Selective use of sonographic cervical length measurement for predicting imminent preterm delivery in women with preterm labor and intact membranes.

OBJECTIVES To determine, in a population of women with preterm labor and intact membranes, whether ultrasound cervical length measurement performed only in patients selected according to the Bishop score predicts imminent preterm delivery better than does systematic cervical length measurement in the entire population. METHODS The Bishop score and sonographic cervical length were recorded prospectively in women with preterm labor between 24 and 34 completed weeks' gestation. Outcome measures were preterm delivery within 48 h and within 7 days. Predictive values were calculated for each marker separately and then in combination. RESULTS Of the study population of 395 women, 17 (4.3%) and 32 (8.1%) delivered within 48 h and within 7 days, respectively, following inclusion. For delivery within 7 days, areas under the Bishop score (0.848) and sonographic cervical length (0.813) receiver-operating characteristics curves did not differ significantly. For the selective use of sonographic cervical length measurement in patients selected according to the Bishop score, the test was considered positive if the Bishop score was >or= 8, or 4-7 with cervical length <or= 30 mm. This test was as sensitive (94%) but more specific (60% vs. 42%, P < 0.001) for predicting preterm birth within 7 days than was sonographic cervical length with a 30-mm cut-off value in the entire population. Results were similar for delivery within 48 h. CONCLUSION For predicting imminent preterm delivery in women with preterm labor, measuring sonographic cervical length only in patients with a Bishop score between 4 and 7, compared with a strategy of systematic measurement in the entire population, reduces by 30% the number of false positives and might thus decrease unnecessary therapeutic intervention.