Background Multiple pregnancies are associated with a higher risk of spontaneous preterm birth (sPTB). Whilst fetal fibronectin (fFN) and cervical length (CL) measurement can predict sPTB in singleton pregnancies (Kurtzman et al, 2009), their value for twin pregnancies is unknown.

Methods Prospective blinded secondary analysis of longitudinal samples of cervicovaginal fluid fFN concentration (nanograms per mililiter) using a bedside 10 qfFN analyzer (HOLOGIC, USA), and transvaginal ultrasound CL of 93 consecutive women with multiple pregnancies attending a Preterm Surveillance Clinic at St. Thomas Hospital from 18 weeks gestation (Oct 2010–Jan 2012). qfFN was assigned 4 ranges; <10, 10–50, 50–200, >200 (ng/ml) to detect spontaneous delivery before 30, 34 and 37 weeks. qfFN was blinded to clinicians using an embedded code in the analyzer.

Results The rate of sPTB (<37 weeks) rose increased with increasing qfFN from 17.5% (<10 ng/ml) to 61.5% (>200 ng/ml) and the negative prediction value for sPTB <30 weeks at <10 ng/ml was 98%. 4/13 (30%) of women with qfFN > 200 ng/ml delivered <30 weeks gestation. Using combined CL/qfFN testing, the positive prediction value of a qfFN value >200 ng/ml and CL < 25 mm was 87.5% for SPTB <37 weeks.

Conclusion This is the first report of 10 qfFN in twins, demonstrating that it adds predictive value to the qualitative results (negative cut-off at 50 ng/ml). High levels, even in early pregnancy, are associated with preterm delivery. Using cervical length and qfFN, management can be targeted to this group; e.g. antenatal maternal steroids. Further research should evaluate interventions to prolong pregnancy in this highest risk group, while lower risk women can be reassured.