Evaluation of the ability of cervical length and fetal fibronectin measurement to predict preterm delivery in asymptomatic women with risk factors

Abstract

Objective: To evaluate the power of transvaginal ultrasound cervical length (CL) and fetal fibronectin (fFN) measurement for the prediction of preterm delivery (PTD) in asymptomatic women with risk factors.

Methods: Between January 2010 and October 2012, 218 women with singleton pregnancies and a high risk for PTD (i.e. prior history of PTD, cervical surgeries, late miscarriages, uterine abnormalities or premature rupture of the membranes) were selected. CL was measured by transvaginal ultrasonography between 14 and 22 weeks, and a CL < 2.5 cm was considered to be short. The fFN was measured with a sterile speculum examination. The primary outcome was the spontaneous PTD rate at <34⁺⁰ and <37⁺⁰ weeks, and the PTD rate was analyzed by either of two methods alone or combined.

Results: There were 213 cases included in the analysis, and 47 women (22.1%) had positive fFN. The average CL of women with positive fFN (2.3 ± 0.9 cm) was significantly shorter than that of women with negative fFN (3.1 ± 0.8 cm; p = 0.02). There were 25 (11.7%) and 38 (18.8%) cases with PTD at <34⁺⁰ and <37⁺⁰ weeks, respectively. By individual analyses, the highest PTD rate was 51.2% in women with positive fFN at <37⁺⁰ weeks. The sensitivity and diagnostic efficiency values for PTD prediction by CL were low at <34⁺⁰ and <37⁺⁰ weeks. The values of diagnostic efficiency by fFN were >80% at <34⁺⁰ and <37⁺⁰ weeks. In the combined analysis, the highest PTD rate was 57.1% in women with CL ≤ 2.5 cm and positive fFN at <37⁺⁰ weeks. The Kaplan–Meier curve indicated that there were significant differences among the four groups (χ²= 54.261; p = 0.000).

Conclusions: These results indicate a significant correlation between fFN and PTD prediction. The PTD prediction could be enhanced by combined CL and fFN measurement.

• Cervical length
• diagnostic prediction
• fetal fibronectin
• preterm delivery

Declaration of interest

The authors report no conflicts of interest.