Objective: To reassess the diagnostic value of 24 hour urinary copper excretion in children with Wilson disease (WD).
Methods: From July 2005 to June 2007, inpatients over three years old in a pediatric liver center were assigned into WD and non-WD group.
Results: 94 patients, including 26 cases in WD and 68 in non-WD group, were enrolled in this study. The median of 24 h urinary copper excretion was 98.5 microg in WD group and 25.8 microg in the non-WD group (Z = -6.111, P equal to 0.000). The area under receiver operator curve (ROC) was 0.909 (95% CI: 0.839-0.979, P equal to 0.000). The sensitivity, specificity, accuracy, positive predictive value and negative predictive value were 84.6%, 91.2%, 89.4%, 78.6% and 93.9% respectively using 52.0 ug as a cutoff value, and 50.0%, 97.1%, 84.0%, 86.7% and 83.5% using 100 microg as a cutoff value. The goodness of fitness of 52 microg criteria was significantly higher than 100 microg criteria (kappacoefficient 0.760, 0.541 respectively, P equal to 0.000).
Conclusion: Comparing to 100, 52 microg of 24 h urinary copper excretion as a cutoff value significantly improves the sensitivity and accuracy for diagnosing WD in children.