UC San Diego

Spontaneous preterm birth prior to 37 weeks gestation is a leading cause of neonatal morbidity and mortality. The preterm birth rate in the United States remains higher than in many other westernized countries. The etiology of preterm birth, identification of women at risk, and effective interventions for preterm birth prevention remain public health, research, and clinical challenges. Fetal fibronectin (FFN), a specific biomarker, seen in cervical and vaginal secretions with disruption of the decidual interface, predicts risk for preterm delivery. The purpose of this retrospective cohort study was to examine whether the availability of clinical FFN testing had an impact on preterm birth, NICU admissions, or use of hospital services. The study compared cohorts seen for assessment of preterm labor symptoms in the six months prior (n=372) to and after (n=390) availability of testing ; 215 tests were performed on 183 subjects. Test use increased significantly during the time of test availability, (trend, p < 0.001) however, overall test utilization rate remained below 50%. The study failed to demonstrate a reduction in preterm delivery, NICU admission, or hospital admissions. Lack of major significant study findings are likely attributable to insufficient power due to limited test utilization, prolonged implementation time, and lack of procedural policies, which reduced the potential impact of testing on patient care. The test negative predictive value (100%) for delivery within 14 days was consistent with previous findings, however, this did not lead to the reduced hospital admissions seen in previous studies. Findings suggest that implementation of FFN testing in clinical settings takes time and requires more rigorous policies for use to replicate previous research findings. Current practice combines FFN testing with cervical length and consideration of other risk factors, especially a history of preterm birth, to improve diagnostic precision. Use of predictive algorithms using the best evidence and predictive factors for subpopulations may lead to improved diagnostic precision in the future