Aust N Z J Obstet Gynaecol 2019; 1–6

DOI: 10.1111/ajo.12947

ORIGINAL ARTICLE

The efficacy of quantitative fetal fibronectin in

predicting spontaneous preterm birth in symptomatic
women: A retrospective cohort study

Anh Duy Nguyen1,†, Cathy Zhenao Liu1,2, Christoph Lehner3 ,

Akwasi Atakora Amoako1,2 and Renuka Sekar3

1
Faculty of Medicine, University
of Queensland, Brisbane, Background: Recent data suggest that quantitative measurements of fetal fi-
Queensland, Australia bronectin can be used accurately to predict increased risk of preterm birth.
2
Department of Obstetrics and Aim: The purpose of this study was to demonstrate that the quantification of
Gynaecology, The Royal Brisbane
and Women's Hospital, Brisbane, fetal fibronectin improves diagnostic accuracy in women who present with symp-
Queensland, Australia toms suggestive of threatened preterm labour (TPL) using a quantitative fetal fi-
3
Centre for Advanced Prenatal
bronectin (qfFN) bedside analyser.
Care, The Royal Brisbane and
Women's Hospital, Brisbane, Study design: This was a retrospective cohort study of pregnant women who
Queensland, Australia
presented between 22+6 and 32+6 weeks gestation with symptoms of TPL who
Correspondence: Dr Akwasi A. had qfFN measured using the Rapid fFN Q10 system. The ability to predict spon-
Amoako, Discipline of Obstetrics and
taneous preterm birth (sPTB) within 48 h, 14 days and <34 weeks gestation at
Gynaecology, University of Queensland,
Teaching and Research, Level 6 Ned qfFN thresholds of 10, 50 and 200 ng/mL was assessed.
Hanlon Building, Royal Brisbane &
Results: The overall rate of sPTB <34 weeks was 4.1% (n = 373). For deliveries
Women's Hospital, Butterfield Street,
Herston, Brisbane, Queensland 4029, within 48 h, within 14 days and <34 weeks, a qfFN threshold of 200 ng/mL had
Australia.
positive predictive values of 26.7%, 42.9% and 46.7%, respectively, when com-
Email: a.amoako@uq.edu.au
†
pared to patients with qfFN values of 0–9 ng/mL. The corresponding relative risks
Present address: The Princess
Alexandra Hospital, Brisbane, were 68.5, 53.8 and 38.0, respectively
Queensland, Australia Conclusion: Quantitative fetal fibronectin testing with thresholds of 10, 50 and
Conflict of Interest: All authors confirm 200 ng/mL allows for more accurate prediction of preterm birth in symptomatic
they have no conflicts of interest in
relation to this work. women. This higher degree of discrimination allows for more directed interven-
tions for high-­risk patients and reduces the cost and burden of unnecessary
Received: 3 September 2018;
Accepted: 9 December 2018 treatment for low-­risk patients.

KEYWORDS
fetal fibronectin, prediction, pregnancy, spontaneous preterm birth, symptomatic women

INTRODUCTION hypothermia, hypoglycaemia, hyperbilirubinaemia and other

In Australia, one in ten births are preterm, meaning that deliv-
1
ery occurs before 37 weeks gestation. Those babies born pre-
maturely are at increased risk of adverse neonatal outcomes
closely linked to the gestational age at delivery, such as respi-
ratory distress syndrome, sepsis, intraventricular haemorrhage,
long-­term morbidities.2 Approximately two-­thirds of preterm ba-
bies are born spontaneously as a result of either spontaneous
preterm birth (sPTB) or premature pre-­labour rupture of mem-
branes (PPROM), while the remaining cases of preterm births
are iatrogenic for a variety of reasons relating to maternal or
fetal health.3

