Safety and efficacy of antibiotics among acutely decompensated cirrhosis patients

Background and Aim: Infection is a leading precipitant of acute-on-chronic liver failure. This study aims to determine the safety and efficacy of antibiotics within acute-on-chronic liver failure. Methods: Retrospective study of 457 acute-on-chronic liver failure patients admitted to the University of Arizona Health Network between January 1 and December 31, 2014. Eligibility criteria were as follows: at least 18 years of age and 6 months follow-up, data available to calculate systemic inflammatory response syndrome (SIRS), and acute-on-chronic liver failure. This study collected patient's clinical features and historical data. Key data points were infection, antibiotic use, and SIRS. This study used Cox proportional hazards to model the effects of clinical factors on risk of death. Results: A total of 521 of 1243 met the inclusion criteria, and 64 had missing data, leaving 457 patients. Infection resulted in higher hazard (hazard ratio [HR] = 1.6, confidence interval [CI]: 1.1–1.3, P = 0.01). Patients with infections and antibiotics, compared with non-infected patients without antibiotics, had higher hazard (HR = 1.633, CI: 1.022–2.609, P =.04). Of those infected patients with antibiotics, SIRS patients experienced higher hazard (HR = 1.9, CI: 1.1–3.0, P =.007). Multivariable Cox proportional hazards associated the following with higher hazard: SIRS (HR = 1.866, CI: 1.242–2.804, P = 0.003), vancomycin (HR = 1.640, CI: 1.119–2.405, P = 0.011), Model for End-Stage Liver Disease (HR = 1.051, CI: 1.030–1.073, P < 0.001), gastrointestinal bleeding (HR = 1.727, CI: 1.180–2.527, P = 0.005), and hepatic encephalopathy (HR = 1.807, CI: 1.247–2.618, P = 0.002). Conclusion: Overall, treatment of infection with antibiotics did not improve survival; however, patients not meeting SIRS criteria had better outcomes, and vancomycin was associated with poorer survival among acute-on-chronic liver failure patients.