Recently, randomized controlled trials have demonstrated a benefit of vaginal progesterone therapy for the reduction of preterm birth in women with a short cervix.1,2 Two trials included women with previous preterm births, nulliparous women, and parous women without a history of preterm birth.1,2 The natural history of short cervix in women with a history of preterm birth and nulliparous women has been well-described.3,4 Conversely, birth outcomes associated with a short cervix in women with obstetric histories that place them at low risk for preterm birth (eg, parous with only previous term births) have not been adequately described. It is important to understand the natural history in this population of women who are generally considered to be at low risk for preterm birth for the following two reasons: to be able to adequately counsel them regarding the potential benefit of progesterone therapy for an ultrasound-detected short cervix and to understand the implications of implementing cervical length screening in this group of women. Therefore, the objective of our study was to understand the relationship between cervical length and the risk of preterm birth in parous women without a history of preterm delivery and to compare them with women with a history of spontaneous preterm birth and nulliparous women.
MATERIALS AND METHODS
This study was a secondary analysis of data collected for a multicenter, prospective, observational cohort study originally conducted by the Maternal Fetal Medicine Units Network of the National Institute of Child Health and Human Development between October 1992 and July 1994.5 This deidentified dataset is available through the Eunice Kennedy Shriver National Institute of Child Health and Human Development. The purpose of the original study was to identify predictors of preterm birth. The investigation was approved by the human subjects review board at each institution. All participating women provided written informed consent. Women with singleton gestations and intact membranes were enrolled before 24 weeks of gestation and underwent serial cervical length measurements. A detailed description of the study protocol, including the cervical length measurement protocol, has been published elsewhere.5 In the original study, the primary outcome was defined as spontaneous preterm birth at less than 35 weeks of gestation. We chose to define spontaneous preterm birth as a birth after preterm labor or rupture of membranes at more than 22 6/7 weeks of gestation and less than 37 0/7 weeks of gestation, and early spontaneous preterm birth was defined as birth occurring after preterm labor or rupture of membranes at less than 34 0/7 weeks of gestation. We subgrouped the study population into women with a history of at least one spontaneous preterm birth (previous preterm birth group), nulliparous women, and parous women with a history of at least one term birth and no preterm birth (low-risk history group). Cervical length data from the first study visit (22–24 6/7 weeks of gestation) were used for this analysis. Cervical length cutoffs of 20 mm or less and 15 mm or less were used to define a short cervix. These cut points were specified before analysis.
Demographic characteristics of the subgroups were compared using analysis of variance and χ2 analyses, as appropriate. For each subgroup, the relationship between cervical length (measured between 22 and 22 6/7 weeks of gestation), as a continuous variable, and preterm birth was examined using logistic regression. The frequencies of a short cervix and of preterm birth when a short cervix was present were calculated for each group. Assuming a 40% risk reduction with the use of vaginal progesterone,1,2 we then calculated the number needed to screen to prevent one preterm birth before 37 and 34 weeks of gestation in each subgroup using cervical length cutoff values of 20 mm or less and 15 mm or less.
There were 3,073 women in the cohort; 3,056 had cervical length assessments between 22 and 24 weeks of gestation, and 2,998 of these women also had data regarding gestational age at delivery. Four hundred sixty-seven (15.6%) had a previous spontaneous preterm birth, 1,237 (41.4%) were nulliparous, and 1,284 (43%) had low-risk histories. The demographic characteristics of the three subgroups are shown in Table 1.
Overall, the frequencies of spontaneous preterm birth were 32.2% for women with a previous preterm birth, 12.5% for nulliparous women, and 10.5% for women at low risk. Cervical length was normally distributed for each subgroup (data not shown). The median cervical lengths (range) were 35 mm (0–63 mm), 34 mm (5–67 mm), and 36 mm (3–70 mm) for women with a previous preterm birth, nulliparous women, and women at low risk, respectively. The estimated probability of preterm birth by cervical length for the three subgroups is shown in Figure 1. Although the shape of the curves differs qualitatively between groups, the negative relationship between cervical length and risk of preterm birth is apparent in each subgroup. As expected, women with a previous preterm birth had the highest risk of preterm birth at any given cervical length when compared with women in the other two groups.
For each subgroup, the frequency of short cervix (95% confidence interval [CI]) and the rate of preterm birth and early spontaneous preterm birth are presented in Table 2. As a reference point, the rate of preterm birth for women with a cervical length more than 15 mm also is presented in Table 2. Using the data from Table 2, and assuming a 40% risk reduction with the use of vaginal progesterone, we estimated the number needed to screen (NNS) to prevent one preterm birth for each subgroup. These data are presented in Table 3. For a cutoff of 15 mm or less, preventing one early spontaneous preterm birth would require screening 167 (95% CI 112–317) women with a previous preterm birth, 344 (95% CI 249–555) nulliparous women, and 1,075 (95% CI 667–2,500) women with a low-risk history.
Our findings confirm that an inverse relationship exists between the length of the cervix, as measured by ultrasonography during the second trimester, and the frequency of preterm birth even in women with a reassuring obstetric history. The overall frequency of preterm birth was lowest for parous women without a history of preterm birth (10.5%); yet, the frequency of preterm birth for these women at low risk with a short cervix (15 mm or less) was 25%. However, their frequency of short cervix was only 0.93% compared with 3.4% in women with a previous spontaneous preterm birth and 2.1% in nulliparous women. Therefore, the NNS to prevent one preterm birth is considerably greater for women who are parous and have had only term births.
