Improvement of Clinical Outcomes in Patients with Acute Liver Failure Through Establishment of a Specialty Kidney-Liver Program

Background: Patients with acute liver failure (ALF) are at high risk of dying from cerebral edema due to hyperammonemia and require multidisciplinary care. The University of Washington (UW) is a large academic center with the largest liver transplant center in our area. To improve the care of those with liver and kidney failure, the UW Division of Nephrology launched a “Kidney-Liver-Program” (KLP) in 2016 which initiated clinical practice changes including early, continuous HD for ALF patients with hyperammonemia to improve clinical outcomes. This abstract compares outcomes in patients with ALF before and after the start of the KLP. Methods: A retrospective cohort study was performed by reviewing the charts of patients admitted with ALF to the UW with hyperammonemia (ammonia level >200mg/dL). The two cohorts in this study were those with ALF who presented before (01/2012-06/2016) and after (09/2016-03/2021) the launch of KLP. The primary outcome was all-cause mortality. Patients with ALF in the setting of cardiogenic shock or ischemia were excluded. Results: 86 patients with ALF and hyperammonemia not from ischemia or cardiogenic shock were identified between 01/2012-03/2021, 43 in the pre-KLP era and 43 in the post-KLP era. Of these, 16 (37%) pre-KLP patients received HD; 25 (58%) patients received HD in the post-KLP cohort, when HD was targeted to achieve a reduction in ammonia levels <200mg/dL (95% CI: 0%-42%; p=0.05). Mortality was numerically lower in the post-KLP era (23%) compared with the pre-KLP era (33%), although the difference was not statistically significant (95% CI: -28%-10%; p=0.33). Conclusion: The launch of a specialized KLP with the aim of initiating HD early to target an ammonia level of <200mg/dL significantly increased receipt of HD in patients with ALF. Although not statistically significant, mortality was numerically lower in the post-KLP cohort.Table 1. ResultsFigure 1. Patient population