wileyonlinelibrary.com/journal/anzjog © 2019 The Royal Australian and New Zealand College of Obstetricians and Gynaecologists 1
2 Quantitative fetal fibronectin and preterm labour
The mechanisms which cause preterm labour are not well
understood. Currently, preterm labour is considered a syndrome
which can be triggered by multiple mechanisms including in-
fection or inflammation, uterine overdistension, uteroplacental
ischaemia or haemorrhage, stress and other immunologically
3
mediated processes. However, it is estimated that no cause is
4
found in 50% of preterm births. Risk factors include prior his-
tory of preterm birth, multiple pregnancy, short cervix, uterine
abnormalities, infection and pregnancy during adolescence or
advanced maternal age, short inter-­pregnancy interval, low body
mass index and low socioeconomic and smoking status.3
Given the increased neonatal morbidity and mortality of those
babies born prematurely, it is necessary to accurately identify
women at risk of preterm birth. Interventions (such as adminis-
tration of antenatal corticosteroids, magnesium sulphate for neu-
roprotection of the fetus, antibiotics and in utero transfer to an
obstetric centre) can mitigate adverse neonatal outcomes. However,
ultimately most women with symptoms of preterm labour do not
deliver within one week and 50% will go on to give birth at full term
(at or beyond 37 weeks gestation).5 Thus, the uncertainty of tim-
ing of delivery may lead to intervention which may not be useful.
Diagnostic tools which help clinicians to define women at risk of
premature delivery can be helpful in guiding clinical management.6
Fetal fibronectin (fFN) is a large glycoprotein found at the inter-
phase of the chorion and decidua. Under normal circumstances, it
is not found in the cervicovaginal fluid between 24 to 34 weeks of
gestation.7 However, its presence suggests a disruption of the chorio-
decidua interphase which results in increased risk of preterm labour
for the woman. The qualitative fFN test is popularly used as a rule-­out
test for preterm birth with a high negative predictive value of 96%.
However, its ability to predict preterm birth is poor in both symptom-
atic and asymptomatic patients (positive predictive value of <30%).8
The qualitative fFN test uses a cut-­off of 50 ng/mL, and it has
been postulated that improved prediction could be achieved with
increasing fFN values. Several studies using the enzyme-­linked
immunosorbent assay (ELISA) technique for quantification of fFN
values have shown a positive correlation with increased risk of
sPTB. Recent development of a bedside test that can accurately
measure fFN concentrations has also been corroborated. 9 Fifty-­one of those were either lost to follow up or had no
The aim of this study was to investigate the diagnostic utility
of the novel bedside fFN analyser in a cohort of Australian women
considered at high risk of preterm birth. We will use these results
to evaluate the clinical management of patients who present with
symptoms of preterm labour.
MA TERIALS AND METHODS
This was a retrospective cohort study of pregnant women with sin-
gleton pregnancies who presented to a tertiary referral centre from
October 2013 to January 2018 with symptoms suggestive of preterm
labour. All cases with a quantitative FFN (qfFN) test were identified in
the patient's integrated electronic medical record (ieMR). The study
was approved by the local Human Research Ethics Committee Board
(HREC/18/QRBW/30). The outcomes measured in this study were
sPTB within 48 h of qfFN testing, sPTB within 14 days of qfFN testing
and sPTB prior to 34 weeks gestation. In order to reduce the number
of false-­positives, women who presented in TPL outside the gesta-
tional range of 22+0 weeks to 32+6 weeks were excluded from data
analysis. In addition, patients who had iatrogenic delivery before the
outcome of interest were also excluded.
Following the recommendations of the Rapid fFN 10Q system
(Hologic, Marlborough, MA, USA), the specimens were collected
during speculum examinations using a sterile swab inserted in the
posterior fornix. Women who had rupture of membranes, active
bleeding or reported recent vaginal intercourse within 24 h were ex-
cluded from having the test. The Rapid fFN 10Q analyser produced
discrete values of qfFN ranging from 0 to >500 ng/mL. Patients’
background information, obstetric history, and birth outcomes
were collected from the hospital's ieMR. Statistical analysis was
done using MedCalc Statistical Software version 18.2.1 (MedCalc
Software bvba, Ostend, Belgium; http://www.medcalc.org; 2018).
Patients were separated into four different categories according
to the qfFN results: 0–9, 10–49, 50–199 and >200 ng/mL with three
different cut-­off points of 10, 50 and 200 ng/mL, respectively, to cal-
culate sensitivity, specificity, likelihood ratios, area under the curve
(AUC), negative predictive value (NPV) and positive predictive value
(PPV) for the outcomes specified. Relative risk (relative to the lowest
category of qfFN 0–9 ng/mL) was also calculated for different fFN
categories, and the χ2 test was used to determine their significance.
Receiver operator characteristic curves were constructed to
evaluate the diagnostic capacities of the qfFN test. Stepwise lo-
gistic regression models were fit for the prediction of sPTB within
two weeks and sPTB <34, and the goodness of fit of both models
was confirmed using Hosmer and Lemeshow test.
RESULTS
There were a total of 605 women presenting in TPL with qfFN
testing performed between October 2013 to January 2018.
birth records on ieMR. In addition, 174 presentations were
excluded as per the predefined exclusion criteria, including
two iatrogenic deliveries that happened within 48 h of qfFN
testing. A total of 380 presentations of TPL between 22+0 to
32+6 weeks for analysis of sPTB within 48 h were eligible for
data extraction.
For the outcome of delivery within two weeks of testing, 181
presentations were excluded as per the predefined exclusion
criteria, including 12 iatrogenic deliveries that happened within
two weeks of qfFN testing. A total of 373 presentations of TPL be-
tween 22+0 to 32+6 weeks for analysis of sPTB within 14 days were
eligible for data extraction (Fig. 1).
For the outcome of sPTB <34 weeks gestation, there were
18 iatrogenic deliveries before 34 weeks gestations that were
A. D. Nguyen et al. 3

