PURPOSE: To investigate the usefulness of the measurement of cervical length and of the test for phosphorylated insulin-like growth factor binding protein-1 (phIGFBP-1) performed sequentially in the prediction of preterm birth and the correlation between tests. METHODS: We analyzed data from 101 asymptomatic pregnant women with a history of premature delivery. The ultrasound measurement of cervical length and phIGFBP-1 test were performed in parallel every three weeks, between 24 and 34 week. The best cutoff value for each cervical evaluation was established by the ROC curve, and the two tests were compared using nonparametric tests. We determined the sensitivity, specificity and predictive values of each test and of the association of the exams for the occurrence of delivery before the 37th weeks. RESULTS: There were 25 preterm births (24.8%). The cervix length showed the highest sensitivity and was able to predict preterm birth in all evaluations, with similar accuracy at different gestational ages. The test for phIGFBP-1 was not helpful at 24 weeks, but was able to predict prematurity when performed at 27, 30 and 33 weeks. The combination of tests increased the sensitivity (81.8%) and negative predictive value (93.7%) when compared to the separate use of each test. The mean cervical length was lower in women with a positive test. CONCLUSIONS: Both cervical length and the test for phIGFBP-1 were able to predict premature delivery, and sequential combination of both tests showed a high sensitivity and high negative predictive value.
Premature birth; Risk assessment; Cervix; Cervical length measurement; Insulin-like growth factor binding protein-1; Biological markers
