Prediction of Spontaneous Preterm Delivery in Women With Preterm Labor

Abstract

To analyze whether specific proteins in amniotic and cervical fluids, alone or in combination with risk factors, can identify women in preterm labor with intact membranes who will deliver spontaneously within 7 days of sampling. In a cohort of 89 women in preterm labor, amniotic and cervical fluids were collected between 22 and 33 weeks of gestation. Twenty-seven proteins were analyzed simultaneously using multiplex technology. Individual levels of each protein were compared and calculations performed to find associations among different proteins, background variables, and spontaneous preterm delivery within 7 days of sampling. The area under the curve (AUC) was calculated using receiver operating characteristic curve analysis, and prediction models were created based on stepwise logistic regression. We found two multivariable models that predicted spontaneous preterm delivery better than any single variable. One combined multivariable prediction model was based on amniotic macrophage inflammatory protein-1beta, cervical interferon-gamma, and monocyte chemotactic protein-1. This model predicted outcome with 91% sensitivity, 84% specificity, 78% positive predictive value, and 94% negative predictive value, with a likelihood ratio of 5.6 and AUC of 0.91. An alternative, noninvasive model based on cervical length, cervical interferon-gamma, interleukin-6, and monocyte chemotactic protein-1 predicted delivery within 7 days with 85% sensitivity, 82% specificity, 74% positive predictive value, and 90% negative predictive value, with a likelihood ratio of 4.7 and AUC of 0.91. A combination of proteins from amniotic fluid and cervical fluid or cervical length can help determine which women will deliver preterm. II.