Jc. Morrison et al., PREDICTION OF SPONTANEOUS PRETERM BIRTH BY FETAL FIBRONECTIN AND UTERINE ACTIVITY, Obstetrics and gynecology, 87(5), 1996, pp. 649-655
Objective: To evaluate the usefulness of fetal fibronectin and home ut
erine contraction assessment in predicting preterm birth (before 34 we
eks) in at-risk asymptomatic women. Methods: One hundred fifty women w
ere enrolled prospectively; five were lost to follow-up, leaving 145 w
omen available for analysis. Because patients with preterm labor befor
e 34 weeks' gestation most commonly develop this problem after 28 week
s, the period of 26-28 weeks' gestation was selected prospectively as
the first window for prediction and study analysis. Eighty-five of 145
asymptomatic women at high risk for preterm birth had both home uteri
ne contraction assessment of 2 hours per day and one or more cervical
sampling(s) for fetal fibronectin measurement at 26-28 weeks. A positi
ve home uterine contraction assessment was defined as contractions exc
eeding two per hour averaged over the 2-week study interval. Positive
fetal fibronectin was defined as greater than 50 ng/mL. Results: Fourt
een of the 85 women (16.5%) delivered before 34 weeks. Home uterine co
ntraction assessment alone had a sensitivity, specificity, positive pr
edictive value, and negative predictive value for preterm birth of 64,
85, 45, and 92%, respectively; fetal fibronectin alone was associated
with values of 43, 89, 43, and 89%, respectively. A positive home ute
rine contraction assessment was associated with a relative risk (RR) f
or preterm birth of 5.9 (95% confidence interval [CI] 2.4-14.2), where
as a positive fetal fibronectin demonstrated an RR of 3.8 (95% CI 1.5-
9.4). When both assessments were positive, all patients delivered befo
re 34 weeks and there was an RR of 27.0 (95% CI 8.7-84.1) compared wit
h those with both tests being negative. Only two patients with both te
sts negative delivered before 34 weeks (negative predictive value 96%)
. Conclusion: Both the home uterine contraction assessment and fetal f
ibronectin accurately predicted preterm birth before 34 weeks. When bo
th tests were combined, the predictive ability improved substantially.