Evaluation of risk scoring in a preterm birth prevention study of indigent patients.

A total of 4591 indigent patients were scored for risk of preterm labor and birth by the Creasy et al. scoring system and delivered at greater than or equal to 20 weeks' gestation. Of 462 preterm deliveries (less than 37 weeks), only 39.6% were contributed by the 18.1% of patients who scored at high risk (score greater than or equal to 10). Of the 44 risk factors in the scoring system, 25 were not statistically significant by univariate chi 2 analysis. The spontaneous preterm birth rate was 8.0%. The positive predictive value of the Creasy et al. system was 18.3%. With multiple logistic analyses we excluded infrequent risk factors and identified five factors that remained significant (prepregnancy weight less than 45.5 kg, black race, single marital status, one preterm labor and delivery, preterm labor, and delivery greater than or equal to 2). With these five risk factors we created an equation that allowed identification of 12% of patients at high risk of preterm delivery with a positive predictive value of 21.9%. Because of the importance of previous preterm delivery, risk scoring of primigravid patients is of limited value.