Total number of
presentations with qfFN
testing
(n = 605)

Lost to follow up or not

available on electronic
medical record
(n = 51)

Presentations with outcomes

available
(n = 554)

Presentations excluded as
per exclusion criteria
- Multiple pregnancy
- GA at testing <22+0 weeks
or >32 +6 weeks
- Iatrogenic delivery before
the outcome of interest
(n = 174, 181, 186
respectively)

F I G U R E   1   Flow diagram
Total presentations for Total presentations for Total presentations for
demonstrating total number of
analysis of sPTB within analysis of sPTB within analysis of sPTB less than
participants involved in the study and 48 h 14 days 34 weeks
those who were excluded as per the (n = 380) (n = 373) (n = 368)
predefined exclusion criteria.

excluded, resulting in a total of 367 presentations for analysis.
+2
The mean gestational age at presentation was 29 weeks, and
the mean gestation age at delivery was 38 weeks. The average la-
tency between qfFN testing and birth was 62 days. The median
maternal age was 28 years (Fig. S1).
Table 1 summarises the patients’ demographics and specific
obstetrics characteristics pertaining to preterm birth, including to-
colysis and steroids usage for those who presented with TPL. As
expected, there is a higher utilisation of tocolysis and steroids for
those who had a positive fFN results of >50 ng/mL as per the current
qualitative cut-­off (89.9% given steroids and 67.6% given tocolysis).
Nevertheless, for negative qualitative fFN results, there were still a
significant number of patients who were given these medications
prophylactically (14.7% given steroids and 9.4% given tocolysis).
The overall rate of sPTB <34 weeks was 4.1%. Women who
delivered within 14 days of testing had a higher mean qfFN com-
pared to women who delivered later than 14 days (214 vs 26 ng/
mL, respectively, Mann–Whitney U-­test, P < 0.00001). The rates of
sPTB <34 weeks gestation (n = 15) increased progressively with
increasing fFN concentrations from 1.2% (3/244), 2.7% (2/75),
9.1 (3/33) and 46.7% (7/15) for fFN concentrations of 0–9, 10–49,
50–199 and >200 ng/mL, respectively (Table S1). The relative risk
of sPTB <34 weeks also increased with higher fFN concentra-
tions when compared to the lowest fFN category of 0–9 ng/mL.
The relative risk was 2.2 (95% CI, 0.4–12.7, P < 0.4) for fFN 10–49;
7.4 (95% CI, 1.6–35.1, P < 0.004) for fFN 50–199; and 38.0 (95% CI,
10.9–132.2, P < 0.0001) for fFN > 200 ng/mL, respectively.
Similarly, the rates of sPTB within 14 days of testing (n = 11) also
increased progressively with increased fFN concentrations from
0.8% (2/251), 2.7% (2/74), 2.9% (1/34) to 42.9% (6/14) for fFN val-
ues of 0–9, 10–49, 50–199 and >200 ng/mL, respectively (Table S1).
There was a significant difference in the mean qfFN values of those
who delivered within 14 days (qfFN = 214 ng/mL) and those who
did not (qfFN = 24.8 ng/mL) of 189 ng/mL (Mann–Whitney U-­test,
P < 0.00001). The overall number of sPTB within 14 days of testing
was 11 (2.9%). The relative risks of sPTB within 14 days relative to
the lowest fFN category of 0–9 ng/mL also increased as fFN concen-
tration increased, with relative risk of 3.4 (95% CI, 0.5–23.7, P < 0.2)
for fFN 10–49; 3.7 (95% CI, 0.3–39.6, P < 0.3) for fFN 50–199; and 53.8
(95% CI, 11.9–242.8, P < 0.0001) for fFN >200 ng/mL, respectively.
There was a total of eight (2.1%) spontaneous deliveries
just within 48 h of qfFN testing. At a latency of within 48 h, the
previous trend of increasing rates of sPTB with increasing fFN
concentrations was again demonstrated with 0.4% (1/257),
2.6% (2/77), 3.2% (1/31) and 26.7% (4/15) for fFN values 0–9,
10–49, 50–199, >200 ng/mL, respectively (Table S1). This re-
sults in a relative risk of sPTB within 48 h when compared with
the lowest fFN category of 0–9 ng/mL of 6.7 (95% CI, 0.6–72.6,
4 Quantitative fetal fibronectin and preterm labour

