Pessary or Progesterone to Prevent Preterm delivery in women with short cervical length
Go to the Quadruple P website
Preterm birth is in quantity and in severity the most important issue in obstetric care in the developed world. Progestagens and cervical pessaries are both considered as potential preventive treatments. We aim to compare the effectiveness of vaginal progesterone and cervical pessary in the prevention of preterm birth in women with singleton and multiple pregnancies and a short cervix.
Nationwide open-label multicenter randomized clinical trial with an economic analysis alongside it.
Women with a singleton undergoing fetal assessment at 18-22 weeks will be offered cervical length measurement; for women with a multiple pregnancy, this happens at 16-22 weeks. Women with a short cervix (singleton pregnancy 35 mm or less [11.5th percentile], multiple pregnancy below 38 mm [25th percentile], will be invited to participate in a randomised clinical trial.
Daily progesterone 200mg will be self-administered vaginally. The silicone cervical pessary will be placed between 16-22 weeks. Both interventions will be applied until 36 weeks gestation or until delivery, whatever comes first.
Primary outcome will be a composite morbidity and mortality rate of children in the two groups. This composite morbidity rate contains the following variables: severe Respiratory Distress Syndrome (RDS), Broncho Pulmonal Dysplasia (BPD), Intraventricular Haemorrhage II B or worse, Necrotizing Enterocolitis (NEC), proven sepsis, stillbirth and death before discharge from the nursery. They will be measured until 10 weeks after the expected term date. Secondary outcomes will be time to delivery, preterm birth rate before 28, 32 and 37 weeks, days of admission in neonatal intensive care unit, maternal morbidity, maternal admission days for preterm labour, and long term effects.
Singleton pregnancies: With our previous experience in the Triple P trial we know that screening 850 low risk women with a singleton pregnancy a month is feasible. With this study we need to screen approximately 200 women a month to reach our sample size. Screening 200 women a month will result in 10.000 screened women in 4 years. Taking the 11.5th percentile of short cervical length (35mm or less) will result in 1150 potential eligible women for our study. To reach our final sample size of 620 women we need a participation rate of 50 to 60%.
Multiple pregnancies: With our previous experience from the AMPHIA and the ProTwin trial we expect to screen 3.600 women in 36 months. Taking the 25th percentile of cervical length (38mm) will result in 900 eligible women for the trial. Assuming a 40% participationrate (from AMPHIA and ProTwin trial we know that women with a multiple pregnancy are very likely to participate in a trial) we can randomize 332 women with a multiple pregnancy.
Summarizing the numbers described above this will lead to:
- 628 randomisations in low risk singletons after screening 10.000
- 332 randomisations in multiples after screening 3.600
Total: 960 randomizations after screening 13.600 women
For each strategy, we will calculate the risk of poor neonatal outcome as well as the average costs. If screening and progesterone and /or cervical pessary treatment lowers the risk of poor neonatal outcome, we will calculate the incremental cost-effectiveness ratio as costs per prevented case of poor neonatal outcome. If the results of the RCT demonstrate differences in neonatal outcome rates between the strategies, a long term projection of downstream costs and outcome will be performed. Whereas medical costs associated with preterm infants have been well described, data concerning long-term developmental and educational costs are scarcer. The cost-effectiveness analysis will assess the balance between number needed to treat and number needed to screen (fixed). The number of women needed to treat with a short cervix will obviously be much smaller when a cut-off value of 15mm is used than when the group with a cervix between 15 and 25mm is considered to be abnormal. These type of analyses have also been performed in the triple P study.
We will perform subgroup analyses based on ethnicity and other patient characteristics, as both the screening results as well as the treatment effect could differ between these groups. We will also assess the effects and costs of screening strategies with different cut-off levels. However we are not able to foresee whether the treatment effect differs across other subgroups. The sensitivity of the results (costs and health outcomes) for various assumptions and parameter estimates is tested in sensitivity analyses and visualised in ICER-graphs and cost-effectiveness acceptability curves.
We will need a run-in period of 3 months for the study set up, and 36 months for inclusion. After inclusion of the last patient, 6 months (additional pregnancy) will be needed for data collection and report of results. The first report on the primary outcome is expected at 4 years after the start of the study.
Prof. dr. E. Pajkrt, department of obstetrics and gynaecology, AMC (Amsterdam)
M.D. van Zijl, AMC