Three large randomized controlled trials investigating the efficacy of universal cervical length screening and treatment with vaginal progesterone or cervical cerclage to prevent preterm delivery have been published over the past several years. None of these trials demonstrate proven efficacy for universal cervical length screening and cerclage placement in women with short cervical length. However, universal cervical length screening and treatment with daily vaginal progesterone in women with short cervical length reduces the risk of preterm birth, but large numbers of women must be screened to prevent a relatively small number of preterm deliveries. Issues that should be considered while implementing universal cervical length screening include: (1) standards of quality and reproducibility for transvaginal ultrasound cervical length ascertainment; (2) implications of screening on the application of therapeutic strategies to populations not known to benefit (so-called “indication creep”); and (3) willingness of obstetricians to prescribe vaginal progesterone formulations that are not approved by the US Food and Drug Administration for preterm birth prevention. Optimal strategies to employ cervical ultrasound and progesterone treatment might be revealed by additional studies investigating cervical length cutoffs, frequency of screening, selective screening in higher-risk groups, and the use of transabdominal cervical length screening as a surrogate for transvaginal cervical length screening.
The incidence of preterm delivery in the United States remains >12% (>500,000 deliveries <37 weeks’ gestation annually). Most successful efforts to reduce the incidence of spontaneous preterm delivery have focused on women with previous preterm deliveries. Unfortunately, only 15% of preterm births occur in women with previous preterm births, so effective interventions that reduce the rate of recurrent preterm birth by 50% will prevent <40,000 preterm births annually. Therefore, successful screening strategies must be developed to prevent preterm births in women with no history of preterm delivery.
Approximately 20 years ago, Smith et al and Andersen et al first used transvaginal sonography to measure cervical length during pregnancy and predict the risk of preterm delivery, and in 1996, Iams et al conclusively demonstrated that the risk of spontaneous preterm delivery was increased in women who were found to have a short cervix by transvaginal ultrasound performed at approximately 24 and 28 weeks’ gestation. Others have since demonstrated the usefulness of cervical length screening and cerclage placement in women with previous spontaneous preterm deliveries. However, the role of cervical length screening in women with no history of spontaneous preterm delivery is less well established. As obstetricians begin to implement universal cervical length screening in pregnant women, issues that must be considered include the number of women who will need to be screened to prevent 1 preterm birth, the logistics of implementing a reliable system of ultrasound screening, the frequency of short cervical length, the effectiveness of treatments (ie, progesterone and/or cervical cerclage), and the overall impact on the preterm birth rate.
Randomized trials utilizing universal cervical length screening
Three large randomized controlled trials investigating the efficacy of universal cervical length screening to prevent preterm delivery have been published over the past 10 years. One of these trials (To et al ) studied cervical cerclage placement as the treatment for women with short cervical lengths, while the other 2 trials (Fonseca et al, Hassan et al ) studied vaginal progesterone as the treatment for women with short cervical lengths. In these trials, women with major fetal abnormalities, painful regular uterine contractions, and history of ruptured membranes or cervical cerclage in situ were excluded. In the most recent trial by Hassan et al, women with chronic medical conditions, uterine malformations, vaginal bleeding, or chorioamnionitis also were excluded. A brief review of these trials is provided here.
In the trial by To et al, women underwent cervical length screening at 22-24 weeks’ gestation and were randomized to Shirodkar cervical cerclage vs no treatment when the cervical length measured ≤1.5 cm. The primary outcome compared between treatment groups was preterm delivery <33 completed weeks (33 0/7 ) of pregnancy. The study population consisted of women with singleton pregnancies at 12 hospitals in Europe, South Africa, and South America. A total of 47,123 women underwent transvaginal cervical length screening, and 470 women (1.0%) had a cervical length ≤1.5 cm. Among these 470 women, 253 were enrolled in the randomized trial: 127 received a cerclage and 126 did not. Approximately 25% of randomized subjects were nulliparous, and 15% had a previous preterm delivery. The frequency of preterm delivery at <33 weeks was 22% (28/127) in the cerclage group vs 26% (33/126) in the no cerclage group (relative risk [RR], 0.84; P = .44). The study was powered to detect a 3-fold reduction in preterm delivery in the cerclage group, and subset analyses exclusive to nulliparous subjects or women with no history of preterm delivery were not presented.
In the trial by Fonseca et al, women underwent cervical length screening between 20-25 weeks’ gestation and were randomized to micronized vaginal progesterone (200 mg capsules each night) vs placebo when the cervical length measured ≤1.5 cm. The primary outcome compared between treatment groups was spontaneous preterm delivery <34 completed weeks (34 0/7 ) of pregnancy. The study population consisted of women with singleton or twin pregnancies undergoing routine ultrasonography at 8 hospitals in the United Kingdom, Greece, Brazil, and Chile. A total of 24,620 women underwent transvaginal cervical length screening, and 413 women (1.7%) had a cervical length ≤1.5 cm. Among these 413 women, 250 were enrolled in the randomized trial: 125 received daily vaginal progesterone and 125 received placebo. Approximately 55% of randomized women were nulliparous, 10% had twins, and 15% had a previous preterm delivery. The frequency of spontaneous preterm delivery at <34 weeks was 19% (24/125) in the daily progesterone group vs 34% (43/125) in the placebo group (RR, 0.56; P = .02). Two women in each group had indicated preterm births <34 weeks.