TABLE 1 Demographics and specific characteristics data relating to preterm birth of participants involved in the study

Characteristic Value (%)

Maternal age† 28 ± 5.6 weeks

Ethnicity
Caucasian 281 (73.9)
Aboriginal and TSI 39 (10.3)
Asian 12 (3.2)
Indian 14 (3.7)
Pacific Islanders 10 (2.6)
Others 24 (6.3)
Obstetric history
Primiparous 130 (34.2)
Multiparous 250 (65.8)
BMI† 25.3 ± 6.4
Smoking 85 (23.5)
Tocolysis/steroids Steroids Tocolysis Both
fFN 0–9 23 (9.6) 18 (7.5) 16 (6.7)
fFN 10–49 14 (19.7) 8 (11.3) 8 (11.3)
fFN 50–199 27 (93.1) 16 (55.2) 16 (55.2)
fFN >200 13 (86.7) 12 (80.0) 12 (80.0)

†Data represented as mean ± SD.

BMI, body mass index; fFN, fetal fibronectin; TSI, Torres Strait Islander

P < 0.08) for fFN 10–49; 8.3 (95% CI, 0.5–129.3, P < 0.08) for
fFN 50–199; and 68.5 (95% CI, 8.2–575.9, P < 0.0001) for fFN
>200 ng/mL, respectively.
In our study, patients with the highest fFN measurable results
of >500 ng/mL had similar outcomes to those with fFN concen-
trations of 200–499 ng/mL. Considering the rates of sPTB within
14 days, there was no significant statistical difference (P = 0.5) be-
tween fFN 200–499 (50%, 5/10) and fFN >500 (25%, 1/3). Rates of
sPTB <34 weeks were 55% (6/11) for fFN 200–499 and 25% (1/4)
for fFN >500 ng/mL (P = 0.5). There was no fFN concentrations
>500 ng/mL in those who delivered within 48 h.
Table 2 summarises the diagnostic statistics of fFN for the two
different outcome measures. The PPVs for sPTB within 14 days
increased from 7.4%, 14.6% to 42.9% for fFN thresholds of 10, 50
and 200 ng/mL, respectively. NPVs remained relatively constant at
99.2%, 98.8% and 98.6%, respectively, with the same increase in
fFN thresholds. Similar numbers were seen with sPTB <34 weeks
gestation with PPVs of 9.8%, 20.8% and 46.7% for fFN thresholds
10, 50 and 200 ng/mL, respectively.
For sPTB within 48 h, there was again a trend of increasing
PPVs (5.7, 10.9 and 26.7) as the fFN threshold increases albeit
at an overall lower accuracy. Analysis of the receiver operating
characteristics of the different fFN thresholds (Fig. 2A–C) demon-
strates that the current cut-­off for the qualitative test of 50 ng/
mL remains valid with the highest AUC values for the outcomes
of sPTB within 14 days and sPTB less than 34 weeks gestation.
Interestingly, for sPTB within 48 h, the fFN threshold of 10 ng/mL
yields the highest AUC value (Fig. 2A).
DISCUSSION
Our study confirms the increased predictive power of qfFN
when compared to the qualitative test as previously postulated.8
However, there was roughly a twofold increase in the PPV when
