The value of the 24-h proteinuria in evaluating the severity of preeclampsia and predicting its adverse maternal outcomes.

ABSTRACT

BACKGROUND: To identify the 24-h proteinuria value with quantitative analysis and how it correlates with the severity of preeclampsia and subsequent adverse maternal outcomes in the Chinese population. STUDY DESIGN: Eleven hospitals in 10 provinces across China were chosen, in which 1,738 pregnant women complicated by hypertensive disorders of pregnancy (HDP) with the records of 24 h proteinuria were enrolled. They were allocated into four groups patients with maximal quantified proteinuria < 0.3 g/24 h (Group 1, n = 328); patients with maximal quantified proteinuria ≥ 0.3 g/24 h and < 2.0 g/24 h (Group 2, n = 638); patients with maximal quantified proteinuria ≥ 2.0 g/24 h and < 5.0 g/24 h (Group 3, n = 353); and patients with maximal quantified proteinuria ≥ 5.0 g/24 h (Group 4, n = 419). Logistic regression analysis were conducted to assess the differences in maternal outcomes between different subgroups of 24-h proteinuria and to identify independent risk factors of adverse maternal outcomes in preeclampsia. The multivariable risk prediction model of adverse maternal outcome for HDP was established with receiver operating characteristic curve (ROC) curve and its predicted value was assessed. RESULTS: Thrombocytopenia and cerebral or visual symptoms were more frequent in Groups 3 and 4 than Groups 1 and 2 but no differences were found between Groups 3 and 4 or Groups 1 and 2. Maternal complications were more frequent in Groups 3 and 4 than in Groups 1 and 2 [Group 3 vs. Group 1, odds ratios (ORs) = 3.359 (1.067-10.571); Group 4 vs. Group 1, OR = 3.628 (1.189-11.086); Group 3 vs. Group 2, OR = 2.845 (1.155-7.003); Group 4 vs. Group 2, OR = 3.082 (1.304-7.288)]. However, no significant difference was found between Groups 4 and 3 or between Groups 2 and 1. The proteinuria ≥ 2 g/24 h had an area under the receiver operating characteristic curve (AUC ROC) of 0.668 (95% confidence interval (CI) 0.632-0.705) for predicting adverse maternal outcome. After adjusting for the effects of other symptoms, signs, and laboratory tests, it was the independent risk factor and predictor factor of the adverse maternal outcome (OR = 3.683, 95% CI 2.439-5.562, P<0.001). The final risk prediction model had an AUC ROC of 0.800 (95% CI 0.769-0.830, P<0.001). CONCLUSION: The proteinuria ≥ 2 g/24 h is an independent predictive factor of adverse maternal outcomes in preeclampsia, but its individual predictive value is limited. The risk prediction model is effective in assessing the risk of adverse maternal outcomes in patients with HDP.