Prognostic factors and scoring systems associated with outcome in pediatric acute liver failure

Pediatric acute liver failure (PALF) is an uncommon, devastating illness with significant mortality. Liver transplantation remains the mainstay of treatment for irreversible PALF. The purpose of this study was to determine the etiology and prognostic factors associated with outcome of PALF in South Africa and to evaluate prognostic scoring systems used.

Records of 45 pediatric patients younger than 16 years of age who presented with PALF from 1 January 2015 till 31 October 2020 were analysed. Patients were divided into two groups with one group consisting of patients with spontaneous recovery of the liver with supportive treatment (6/45:13.3%) and the second group consisting of patients with poor outcomes who demised (19/45: 42%) or underwent liver transplantation (20/45: 44%).

The median age of presentation was 3.3 years (IQR 1.8-6.9) with the 1-5 years age group constituting majority of patients (55.6%). Median time to follow up was 6.1 months (IQR 0.2-28.8). Higher liver injury unit scores were observed in patients who had poorer outcomes (P = 0.008) with a threshold of greater than 246 having a sensitivity of 84% and specificity of 83% (P < 0.001). Higher peak PELD/MELD (P = 0.006) and admission UKELD (P = 0.002) scores, were found in patients with poorer outcomes. Kings College Hospital criteria (KCHC) was useful in predicting which patients would die without liver transplantation (P = 0.002). Liver transplantation was performed in 20/45 (44%) patients with a post transplantation 1 year patient and graft survival of 80%.

Although, survival of PALF patients was lower than high and other low-middle income countries, outcomes post transplantation were good. Our study demonstrates the utility of dynamic scoring systems in PALF patients, it underscores the need for early referral and clinical monitoring in a tertiary center once the criteria for PALF have been met.