King's College London

Research portal

The effect of blood staining on cervicovaginal quantitative fetal fibronectin concentration and prediction of spontaneous preterm birth

Research output: Contribution to journalArticlepeer-review

Original languageEnglish
Pages (from-to)103-108
Number of pages6
JournalEuropean Journal of Obstetrics Gynecology and Reproductive Biology
Early online date28 Nov 2016
Accepted/In press27 Nov 2016
E-pub ahead of print28 Nov 2016
Published1 Jan 2017

King's Authors


OBJECTIVE: Spontaneous preterm birth is the leading cause of neonatal morbidity and mortality. Cervicovaginal fetal fibronectin (fFN) has enhanced prediction of preterm birth and, more recently, quantified results have become available so that management can planned more effectively and targeted to individual women. Manufacture guidelines stipulate that fetal fibronectin (fFN) samples should be discarded in the presence of moderate to heavy vaginal bleeding but there hasn't yet been any formal investigation into the effect of blood staining on fetal fibronectin concentration and subsequent preterm birth prediction. The objective for this study was to determine the impact of blood stained swabs on quantitative fetal fibronectin (qfFN) concentration and prediction of spontaneous preterm birth (sPTB) in asymptomatic high-risk women.

STUDY DESIGN: Predefined blinded sub-analysis of a larger prospective study of qfFN in asymptomatic women at high-risk of preterm labour. Women with and without blood stained swabs were matched for gestational age at testing and delivery, risk factors and cervical length measurement.

RESULTS: Median fFN concentration in blood stained swabs (n=58) was 66ng/ml vs. 7.5ng/ml in the controls (n=58) (p<0.0001). At ≥50ng/ml threshold the false positive ratio (FPR) in blood stained was 25/33 (75.8%) vs. 8/15 (53%) in controls, (risk difference 22.4; -6.8 to 51.6, p=0.18). At ≥50ng/ml threshold the false-negative ratio (FNR) in blood stained was 2/25 (8.0%) vs. 1/43 (2.3%) in controls (risk difference -5.7; -17.2 to 5.9, p=0.55). At each threshold 10, 50 and 200ng/ml blood stained swabs had higher sensitivity but lower specificity for predicting preterm birth. Receiver Operating Characteristic (ROC) curve, the strongest global measure of test performance, for prediction of delivery at <34 weeks gestation was similar in blood stained vs. control groups. (0.78 vs. 0.84) in blood stained vs. control groups respectively.

CONCLUSION: Blood stained swabs have elevated qfFN concentrations but may still have predictive value, and clinical utility. Very low fFN values (<10ng/ml) are especially reassuring and indicate lower risk of delivery than non-blood stained swabs. The higher false positive rate must be noted and explained to the patient.

View graph of relations

© 2020 King's College London | Strand | London WC2R 2LS | England | United Kingdom | Tel +44 (0)20 7836 5454