Introduction Hospitalised patients with cirrhosis often require admission to intensive care (ITU), usually for management of secondary complications such as GI bleeding, sepsis, or organ failure. Prognosis of these patients is poor. Despite demonstration of modest improvements in recent years, overall in-hospital mortality in this group remained 54.6% in a recent study.¹ Our study aimed to explore the characteristics and outcomes of cirrhotic patients admitted to ITU, and whether this was comparable to published data. We also aimed to explore whether parameters such as APACHE 2 score and admission blood lactate differ between those who survived their ITU stay and those that did not, and if this may help predict discharge.

Methods A retrospective analysis was performed of patients admitted to the Whittington hospital ITU from January 2011-June 2013. Information regarding patients with a diagnosis of cirrhosis was gathered from the Intensive Care National Audit and Research unit (ICNARC) database, and discharge summaries.

Results We identified 60 patients with cirrhosis, 3.07% of total ITU admissions, mean age 54.8 years (range 19–78). 49/60 patients had alcoholic cirrhosis, 5/60 had non-alcoholic steatohepatitis related cirrhosis. Primary reason for admission included GI bleed (24/60, 40%), pneumonia (16/60, 26.7%), other sepsis (10/60, 16.7%), encephalopathy (8/60, 13.3%). Overall mortality figures were 41.7% in-ITU, 48.3% at 30 days, and 70% at 1 year. In-ITU mortality for patients requiring only ventilatory support was 48.9%, those requiring inotropic support was 61.2% and those requiring renal support 64.2%. 24/60 patients required all 3 methods of support, 66.7% of those died in ITU and 75% at 30 days. 23 patients were admitted with sepsis and decompensated liver disease, 65% died in ITU and 73.9% at 30 days. Mean APACHE 2 scores for patients that died in ITU vs. those discharged from ITU were 23 (range 9–30) and 20 (range 8–36) respectively, with a statistically significant difference between the two groups (p = 0.036). Mean admission serum lactate for patients who died in ITU vs. those discharged from ITU was 7.6 (range 1–23) and 4.6 (range 1–17) respectively, demonstrating a statistically significant difference (p = 0.015).

Conclusion As expected for a non liver-specialist unit, most patients had alcohol related cirrhosis. Mortality was high but comparable to other published data. The worst outcomes were seen in patients with sepsis and decompensated liver disease, and those requiring organ support. Admission lactate levels and APACHE 2 scores were significantly lower in patients successfully discharged from ITU; admission lactate could potentially aid prediction of successful discharge. Further study is needed.

Reference

1. Intensive Care Med. Jun;2012;38(6):991–1000

Disclosure of Interest None Declared.