the fFN threshold was increased to 200 ng/mL for all studied out-
comes, while the NPV remained relatively constant. The relative
risk of sPTB within the three specified outcomes also substantially
rose when fFN values exceeded 200 ng/mL. Thus, the qfFN test
adds greater descriptive values for the risk of sPTB compared to
the qualitative test, especially at and beyond the cut-­off value of
50 ng/mL and allows for better stratification of a patient's risk of
preterm delivery.
Our data demonstrate that 97% of women who presented
with symptoms of preterm labour did not go on to deliver sponta-
neously within 48 h and 14 days. Neither are they likely to deliver
prior to 34 weeks gestation and therefore be at risk of complica-
tions of preterm birth. Thus, the ability of qfFN to better discrimi-
nate between high-­risk and low-­risk patients may be valuable from
an economic and treatment perspective. This may include the need
for transfer to a tertiary centre, admission to a 24 h birth suite or
the antenatal ward and interventions such as corticosteroids and
tocolytics in high-­risk patients to mitigate the deleterious compli-
cations of preterm birth. In contrast, patients with very low fFN
concentrations (<10 ng/mL) can be reassured of being at low-­risk
of preterm delivery and thus do not need to be transferred or ad-
mitted to hospital. This is partly demonstrated in our tocolysis and
A. D. Nguyen et al. 5

TABLE 2 Diagnostic statistics for predicting sPTB within 48 h, sPTB within 14 days and sPTB < 34 weeks gestations according to
different fFN thresholds

sPTB within 48 h sPTB within 2 weeks sPTB < 34 weeks gestations
fFN threshold
(ng/mL) 10 50 200 10 50 200 10 50 200

Sensitivity 87.5 62.5 50.0 81.8 63.6 54.6 80.0 66.7 46.7
Specificity 68.8 89.0 97.0 68.8 88.7 97.8 68.5 89.2 97.7
ROC curve area 0.782 0.757 0.735 0.753 0.762 0.762 0.742 0.779 0.722
Positive likelihood 2.8 5.7 16.9 2.62 5.62 24.7 2.54 6.18 20.53
ratio
Negative 0.18 0.42 0.52 0.26 0.41 0.47 0.29 0.37 0.55
likelihood ratio
Positive predictive 5.7 10.9 26.7 7.4 14.6 42.9 9.8 20.8 46.7
value
Negative 99.6 99.1 99.2 99.2 98.8 98.6 98.8 98.4 97.7
predictive value
Relative risk 6.7 8.3 68.5* 3.4 3.7 53.8* 2.2 7.4** 38.0**
(relative to fFN
category (0–9)

χ2 test: *P < 0.0001, **P < 0.001

fFN, fetal fibronectin; ROC, receiver operating characteristic; sPTB, spontaneous preterm birth

F I G U R E   2   Receiver operating characteristic (ROC) curves demonstrating predictive accuracy of the quantitative fetal fibronectin
(qfFN) test for the three different outcomes. sPTB, spontaneous preterm birth

