ATLANTA – The combination of cervical length and self-collected fetal fibronectin did not accurately predict preterm birth in a cohort of nulliparous women.
Although each measurement in itself had some predictive value, the best measure – cervical length measured by transvaginal ultrasound at 22-30 weeks’ gestation – was only marginally clinically useful, identifying about 24% of those who had a preterm birth. The addition of fetal fibronectin added nothing to the predictive value of the cervical measurement, Dr. Michael Esplin reported at the annual Pregnancy Meeting sponsored by the Society for Maternal-Fetal Medicine.
“Based on this study, we must conclude that the routine use of self-collected quantitative fetal fibronectin and transvaginal cervical length is not a viable way to screen for preterm birth in a nulliparous population,” said Dr. Esplin of the University of Utah, Salt Lake City.
©Photodisc/Thinkstockphotos.com
He reported the results of Nulliparous Pregnancy Outcomes Study: Monitoring Mothers-to-be (nuMoM2b). The prospective studycomprised more than 9,300 women and was conducted at eight sites in the United States. There were two parts to the investigational screening protocol: collection of a fetal fibronectin swab at three gestational time points (8-14 weeks, 16-22 weeks, and 22-30 weeks) and transvaginal ultrasound to measure cervical length at time of the second and third fibronectin measurement.
The main outcome measure was a live birth at less than 37 weeks, after spontaneous preterm labor; the secondary outcome was a live birth after spontaneous preterm labor at less than 32 weeks’ gestation.
Mean age of the women was 27 years. All the women were carrying a singleton pregnancy. There were 471 spontaneous preterm live births. These were more likely to occur in the youngest and oldest women in the cohort. Preterm births were more common in women who smoked during pregnancy (24% vs. 17%), those who had less than a high school education (11.5% vs. 7.8%), and those with a prepregnancy history of diabetes (3.6% vs. 1.5%).
At the second visit, cervical length of 25 mm or less identified 8% of those who went on to have a preterm birth, for a positive predictive value of 16% and an area under the curve (AUC) of 0.61. At the third visit, cervical length of 25 mm or less identified 24% of women with preterm birth, for a positive predictive value of 15% and an area under the curve of 0.67.
Fetal fibronectin alone performed similarly. No matter what visit, and whether the cutoff was 10, 50, or 200 ng/mL, the AUC stayed between 0.51-0.54. The positive predictive values of these measurements ranged from 6%-12%.
“All of these measures had a relatively low positive predictive value,” Dr. Esplin said. “And while that may be acceptable if the tests are very sensitive, these, in fact were not. Cervical length, for example, had only 7.9% sensitivity at visit 2 and 23% sensitivity at visit 3.”
Combining cervical length and fetal fibronectin did nothing to improve the test’s validity or clinical usefulness. For a test to be clinically useful, Dr. Esplin said, the AUC should be at least 0.73. None of the measures came close to meeting this standard.
The National Institute of Child Health and Human Development sponsored the study. Dr. Esplin reported having no financial disclosures.