Accuracy of computed tomography for detecting hepatic steatosis in donors for liver transplantation: A meta-analysis.

Abstract

The accuracy of computed tomography (CT) for detecting donor hepatic steatosis (HS) before liver transplantation is not well established. A meta-analysis was performed to determine the accuracy of CT for HS detection in liver donor candidates. Pooled sensitivity, specificity, positive and negative likelihood ratios, hierarchical summary receiver operating characteristic (HSROC) curves, and the area under the curve (AUC) were estimated using HSROC and bivariate random-effects models. Twelve studies involving 1782 subjects were eligible for this meta-analysis. For detecting significant HS (>10%-30% steatosis in liver pathology) with CT in liver donors, the pooled sensitivity, specificity, positive likelihood ratio, and negative likelihood ratio were 0.81 (95% confidence interval [CI]: 0.70-0.89), 0.94 (95% CI: 0.90-0.96), 13.7 (95% CI: 8.1-23.1), and 0.20 (95% CI: 0.12-0.33). The AUC was 0.95 (95% CI: 0.92-0.96). For detecting the presence of HS, these corresponding diagnostic estimates were 0.50 (95% CI: 0.36-0.64), 0.90 (95% CI: 0.83-0.95), 5.2 (95% CI: 3.1-8.9), 0.55 (95% CI: 0.42-0.72), and 0.80 (95% CI: 0.76-0.83). Moderate-to-high heterogeneity was detected. Computed tomography shows high accuracy in detecting significant HS while poor accuracy in detecting the presence of HS in liver donors. Donors estimated to have significant HS by CT may avoid unnecessary liver biopsy.