steroids data in Table 1 which show ongoing utilisation of these
medications at the lowest fFN category of 0–9 ng/mL despite hav-
ing the lowest rate of sPTB <34 weeks gestation of 1.6%. Indeed,
it has been postulated that qfFN improves cost savings of health
7
resources that have been elucidated with the qualitative test.
9
Compared to the UK-­based study by Abbott et al., we were able
to demonstrate a higher PPV (42.9% vs 37.0%) and relative risk (53.8
vs 16.1) at the fFN threshold of 200 ng/mL. However, at the highest
fFN concentrations of >500 ng/mL, we were not able to show an in-
creased predictive power over the fFN threshold of 200 ng/mL that
was demonstrated in the Abbott study. Similarly, when comparing
the outcome for sPTB <34 weeks gestation, we were not able to
demonstrate an increase in PPV between fFN thresholds of 200
and 500 ng/mL which has been reported by Abbott et al. (61.1%
and 75.0% vs 46.7% and 25.0% for 200 and 500 ng/mL, respec-
9
tively). This is despite similar sample sizes in both studies (n = 373
in this study vs n = 300 in Abbott's). Nevertheless, this is likely due
to insufficient sample size with only four presentations having fFN
>500 ng/mL. A proposed reason for the smaller incidence of pa-
tients with fFN >500 ng/mL is that Abbott et al.'s sample included
women who were further along in their gestation (up to 35+6).
Thus, there is a greater chance of uterine activity to occur closer to
term and hence increased incidence of elevated qfFN.
The utility of qfFN was first demonstrated in a secondary analysis
of the Preterm Prediction Study which showed a correlation between
risk of sPTB and rising fFN concentrations from 20 to 300 ng/mL.10
Kurtzman et al.11 subsequently corroborated these findings using a
qfFN testing at 24 weeks in their cohort of high-­risk asymptomatic
patients.11 In both studies, fFN concentrations were measured using
ELISA techniques which are considered time-­consuming and require
dedicated laboratories. Abbott et al.9 and Centra et al.12 confirmed
the increased diagnostic accuracy of qfFN using the Hologic fFN 10Q
6 Quantitative fetal fibronectin and preterm labour
analyser which greatly improved the utility of qfFN to be used as a
point of care test. In addition, the usefulness of qfFN has also been
demonstrated in women with prior cervical surgery.13
Transvaginal (TV) cervical length measurements have been
shown to be useful to assess the individual risk of preterm de-
livery. A cervical length of less than 25 mm has a sensitivity of
78.3% and specificity of 70.8% in predicting preterm birth within a
week.14 Recent studies examined the relation between qfFN and
cervical length measurements in predicting sPTB. The results have
shown that the combination of both failed to elucidate any addi-
tional benefit in both a cohort of low-­risk asymptomatic patients
as well as symptomatic patients.15,16
Considering that preterm labour is associated with placental
inflammation, Krogt et al.17 illustrated that increasing fFN con-
centrations correlated with inflammatory placental pathology in
women with sPTB. Intrauterine infection is thought to drive the
local release of inflammatory chemokines and cytokines which
stimulate the production of prostaglandins and leads to uterine
contraction and preterm labour. This was demonstrated in an
analysis of the inflammatory mediators such as RANTES (regu-
lated on activation, normal T cell expressed and secreted) and
interleukin-­8β in cervicovaginal fluid in asymptomatic women for
preterm birth.18 These studies illustrate that levels of cytokines
correlate with qfFN levels and are highly predictive for sPTB.
We were able to replicate results that corroborate other large
centre studies9 with our study. Limitations of our study include the
retrospective analysis of data which lacks the rigorous predefined
study protocol of prospective studies and the significant number
of iatrogenic deliveries which can affect the overall results.
CONCLUSION
Our study confirms that qfFN testing is helpful in predicting
spontaneous preterm birth in symptomatic women. Specified
fFN thresholds can be used to define women at risk, and hence,
direct necessary interventions for high-­risk patients only deter-
