Infection of the amniotic cavity is an important driver and/or consequence of preterm prelabour rupture of membranes (PPROM). Prediction of infection is challenging, limiting guidance for interventions during the antenatal period. Infection typically triggers a host inflammatory response, and non-invasive indirect markers of the maternal or fetal inflammatory response have been reported in the context of PPROM and intra-amniotic infection. Some of these markers have also been tested in amniotic fluid (AF) samples. This study compared markers of the inflammatory response in women with PPROM against the outcome standard of histological chorioamnionitis (HCA) or funisitis (FUS). Searches were conducted for studies reporting diagnostic test sensitivity and specificity for proven HCA or FUS in pregnant women with PPROM after 20 weeks' gestation. Weighted mean pooled sensitivity (Se), specificity (Sp), positive predictive value, negative predictive value, diagnostic odds ratio and 95% confidence intervals were calculated for each of the selected diagnostic tests. Except ultrasonographic detection of fetal thymic involution, almost all index tests analysed showed relatively low sensitivity. Maternal white cell count, interleukin-6 (IL-6) and AF IL-6 had credible specificity. Testing of AF markers, while more consistent than serum markers, showed no clear diagnostic accuracy improvement. There is a clear lack of evidence for the reliability of any individual diagnostic test to assist in the detection of HCA or FUS in women with PPROM. Combining several markers into a predictive model for improved diagnosis may be worth investigating.
Read full abstract