Evaluation of quantitative fFn test in predicting the risk of preterm birth.

ABSTRACT

OBJECTIVE: To evaluate diagnostic accuracy of quantitative fetal fibronectin (qfFN) test in predicting preterm birth (PTB) risk <34 weeks' gestation or within 14 days from testing. We explored the predictive potential of the test in five-predefined PTB risk categories based on predefined qfFN thresholds (<10, 10-49, 50-199, 200-499 and ≥500 ng/mL). METHODS: Measurement of cervicovaginal qfFN with Rapid fFN 10Q System (Hologic) in 126 women with singleton pregnancy (23-33 weeks' gestation) reporting signs and symptoms indicative of preterm labour (PTL). RESULTS: For PTB prediction risk <34 weeks' gestation, sensitivity decreased from 100% to 41.7% and specificity increased from 0% to 99.1% with increasing fFN thresholds. Positive predictive value (PPV) increased from 9.5% to 83.3% with increasing qfFN thresholds, while negative predictive value (NPV) was higher than 90% among the fFN-predefined categories. Diagnostic accuracy results showed an area under a receiving operator characteristic (ROC) curve of 84.5% (95% CI, 0.770-0.903). For delivery prediction within 14 days from the testing, sensitivity decreased from 100% to 42.8% and specificity increased from 0% to 100% with increasing fFN thresholds. Diagnostic accuracy determined by the ROC curve was 66.1% (95% CI, 0.330-0.902). CONCLUSIONS: The QfFN thresholds of tests are a useful tool to distinguish pregnant women for PTB prediction risk <34 weeks' gestation.