mined by increasing fFN levels. A planned prospective study will
help us to validate these results. Our ultimate goal is to create
a hospital-­based guideline for managing women considered at
high-­risk of preterm birth and thus improve outcomes for this
group. Such a guideline would also help us to avoid unnecessary
transfer from rural hospitals to our tertiary obstetric centre. Our
findings confirm previously published data that using qfFN at a
cut-­off level of 50 ng/mL reduces the burden and associated costs
of unnecessary treatment.
REFERENCES
1. Australian Institute of Health and Welfare. Australia's Mothers And
Babies 2016—in Brief. Canberra: Australian Institute of Health and
Welfare, 2018.
2. Purisch SE, Gyamfi-Bannerman C. Epidemiology of preterm birth.
Semin Perinatol 2017; 41: 387–391.
3. Goldenberg RL, Culhane JF, Iams JD, Romero R. Epidemiology and
causes of preterm birth. Lancet 2008; 371: 75–84.
4. Blencowe H, Cousens S, Chou D, et al. Born too soon: The global epi-
demiology of 15 million preterm births. Reprod Health 2013; 10: S2.
5. Georgiou HM, Di Quinzio MKW, Permezel M, Brennecke SP.
Predicting preterm labour: current status and future prospects.
Dis Markers 2015; 2015: 1–9.
6. Norwitz ER, Robinson JN. A systematic approach to the manage-
ment of preterm labor. Semin Perinatol 2001; 25: 223–235.
7. Dutta D, Norman JE. The efficacy of fetal fibronectin testing in
minimising hospital admissions, length of hospital stay and cost
savings in women presenting with symptoms of pre-­term labour.
J Obstet Gynaecol 2010; 30: 768–773.
8. Hezelgrave NL, Shennan AH. Quantitative fetal fibronectin to pre-
dict spontaneous preterm birth: a review. Womens Health 2016;
12: 121–128.
9. Abbott DS, Radford SK, Seed PT, et al. Evaluation of a quantitative
fetal fibronectin test for spontaneous preterm birth in symptom-
atic women. Am J Obstet Gynecol 2013; 208: 122.e1–e6.
10. Goepfert AR, Goldenberg RL, Mercer B, et al. The preterm prediction
study: quantitative fetal fibronectin values and the prediction of spon-
taneous preterm birth. Am J Obstet Gynecol 2000; 183: 1480–1483.
11. Kurtzman J, Chandiramani M, Briley A, et  al. Quantitative fetal
fibronectin screening in asymptomatic high-­risk patients and
the spectrum of risk for recurrent preterm delivery. Am J Obstet
Gynecol 2009; 200: 263.e1–e6.
12. Centra M, Coata G, Picchiassi E, et al. Evaluation of quantitative fFn test
in predicting the risk of preterm birth. J Perinat Med 2017; 45: 91–98.
13. Vandermolen BI, Hezelgrave NL, Smout EM, et  al. Quantitative
fetal fibronectin and cervical length to predict preterm birth in
asymptomatic women with previous cervical surgery. Am J Obstet
Gynecol 2016; 215: 480.e1–e10.
14. Sotiriadis A, Papatheodorou S, Kavvadias A, Makrydimas G.
Transvaginal cervical length measurement for prediction of
preterm birth in women with threatened preterm labor: a meta-­
analysis. Ultrasound Obstet Gynecol 2010; 35: 54–64.
15. Jwala S, Tran TL, Terenna C, et  al. Evaluation of additive effect
of quantitative fetal fibronectin to cervical length for prediction
of spontaneous preterm birth among asymptomatic low-­risk
women. Acta Obstet Gynecol Scand 2016; 95: 948–955.
16. Bruijn MMC, Vis JY, Wilms FF, et  al. Quantitative fetal fibronec-
tin testing in combination with cervical length measurement in
the prediction of spontaneous preterm delivery in symptomatic
women. BJOG 2015; 123: 1965–1971.
17. van der Krogt L, Ridout AE, Seed PT, Shennan AH. Placental in-
flammation and its relationship to cervicovaginal fetal fibronec-
tin in preterm birth. Eur J Obstet Gynecol Reprod Biol 2017; 214:
173–177.
18. Amabebe E, Chapman DR, Stern VL, et al. Mid-­gestational changes
in cervicovaginal fluid cytokine levels in asymptomatic pregnant
women are predictive markers of inflammation-­associated spon-
taneous preterm birth. J Reprod Immunol 2018; 126: 1–10.
SUPPORTING INFORMATION
Additional supporting information may be found online in the
Supporting Information section at the end of the article.
Table S1. Breakdown of spontaneous preterm birth rates be-
tween different fetal fibronectin (fFN) concentrations.
Figure S1. Distribution graph of gestational age at presentation.