An open randomised trial of the Arabin pessary to prevent preterm birth in twin pregnancy - STOPPIT - 2

DESIGN. A multicentre open randomised trial of the Arabin cervical pessary versus conventional treatment in women with twin pregnancy and a short cervix with two phases for each participant (i) a SCREENING phase to determine cervical length and (ii) a TREATMENT phase in which those with a short cervix (less than or equal to 35mm) are randomized to the Arabin pessary or standard care. SETTING. Antenatal clinics in the UK NHS. TARGET POPULATION. Women with twin pregnancy at less than 20+6 weeks gestation. HEALTH TECHNOLOGIES BEING ASSESSED. The cervical (Arabin) pessary, which appears effective in preventing preterm birth in singletons (Goya et al. 2012) and whose efficacy in twins has been suggested in a subgroup analysis of the ProTwin trial (Liem 2013). There are no alternative therapies (drug or non-drug) that prevent preterm birth in twins. MEASUREMENTS OF COSTS. A cost-effectiveness analysis of the Arabin pessary will be undertaken from the perspective of the UK NHS. MEASUREMENT OF OUTCOMES: The primary obstetric outcome is spontaneous onset of labour leading to delivery before 34+0 weeks gestation, measured at the mother level. The primary neonatal outcome is a composite of stillbirth, periventricular leucomalacia, early respiratory morbidity, intraventricular haemorrhage, necrotising enterocolitis, proven sepsis, and neonatal death, on either or both twins, very similar to the primary neonatal outcome in ProTwin (Liem 2013). Secondary outcomes will include acceptability, health economics, and outcomes likely to facilitate meta-analyses (Alfirevic 2014) – and will be informed by work ongoing under the COMET initiative (PI Duley). We will also calculate preterm birth before 34 weeks gestation for a variety of cervical lengths. SAMPLE SIZE: 2500 having consented to ultrasound screening and 500 randomised 1:1 to Arabin pessary or standard care there is 89% power at the 5% significance level to detect a difference in the primary obstetric outcome between the groups, assuming standard care event rate of 32% in the short cervix group and an (Arabin) treatment effect (RR) of 0.6. For the primary neonatal outcome, the study has 88% power for a treatment effect (RR). TIMETABLE. Set up - 4 months, pilot recruitment over 10 months / 28 centres (112 centre-months recruitment with randomization of 51 women – see case for support for assumptions) initiation of remaining 30 centres over next 14 months, during the recruitment period of a further 36 months, final outcomes collected - 7 months, data cleaning - 1 month, analysis - 2 months. EXPERTISE IN TEAM: Twin pregnancy and prevention of preterm birth, clinical trials, neonatal outcomes, health economics, social science (acceptability) and PPI (each of leading consumer organizations and regional level PPI groups). The team have experience of conducting large multicentre pregnancy studies in the UK (eg STOPPIT, OPPTIMUM and EMPOWaR – all whose recruitment targets have now been achieved).