures, parameters of organ failure, presence of gastrointestinal bleeding, and sepsis. The need for mechanical ventilation, renal replacement therapy and pulmonary artery catheter placement was recorded. The primary endpoint was hospital outcome. RESULTS: Cirrhotic patients admitted to the ICU had high hospital mortality (73.6%). However, the actual mortality was not significantly different from the predicted mortality using prediction systems. There was an association between the number of organs failing and mortality. Coma and acute renal failure emerged as independent predictors of mortality. All patients who were monitored with pulmonary artery catheterisation in this study died. Patients requiring mechanical ventilation and renal replacement therapy had very high mortalities (84% and 89%, respectively). All 13 cirrhotic patients admitted to ICU immediately post-cardiac arrest in this study died. CONCLUSIONS: Cirrhotic patients admitted to ICU have a poor prognosis, especially when admitted with coma, acute renal failure or post-cardiac arrest. The consistently poor prognosis associated with certain ICU interventions should raise new awareness regarding limitations of medical therapy. These mortality statistics compel a critical re-examination of uniformly aggressive life support for the critically ill cirrhotic patient, a percentage of whom will not benefit from invasive measures.