Cost-effectiveness of universal routine sonographic cervical-length measurement at 19 to 25 weeks’ gestation

Abstract

National second-trimester scanning of cervical length was introduced in Israel in 2010, and in the decade thereafter, a significant systematic reduction in preterm birth and in the delivery of low birthweight babies was found among singletons. In this study, we sought to estimate the cost-effectiveness of a national policy mandating second-trimester cervical length screening by ultrasound, followed by vaginal progesterone treatment for short cervical length in comparison with no screening strategy. We constructed a decision model comparing 2 strategies, namely (1) universal cervical length screening, and (2) no screening strategy. This study used the national delivery registry of Israel's Ministry of Health. All women diagnosed with a second-trimester cervical length <25 mm were treated with vaginal progesterone and were monitored with a bimonthly ultrasound scan for cervical dynamics and threat of early delivery. Preterm birth prevalence associated with short cervical length, the efficacy of progesterone in preterm birth prevention, and the accuracy of cervical length measurements were derived from previous studies. The cost of progesterone and bimonthly sonographic surveillance, low birthweight delivery, newborn admission to intensive care units, the first-year costs of managing preterm birth and low birthweight, and instances of handicaps and the cost of their follow-up were extracted from the publicly posted registry of Israel's Ministry of Health and Israel Social Securities data. Monte Carlo simulations decision tree mode, Tornado diagrams, and 1- and 2-way sensitivity analyses were implemented and the base case and sensitivity to parameters that were predicted to influence cost-effectiveness were calculated. Without cervical length screening, the discounted quality-adjusted life years were 30.179, and with universal cervical length screening, it increased to 30.198 (difference of 0.018 quality-adjusted life years). The average cost of no screening for cervical length strategy was $1047, and for universal cervical length screening, it was reduced to $998. The calculated incremental cost-effectiveness ratio was -$2676 per quality-adjusted life year (dividing the difference in costs by the difference in quality-adjusted life years). Monte Carlo simulation of cervical length screening of 170,000 singleton newborns (rounded large number close to the number of singleton newborns in Israel) showed that 95.17% of all babies were delivered at gestational week ≥37 in comparison with 94.46% of babies with the no screening strategy. Given 170,000 singleton births, the national savings of screening for short cervical length when compared with no cervical length screening amounted to $8.31M annually, equating to $48.84 for a base case, and the incremental cost-effectiveness ratio for each case of low birthweight or very low birthweight avoided was -$14,718. A cervical length <25 mm was measured for 30,090 women, and of those, 24,650 were false positives. The major parameters that affected the incremental cost-effectiveness ratio were the incidence of preterm birth, the specificity of cervical length measurements, and the efficacy of progesterone treatment. At a preterm birth incidence of <3%, universal screening does not lead to a cost saving. National universal cervical length screening should be incorporated into the routine anomaly scan in the second trimester, because it leads to a drop in the incidence of preterm birth and low birthweight babies in singleton pregnancies, thereby saving costs related to the newborn and gaining quality-adjusted life years.