• J M Crane, B A Armson, L Dodds, R F Feinberg, W Kennedy, and S A Kirkland.
• Department of Obstetrics and Gynecology, Memorial University of Newfoundland, St. John's, Canada.
• Obstet Gynecol. 1999 Apr 1; 93 (4): 517-22.

ObjectiveTo determine the value of markers for predicting spontaneous preterm birth.MethodsOne hundred forty asymptomatic gravidas were recruited from 20-24 weeks' gestation. Risk score was assessed, vaginal swabs were analyzed for bacterial vaginosis, and cervical and vaginal swab were tested for fetal fibronectin FDC-6, X18A4, and CAF. Univariate analysis was used to determine potential predictors (and combinations of predictors) of outcome. Multiple logistic regression was done to identify independent predictors of spontaneous preterm birth. Sensitivity, specificity, positive and negative predictive values; and odds and likelihood ratios were calculated for significant predictors.ResultsPredictors significantly associated with the primary outcome were preterm birth-risk score and vaginal fetal fibronection FDC-6 (logistic regression odds ratio [OR] 16.9 [95% confidence interval (CI) 3.1, 92.8]) and 8.0 ([95% CI 1.6, 38.2], respectively). Bacterial vaginosis, fetal fibronectin X18A4, fibronectin CAF, and cervical fetal fibronectin FDC-6 were not associated with spontaneous preterm birth; however, the statistical power to assess these variables was limited. The combination of positive preterm birth-risk score and vaginal fetal fibronectin FDC-6 had a sensitivity of 44.4%, specificity of 97.7%, positive predictive value of 57.1%, negative predictive value of 96.2%, and a significant likelihood ratio for a positive test of 19.4 (95% CI 5.1, 73.8).ConclusionThe combination of preterm birth-risk score and vaginal fetal fibronectin FDC-6 predicted spontaneous preterm birth. Intervention trials are required to determine whether a combination of screening tests will reduce rates of spontaneous preterm birth.

21 keywords...

hide…