A mathematical model for predicting outcome in preterm labour

Abstract

Objective: This study aimed to develop a model for predicting the outcome and evaluating the treatment of patients with threatened of preterm labour.

Methods: Clinical data from 236 patients at <32 weeks gestation who were in preterm labour were analysed to develop a discriminant function using multiple logistic regression to identify significant risk factors. The function was validated retrospectively in a further 501 patients and prospectively in 63 patients with premature labour.

Results: Factors that increased the risk of preterm birth were premature rupture of the membranes, intrauterine infection, dilatation of the cervix and uterine bleeding. Factors that decreased the risk of preterm birth were hospital admission after 28 weeks of gestation and intravenous administration of ritodrine. The predictive accuracy of the function was 75.4% in the 236 patients analysed, 84.8% in the further 501 retrospectively studied patients and 85.7% in the prospective group.

Conclusions: The discriminant function described was clinically useful for predicting the outcome of threatened preterm labour before initiating treatment and for determining the medical care of patients, including maternal transfer to a high-level perinatal care centre.