Accuracy of serum uric acid as a predictive test for maternal complications in pre-eclampsia: bivariate meta-analysis and decision analysis.

The aim of this study is to determine the accuracy and clinical value of serum uric acid in predicting maternal complications in women with pre-eclampsia. An existing meta-analysis on the subject was updated. The accuracy of serum uric acid for the prediction of maternal complications was assessed with a bivariate model estimating a summary Receiver Operating Characteristic (sROC) curve. Subsequently, a clinical decision analysis was performed, in which three alternative strategies were modelled: (I) expectant management with monitoring until spontaneous labour; (II) induction of labour; (III) serum uric acid as test for predicting maternal complications. In the latter strategy, accuracy data of serum uric acid derived from the sROC curve were used to assess the value of serum uric acid in the management of women with pre-eclampsia. In this strategy, women with an increased serum uric acid were supposed to have labour induced, whereas women with serum uric acid levels below the threshold were managed expectantly. The decision whether to use the policy expectant management, to induce labour or to test serum uric acid levels, is based on the expected utility of each strategy. The expected utility depends on the probability of occurrence of severe maternal complications (i.e. severe hypertension, haemolysis, elevated liver enzymes and low platelet count (HELLP syndrome) or eclampsia) and the mode of delivery (caesarean section versus vaginal delivery). Valuation of the outcomes was performed using a distress ratio, which expresses how much worse a complication of pre-eclampsia is valued as compared to a caesarean section. Eight studies, testing 1565 women with pre-eclampsia, met the inclusion criteria. If the distress ratio was 10, the strategy regarding serum uric acid would be the preferred strategy when the probability of complications was between 2.9 and 6.3%. At higher complication rates induction of labour would be preferred, whereas at lower complication rates expectant management would be the best treatment option. These findings were stable in sensitivity analyses, using different distress ratios. Based on the decision analysis, serum uric acid seems to be a useful test in the management of pre-eclampsia under realistic assumptions.