Professor, Department of Obstetrics and Gynecology, University of Colorado School of Medicine, Aurora
Dr. Hobbins reports no financial relationships relevant to this field of study.
SYNOPSIS: A recent letter to the editor disputes the conclusion of an earlier study that fetal fibronectin is of little value in threatened preterm labor, despite other studies suggesting that when used in conjunction with cervical length measurements, it can diminish unnecessary hospitalizations appreciably.
SOURCE: van Baaren GJ, Bruijn MMC, Mol BW, et al. Randomized trials are not always the best way to assess diagnostic tests: The case of fetal fibronectin testing. Am J Obstet Gynecol 2017; Nov. 21. pii: S0002-9378(17)31230-9. doi: 10.1016/j.ajog.2017.10.037. [Epub ahead of print].
The featured letter to the editor, along with two responses,1,2 were published in the January issue of the American Journal of Obstetrics and Gynecology, all pertaining to a 2016 article by Berghella et al3 about the use of fetal fibronectin (fFN) in patients with threatened preterm labor (PTL). This interplay provided very interesting commentary regarding the interpretation of the original study results and the benefits/shortcomings of randomized, controlled trials (RCTs) and observational studies. First, the index study: Berghella et al trolled the Cochrane database for RCTs involving fFN alone as a predictor of preterm delivery < 37 weeks.3 They found six studies that satisfied their inclusion requirements. These studies all involved patients being evaluated for threatened PTL. Half of 546 patients, who were between 23 and 35 weeks’ gestation, were randomized to having fFN testing available to their managing providers shortly after the patients presented. In the other half, the test results were not available. None of the studies dealt with cervical length (CL) measurements. In three studies, a common protocol for management and/or therapy was used, but in the remaining three studies, the management was ad lib.
The authors found no statistically significant differences in the rates of delivery < 37 weeks (their primary outcome measure), delivery < 32 weeks, length of hospitalization, use of steroids or tocolytics, neonatal respiratory distress syndrome, or length of neonatal stays. The only difference was an average increased cost of $153 per patient in the fFN group. The authors’ concluded that use of fFN was not associated with a decrease in preterm birth or any improvement in perinatal outcome.
In a companion editorial in the same issue, Macones went one step further in suggesting that the use of fFN in threatened preterm labor cannot be justified.4
The above study and editorial sparked the featured letter to the editor in this month’s issue of the American Journal of Obstetrics and Gynecology from van Baaren and colleagues, who pointed out that most patients with threatened PTL are not in labor and the benefit of selectively used fFN is in its high negative predictive value (NPV). Their own observational study showed that a negative fFN, used in conjunction with CL in this setting, would allow most patients to be discharged safely without the need for tocolysis or steroids.5 They believed that before concluding that fFN was worthless in threatened PTL, the study should include a uniform management protocol regarding how to deal with the results of the fFN test.
In response to this letter to the editor, both Berghella and Macones seemed to agree with van Baaren that a predictive test such as fFN alone cannot change an outcome unless it segues into some type of action that works. Steroids do not stop PTL and even the overall benefit of tocolysis in PTL has been questioned over the years. However, Macones noted that the investigative power of RCTs was superior to observational studies, since they show what “does” happen with a particular test or action, rather than what “could” happen (with observational studies).
After wading through the entertaining collegial jousting, it should be clear that fFN does not have significant benefit in a vacuum. However, when used selectively in conjunction with CL in a regimented protocol — as suggested by van Baaren in the letter to the editor, Berghella in his response to the letter to the editor,1 and in a 2007 RCT for which Berghella was an author6 — it “will” decrease unnecessary hospitalizations, interventions, and healthcare costs.
If using fFN, the following is a re-formatted evaluation of threatened PTL initially suggested by van Baaren et al:5 On initial evaluation, patients with painful preterm contractions can have a sample taken with a swab from the posterior fornix of the vagina, which will be held in abeyance until the rest of the initial evaluation has been completed. This would include transvaginal sonography, in conjunction with a pelvic exam, to assess the status of the cervix. A CL < 1.5 cm would suggest the need for hospital admission (and the swab would be discarded since fFN would offer no additional benefit). If CL > 3 cm, the high NPV of this measurement alone would not be enhanced by fFN and the patient could be discharged. However, if the cervical length is between 1.5 cm and 3.0 cm, a positive fFN result would trigger admission and a negative result would mean that the patient would simply be followed carefully with outpatient visits.
Although both RCTs and observational studies have strong merit alone, they are, in fact, linked symbiotically. Often, an evidence-based evaluation of a concept or method can be accomplished only through an RCT. However, without observational studies there would be no concepts to test. Also, at times, the data from an observational study are so compelling — and the likelihood is so low of getting the results tested through a timely, flawlessly designed RCT — that management could be guided legitimately by the results of the observational study. The focus should not necessarily be on whether fFN decreases the rate of preterm birth, but should be on whether it can keep patients from having an unnecessary (and disruptive) hospitalization at a time when healthcare costs are out of control.
- Berghella V, Saccone G. Reply L17-093AR1. Am J Obstet Gynecol 2017; Nov. 4. pii: S0002-9378(17)31231-0. doi: 10.1016/j.ajog.2017.10.038. [Epub ahead of print].
- Macones GA. Reply. Am J Obstet Gynecol 2017; Nov 4. pii: S0002-9378(17)31232-2. doi: 10.1016/j.ajog.2017.10.039. [Epub ahead of print].
- Berghella V, Saccone G. Fetal fibronectin testing for prevention of preterm birth in singleton pregnancies with threatened preterm labor: A systematic review and meta-analysis of randomized controlled trials. Am J Obstet Gynecol 2016;215:431-438.
- Macones GA. Fetal fibronectin testing in threatened preterm labor: Time to stop. Am J Obstet Gynecol 2016;215:405.
- van Baaren GJ, Vis JY, Wilms FF, et al. Predictive value of cervical length measurement and fibronectin testing in threatened preterm labor. Obstet Gynecol 2014;123:1185-1192.
- Ness A, Visitine J, Ricci E, Berghella V. Does knowledge of cervical length and fetal fibronectin affect management of women with threatened preterm labor? A randomized trial. Am J Obstet Gynecol 2007;197:426.e1-7.