RESUMEN
OBJECTIVE: To assess the efficacy of the cervical phosphorylated insulin-like growth factor binding protein-1 (pIGFBP-1) in combination with cervical length measurement for the prediction of preterm delivery. DESIGN: Observational prospective study. SETTING: University Hospital, Spain. SAMPLE: A total of 276 women between 24 and 34 weeks gestational age complaining of uterine contractions and intact membranes. METHODS: Transvaginal scan to assess cervical length and cervical swabs for pIGFBP-1 detection. MAIN OUTCOME MEASURES: Prevalence of preterm delivery within 48 hours and 7 days, delivery at < 32, < 34 weeks, admission-to-delivery interval by cervical length and the presence of pIGFBP-1. Results. The prevalence of preterm delivery at < 34 weeks was 14.1% (39/276) and 7.6% (21/276) at < 32 weeks. Cervical pIGFBP-1 was positive in 38.7% (107/276), whereas cervical length < 15 mm was present in 7.2% (20/276) and was > 30 mm for 64.8% (179/276) of the women. The presence of a positive pIGFBP-1 appeared useful for prediction of delivery before 32 (likelihood ratio (LR) = 2.21, confidence interval (CI): 1.63-3.00) and 34 weeks (LR = 1.76, CI: 1.25-2.41), respectively. Receiver operating characteristic curves were significant for cervical length and pIGFBP-1 for prediction of delivery at < 34 weeks, < 32 weeks, < 48 hours and < 7 days. Combining pIGFBP-1 and cervical length had a median interval that was shorter when both were positive as compared to only one positive or both negative, but the CIs overlapped. CONCLUSIONS: Both cervical length measurement and pIGFBP-1 are useful for prediction of spontaneous preterm delivery. A pIGFBP-1 positive result associated with cervical length does not improve preterm delivery prediction rates.