The data presented in this analysis originated from a research cohort recruited 20 years ago. Not surprisingly, the demographics and health behaviors across each subgroup are not necessarily reflective of the current U.S. population. For example, when compared with many modern pregnancy cohorts, this group of women had lower prepregnancy weights and a higher frequency of tobacco use. These demographic and social differences may have contributed to a higher-than-expected spontaneous preterm birth rate in our low-risk group (10.5%). Supposing the rate of preterm birth among contemporary women at low risk is less than what we observed in this population, the NNS to prevent one preterm birth with vaginal progesterone would be even greater.
Two randomized controlled trials have demonstrated a benefit of vaginal progesterone therapy to reduce the risk of preterm birth in women with an ultrasonographically detected short cervix. Fonseca et al1 studied a general obstetric population, identifying 250 women with a cervical length of 15 mm or less; 29% of participants were parous without a history of preterm birth. They reported that for the population as a whole, progesterone was associated with a reduced risk of spontaneous birth before 34 weeks of gestation (relative risk 0.56; 95% CI 0.36– 0.86) and demonstrated that the relative risk of spontaneous preterm birth with vaginal progesterone therapy did not appear to vary significantly according to obstetric history.1 Hassan et al2 studied an unselected population identifying 465 women with a cervical length between 10 and 20 mm. They reported that even in women without a history of a preterm birth (defined as nulliparous women and parous women without a previous birth between 20 and 35 weeks of gestation), vaginal progesterone was associated with a significant reduction in the rate of preterm birth before 33 weeks of gestation (relative risk 0.5; 95% CI 0.27–0.90).2 For our NNS analysis, we assumed a 40% reduction in preterm birth with the use of vaginal progesterone based on the data described. It should be noted that 95% CIs for the risk reduction associated with vaginal progesterone in the trials performed by Hassan et al2 and by Fonseca et al1 were 0.08–0.67 and 0.14–0.64, respectively. In terms of a sensitivity analysis, if the “true” effect of vaginal progesterone were as low as 0.08, the NNS to prevent one early spontaneous preterm birth in women at low risk would be 5,376. If the “true” effect of vaginal progesterone were as great as 0.67, the NNS would be 641.
When summarizing the merits and disadvantages of universal cervical length screening of women without a previous preterm birth, two main concerns arise: cost and benefit. In the trial of Fonseca et al,1 to identify 413 eligible women with a short cervix, more than 24,000 women, both at low risk and at high risk, were screened. Given the low prevalence of short cervix in parous women with no previous preterm birth, the cost-effectiveness of screening this low-risk population is called into question. Recently, Cahill et al6 and Werner et al7 have published economic analyses of cervical length screening. Both analyses reported that universal cervical length screening appears to be cost-saving even if the prevalence of a short cervix is as low as 0.9–1%. Regarding treatment benefit, although vaginal progesterone therapy is generally regarded as safe, the evidence for benefit in women at low risk is not robust. The data of Fonseca et al1 suggest that the benefit of vaginal progesterone does not vary significantly (relative risk 0.3–0.7) according to maternal age, body mass index, race, or obstetric history, but definitive subgroup risk estimates are limited by insufficient sample size.1
Our data demonstrate that a short cervix is associated with a significantly increased risk of preterm birth even among parous women without a previous preterm birth; however, the incidence of short cervix (15 mm or less) is less than 1% in this population. It is beyond the scope of this article to propose a formal policy recommendation regarding who should or should not undergo cervical length screening during pregnancy, but these data can help direct policy in the future. A sound understanding of the epidemiology and natural history of a short cervix in women at low risk will help equip obstetric providers, policymakers, and patients to make an informed decision weighing the risks, benefits, and acceptability of universal cervical length screening and the use of vaginal progesterone treatment for parous women without a history of preterm birth.
1. Fonseca EB, Celik E, Parra M, Singh M, Nicolaides KH. Fetal Medicine Foundation Second Trimester Screening Group. Progesterone and the risk of preterm birth among women with a short cervix. N Engl J Med 2007;357:462–9.
2. Hassan SS, Romero R, Vidyadhari D, Fusey S, Baxter JK, Khandelwal M, et al.. Vaginal progesterone reduces the rate of preterm birth in women with a sonographic short cervix: a multicenter, randomized, double-blind, placebo-controlled trial. Ultrasound Obstetrics Gynecol 2011;38:18–31.
3. Owen J, Hankins G, Iams JD, Berghella V, Sheffield JS, Perez-Delboy A, et al.. Multicenter randomized trial of cerclage for preterm birth prevention in high-risk women with shortened midtrimester cervical length. Am J Obstet Gynecol 2009;201:375.e1–8.
4. Grobman WA, Thom EA, Spong CY, Iams JD, Saade GR, Mercer BM, et al.. 17 alpha-hydroxyprogesterone caproate to prevent prematurity in nulliparas with cervical length less than 30 mm. Am J Obstet Gynecol 2012;207:390.e1–8.
5. Iams JD, Goldenberg RL, Meis PJ, Mercer BM, Moawad A, Das A, et al.. The length of the cervix and the risk of spontaneous premature delivery. National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network. N Engl J Med 1996;334:567–72.
6. Cahill AG, Odibo AO, Caughey AB, Stamilio DM, Hassan SS, Macones GA, et al.. Universal cervical length screening and treatment with vaginal progesterone to prevent preterm birth: a decision and economic analysis. Am J Obstet Gynecol 2010;202:548.e1–8.
© 2013 by The American College of Obstetricians and Gynecologists. Published by Wolters Kluwer Health, Inc. All rights reserved.
7. Werner EF, Han CS, Pettker CM, Buhimschi CS, Copel JA, Funai EF, et al.. Universal cervical-length screening to prevent preterm birth: a cost-effectiveness analysis. Ultrasound Obstet Gynecol 2011;38:32–7.