Accuracy of C-reactive protein determination in predicting chorioamnionitis and neonatal infection in pregnant women with premature rupture of membranes: a systematic review.

Preterm premature rupture of the fetal membranes (PPROM) is associated with intra-uterine infection. Early detection of intra-uterine infection may help prevent neonatal sepsis. C-reactive protein (CRP) is an acute phase protein often elevated when inflammation is present. The aim of this review was to assess whether CRP accurately predicts chorioamnionitis and/or neonatal sepsis in women with PPROM. We searched Medline and Embase databases for articles reporting on CRP and chorioamnionitis and/or neonatal sepsis. Two reviewers extracted clinical and methodological study characteristics and test accuracy data. Accurate data were used to form 2 x 2 data tables comparing CRP and the occurrence of infection. For the selected studies, sensitivity and specificity of CRP in the prediction of histological chorioamnionitis, clinical chorioamnionitis and neonatal sepsis were calculated separately. A bivariate meta-regression model was used to calculate pooled estimates of sensitivity and specificity. The search revealed 200 articles, of which only five met the inclusion criteria. These five articles reported on 381 patients, of which four articles (227 patients) reported on CRP as a predictor for histological chorioamnionitis and four studies (330 patients) reported on CRP as a predictor for clinical chorioamnionitis. None of the selected articles fulfilled our criteria for the use of CRP as a predictor of neonatal sepsis. CRP was moderately predictive of histological chorioamnionitis. Unfortunately, the studies of clinical chorioamnionitis were too heterogeneous to pool data. Current literature does not support the use of CRP in women with PPROM.