In the recent trial by Hassan et al, women underwent cervical length screening between 19-24 weeks’ gestation and were randomized to vaginal progesterone (90 mg) gel daily vs placebo when the cervical length measured 1.0-2.0 cm. The primary outcome compared between treatment groups was preterm delivery <33 completed weeks (33 0/7 ) of pregnancy. The study population consisted of women with singleton pregnancies undergoing routine ultrasonography at 44 hospitals in 10 countries. A total of 32,091 women underwent transvaginal cervical length screening. Among these 733 women, 465 were enrolled in the randomized trial: 235 received daily vaginal progesterone, 223 received placebo gel, and 7 were lost to follow-up (not analyzed). Approximately 55% of randomized subjects were nulliparous, and 13% had a previous preterm delivery. The frequency of preterm delivery at <33 weeks was 8.9% (21/235) in the daily progesterone group vs 16.1% (36/223) in the placebo group (RR, 0.55; P = .02). Frequencies of preterm birth <28 weeks, preterm birth <35 weeks, neonatal respiratory distress syndrome, and birthweight <1500 g were significantly lower ( P < .05) in the daily progesterone group. In a subset analysis of women without a history of delivery between 20-32 weeks (84% of the study population), the frequency of preterm delivery at <33 weeks was 7.6% (15/197) in the daily progesterone group vs 15.3% (29/189) in the placebo group (RR, 0.50; P = .02). A subset analysis exclusive to nulliparous subjects was not presented.
Analysis of randomized controlled trials utilizing universal cervical length screening
No randomized trials demonstrate proven efficacy for universal cervical length screening and cerclage placement in women with short cervical length. A recent metaanalysis of the 2 large trials that used universal cervical length screening and treatment with daily vaginal progesterone in women with short cervical length and 3 significantly smaller trials (22-42 subjects) demonstrated that treatment with daily vaginal progesterone reduces the risk of preterm birth. However, large numbers of women must be screened to prevent a relatively small number of preterm deliveries ( Table 1 ). Based on rates of preterm birth and short cervical length in the previously published universal screening cohorts, and the efficacy of treatment with daily vaginal progesterone, universal cervical length screening will prevent 25 preterm births <34 weeks and 17 preterm births <33 weeks when 10,000 women undergo cervical length screening according to the trials performed by Fonseca et al and Hassan et al, respectively. Rates of late preterm birth (34-37 weeks) were not compared between treatment groups in the trial performed by Fonseca et al, but rates of preterm birth <37 weeks were not different in the vaginal progesterone and placebo groups in the trial performed by Hassan et al. Based on these same trials, the number of ultrasound examinations that will need to be performed and the number of women with short cervical lengths who will need to be treated to prevent 1 preterm birth are presented in Table 2 . These estimates are based on the large cohorts screened in the trials performed by Fonseca et al and Hassan et al, but estimates might be different in populations with different characteristics (eg, race, access to prenatal care).
| Trial | Rate of short CL | Women with short CL | Rate of PTB (daily P) | Rate of PTB (placebo) | No. of PTBs (daily P) | No. of PTBs (placebo) | PTBs prevented with daily P |
|---|---|---|---|---|---|---|---|
| Fonseca et al | 1.7% | 170 | 19.2% | 34.4% | 33 | 58 | 25 |
| Hassan et al | 2.28% | 228 | 8.9% | 16.1% | 20 | 37 | 17 |
| Trial | No. of US studies to prevent 1 PTB | No. of women treated with P to prevent 1 PTB |
|---|---|---|
| Fonseca et al | 10,000/25 a = 400 | 170/25 a = 7 |
| Hassan et al | 10,000/17 b = 588 | 228/17 b = 13.4 |
a In hypothetical cohort of 10,000 women, 25 preterm births <34 wk will be prevented by universal cervical length screening and treatment with daily progesterone according to trial performed by Fonseca et al ;
b in hypothetical cohort of 10,000 women, 17 preterm births <33 wk will be prevented by universal cervical length screening and treatment with daily progesterone according to trial performed by Hassan et al.
Several subset analyses were performed in the trials published by Fonseca et al and Hassan et al. In the earlier trial, the risk of spontaneous preterm delivery in the current pregnancy among women with no history of preterm birth was 31.2% (34/109) in the placebo group vs 17.9% (20/112) in daily progesterone group (RR, 0.57; P = .03). In the trial by Hassan et al, the risk of preterm delivery in the current pregnancy among women with no history of preterm birth was 15.3% (29/189) in the placebo group vs 7.6% (15/197) in daily progesterone group (RR, 0.50; P = .02). In the trial performed by Fonseca et al, the risk of spontaneous preterm delivery among women with singleton pregnancies was 32.1% (36/112) in the placebo group vs 17.5% (20/114) in the daily progesterone group (RR, 0.54; P = .02). Women with multiple pregnancies were not enrolled in the trial performed by Hassan et al, so all analyses were exclusive to singleton pregnancies. Unfortunately, subset analyses exclusive to nulliparous subjects were not presented in either trial.
In the trial conducted by Fonseca et al, 25.8% of all spontaneous deliveries <34 weeks occurred in women whose cervical length measured ≤1.5 cm (23,795 women with data on pregnancy outcomes, 489 spontaneous preterm births <34 weeks in the overall cohort, and 126 spontaneous preterm births <34 weeks in women with cervical lengths ≤1.5 cm). Although the overall preterm birth rate (2.1%) in this study appears low, it should be noted that iatrogenic preterm deliveries and preterm deliveries occurring at 34-37 weeks were not included in the primary outcome. In this trial, 1.7% of women had a cervical length of ≤1.5 cm, and some were randomized to daily progesterone or placebo while others were not enrolled. Based on the spontaneous preterm birth rates of women with a short cervix treated with daily progesterone (19.2%) vs placebo (34.4%), the overall rate of birth <34 weeks would be 1.85% if all women with a cervical length <1.5 cm received daily progesterone vs 2.11% if universal cervical length screening was not performed and women with short cervical lengths were not treated with daily progesterone ( Figure ). Preterm deliveries for their overall cohort were not provided by Hassan et al, so similar analyses could not be performed for that study.
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
