Universal cervical length screening–saving babies lives

Images: Courtesy of Perinatal Quality Foundation

Transvaginal ultrasound (TVU) cervical length (CL) screening for prediction and prevention of spontaneous preterm birth (SPTB) is among the most transformative clinical changes in obstetrics in the last decades. TVU CL screening should now be offered to all pregnant women: hence the appellative ‘universal CL screening.’

TVU CL screening is an excellent screening test for several reasons. It screens for SPTB, which is a clinically important, well-defined disease whose prevalence and natural history is known, and has an early recognizable asymptomatic phase in CL shortening detected by TVU. TVU CL screening is a well-described technique, safe and acceptable, with a reasonable cutoff (25 mm) now identified for all populations, and results are reproducible and accurate. There are hundreds of studies proving these facts. In the last 10 years, TVU measurement of CL as a screening test has been accepted^1,2: it identifies women at risk for SPTB, and an early intervention (progesterone or cerclage depending on the clinical situation) is effective in preventing SPTB. Screening and treatment of short cervix is cost-effective and readily available as an early intervention (progesterone or cerclage depending on the clinical situation), is effective in preventing the outcome (SPTB), treating abnormal results is cost-effective, and facilities for screening are available and treatments are readily available.^3–5 It is also important to emphasize that CL screening for prevention of SPTB should be done by TVU, and not by transabdominal ultrasound.⁶It is best to review TVU CL screening by populations: singletons without prior SPTB, singletons with prior SPTB, and twins (Table).

Related Article:
Can transabdominal ultrasound exclude short cervix?

Singletons without prior SPTB

Women with no previous SPTB who are carrying a singleton pregnancy is the population in which TVU CL could have the greatest impact on decreasing SPTB, for several reasons:

1. Up to 60% to 90% of SPTB occur in this population.
2. More than 90% of these women have risk factors for SPTB.^7,8
3. Vaginal progesterone has been associated with a significant 39% decrease in PTB at <33 weeks of gestation and a significant 38% decrease in perinatal morbidity and mortality in a meta-analysis of randomized controlled trials (RCTs) including 606 women without prior PTB.^9,10
4. Cost-effectiveness studies have shown that TVU CL screening in this specific population prevents thousands of preterm births, saves or improves from death or major morbidity 350 babies’ lives annually, and saves approximately $320,000 per year in the US alone.³ These numbers may be even higher now as the TVU CL cutoff for offering vaginal progesterone has moved in many centers from ≤20 mm to ≤25 mm, including more women (from about 0.8% to about 2% to 3%, respectively¹¹) who benefit from screening.
5. Real-world implementation studies have indeed shown significant decreases in SPTB when a policy of universal TVU CL screening in this specific population is implemented.^12,13

Universal TVU CL screening recently called into question

In a recent article published in the Journal of the American Medical Association,¹⁴ TVU CL screening in this population, in particular for nulliparous women, has come under interrogation. The authors found only an 8% sensitivity of TVU CL screening for SPTB using a cutoff of ≤25 mm at 16 0/7 to 22 6/7 weeks of gestation in 9,410 nulliparous women. This result is different compared with other previous cohort studies in this area, however, and is likely related to a number of issues in the methodology.

First, TVU CL screening was done in many women at too early a gestational age. The earlier the CL screening, the lower the sensitivity of the procedure. Data at 16 and 17 weeks of gestation should have been excluded, as almost all RCTs and other studies on universal TVU CL screening in this population recommended doing screening at about 18 0/7 to 23 6/7 weeks.

Second, women with TVU CL <15 mm received vaginal progesterone. This would decrease the incidence of PTB and, therefore, sensitivity.

Third, outcomes data were not available for 469 women and, compared with women analyzed, these women were at higher risk for SPTB as they were more likely to be aged 21 years or younger, black, with less than a high school education, and single, all significant risk factors for SPTB. (Not all risk factors for SPTB were reported in this study.)

Fourth, pregnancy losses before 20 weeks were excluded, and these could have been early SPTB; therefore, the sensitivity could have been decreased if women with this outcome were excluded.

Fifth, prior studies have shown that TVU CL screening in singletons without prior SPTB has a sensitivity of about 30% to 40%.^15,16 In nulliparas, the sensitivity of TVU CL ≤20 mm had been reported previously to be 20%.¹⁶ Additional data from 2012–2014 at our institution demonstrate that the incidence of CL ≤25 mm is about 2.8% in nulliparous women, with a sensitivity of 19.5% for SPTB <37 weeks. These numbers show again that 8% sensitivity was low in the JAMA study¹⁴ due the shortcomings we just highlighted. Furthermore, the reported sensitivity of TVU CL ≤25 mm for PTB <32 weeks was 24% in Esplin and colleagues’ study,¹⁴ while 60% in our data. Given that early preterm births are the most significant source of neonatal morbidity and mortality, women with a singleton gestation and no prior SPTB but with a short TVU CL are perhaps the most important subgroup to identify.

Sixth, a low sensitivity in and of itself is not reflective of a poor screening test. We have known for a long time that SPTB has many etiologies. No one screening test, and no one intervention, would independently prevent all SPTBs. In a population that accounts for more than half of PTBs and for whom no other screening test has been found to be effective, much less cost effective, it is important not to cast aside the dramatic potential clinical benefit to TVU CL screening.

Related Article:
A stepwise approach to cervical cerclage