Neurological events after liver transplantation: a single-center experience.

The aim of this study was to identify potential risk factors linked to neurologic events (NE) occurring after liver transplantation (LT) and use them to construct a model to predict such events. From odds ratios (OR) of risk factors, a scoring system was assessed using multivariate regression analysis. Forty-one of 307 LT patients presented NE (13.3%), with prolonged hospital stay and decreased post-LT survival. On multivariate analysis, factors associated with NE included: severe pre-LT ascites OR 3.9 (1.80-8.41; P = 0.001), delta sodium ≥12 mEq/l OR 3.5 (1.36-8.67; P = 0.01), and post-LT hypomagnesemia OR 2.9 (1.37-5.98; P = 0.005). Points were assigned depending on ORs as follows: ascites 4 points, and hypomagnesemia and delta sodium ≥12 mEq/l, 3 points each (score range = 0-10 points). ROC curve analysis suggested good discriminative power for the model, with a c-statistic of 0.72 (CI 0.62-0.81; P < 0.0001), best performance for a cutoff value >3 points (71% sensitivity, 60% specificity). NE risk increased progressively from 6.4%, to 10.3%, 12.8%, 31.5% and 71.0% as scores rose from 0 to 3, 4, 6-7 and 10 cumulative points, respectively. The score described helps to identify patients potentially at risk for neurologic events, and its prevention would decrease morbidity and mortality after LT.