Introduction Fetal fibronectin (fFN) is an excellent predictor of spontaneous preterm birth (sPTB) and is used qualitatively (<50 ng/ml negative threshold) from 22 weeks gestation. Its value at earlier gestations using a quantitative test (qfFN) is unknown.

Methods A prospective secondary analysis of 431 asymptomatic women at high risk of sPTB, who underwent qfFN testing at 18–21 + 6 weeks. 327 women underwent later testing at 22–26 + 6 weeks (acting as their own controls). The end-points were sPTB/preterm premature rupture of membranes and delivery before 30, 34 and 37 weeks gestation and within 8 weeks of testing.

Results Early qfFN predicted delivery within 8 weeks of testing, <30, <34 and <37 weeks with receiver operating characteristics (ROC) areas of 0.66 (0.54–0.80, p < 0.05), 0.68 (0.56–0.79, p < 0.01), 0.68 (0.58–0.78 p < 0.001) and 0.64 (0.57–0.72, p < 0.001). 22-week test prediction was ROC areas of 0.77 (0.63–0.91 p < 0.001), 0.78 (0.61–0.95, p < 0.001) and 0.79 (0.70–0.89, p < 0.001) respectively. A qFFN result of <10 ng/ml at earlier gestations had only 1%, 2% and 4.3% of women delivered within 8 weeks, <30 and <34 weeks gestation, rising to 6.7%, 8.1% and 14.1% with values between 10–49.9 ng/ml (all differences statistically significant, p = 0.03, 0.02, 0.004 by Fishers-Exact). The 22-week test had 1.0%, 1.0% and 2% respectively, rising to 6.7%, 3.3% and 14% with values between 10–49.9 ng/ml (p = 0.02, 0.23, 0.004).

Conclusion qfFN is valid for screening for sPTB at 18 weeks, but has inferior predictive value to 22 weeks. Early identification may enable earlier targeted management. A threshold of <10 ng/ml is more appropriate than current practise to define low risk at 18 weeks.