Aim: Various predictive markers for preterm delivery have been proposed in previous studies. We investigated which marker is most reliable.

Methods: In 126 patients with preterm labor before 32 weeks of gestation and intact membranes, who had regular uterine contractions with cervical changes effaced > or =50%, we evaluated seven markers: interleukin (IL)-8, glucose, and granulocyte count in amniotic fluid (AF); fetal fibronectin (fFN) in vaginal secretions; IL-8 in cervical mucus; cervical length; and preterm labor index (PLI) based on clinical variables. The relationships of these variables to the occurrence of preterm delivery before 34 weeks were examined by logistic regression analysis.

Results: Values for AF IL-8, AF granulocyte count, fFN in vaginal secretions, and PLI were significantly higher, while the value for AF glucose was significantly lower, in patients delivering before 34 weeks than those in patients delivering at or following 34 weeks (P < 0.0001 for all). The most sensitive marker for predicting delivery before 34 weeks was AF IL-8 (sensitivity, 67.8%). The most specific markers were AF IL-8 (specificity, 95.5%) and PLI (specificity, 95.5%). By the logistic regression analysis, AF IL-8, AF glucose, fFN in vaginal secretions and PLI showed independent relationships with delivery before 34 weeks (P = 0.0009, P = 0.0032, P = 0.0131 and P = 0.0038, respectively).

Conclusions: In preterm labor with intact membranes, AF IL-8, AF glucose, fFN in vaginal secretions and PLI were highly predictive markers for detecting preterm delivery before 34 weeks.

Download full-text PDF

Source
http://dx.doi.org/10.1111/j.1447-0756.2007.00474.xDOI Listing

Publication Analysis

Top Keywords

vaginal secretions
20
preterm labor
16
preterm delivery
16
ffn vaginal
16
delivery weeks
16
predictive markers
12
intact membranes
12
il-8 glucose
12
secretions pli
12
amniotic fluid
8

Similar Publications

Want AI Summaries of new PubMed Abstracts delivered to your In-box?

Enter search terms and have AI summaries delivered each week - change queries or unsubscribe any time!