Background

The prognosis of critically ill patients with cirrhosis who require mechanical ventilation is guarded. Data are lacking for the optimal therapeutic approach to hepatic encephalopathy (HE) in the ventilated patient.

Methods

Retrospective cohort analysis of 314 encounters (298 patients) with cirrhosis who underwent mechanical ventilation in a medical ICU and were ordered at least 1 dose of lactulose. Hazard of extubation alive was determined using a competing risk model. Primary exposures were HE therapy (lactulose and rifaximin) which were adjusted for the indication for ventilation (HE, procedures, respiratory failure), age, MELD-Na, and compensation status.

Results

Indications for ventilation were 22.3% for grade 4 HE, 29.9% for procedures, and 47.8% for respiratory or cardiovascular failure. Median length of intubation was 2.63 days; death rate on ventilator was 31.2%. Relative to intubation for procedure, hazard of extubation for intubation for HE was 0.34 (95% confidence interval (CI): 0.22–0.52) and 0.33 (CI: 0.23–0.47) for respiratory failure. Hazard of extubation for rifaximin administration within 24-h after intubation was significant at 1.74 (1.21–2.50). Lactulose dosing was not significant for hazard of extubation.

Discussion

Mortality is high for all patients with cirrhosis requiring mechanical ventilation, including those intubated for grade 4 HE. Efforts to optimize the odds of successful extubation are urgently needed. Our findings suggest improved incidence of extubation associated with rifaximin administration in the first 24-h after intubation. Prospective, multi-center data to confirm these findings in this vulnerable population are warranted.