Abstract: Postal questionnaires are widely used to collect data in healthcare research but a poor response rate may reduce the validity and reliability of results. There was a lack of evidence available relating to use of a monetary incentive to improve the response rate in the healthcare setting. Methods The MRC ORACLE Children Study is assessing the health and development of nearly 9000 seven year old children whose mothers' joined the MRC ORACLE Trial. We carried out a randomised controlled trial of inclusion of monetary incentive (five pound voucher redeemable at many high street stores) with the reminder questionnaire to parents. This trial took place between April 2002 and November 2003. When the parents were sent the reminder questionnaire about their child's health and development they were randomly assigned by concealed computer-generated allocation stratified by week of birthday to receive a five pound voucher or no incentive. The population were 722 non-responders to the initial mailing of a 12-page questionnaire. Main outcome measures: Difference in response rate between the two groups. Results Inclusion of the voucher with the reminder questionnaire resulted in a 11.7%(95% CI 4.7% to 18.6%) improvement in the response rate between the two groups. Conclusion This improvement in response rate and hence the validity and reliability of results obtained appears to be justified ethically and financially.

Abstract: Studies of prematurely born infants contain a relatively large percentage of multiple births, so the resulting data have a hierarchical structure with small clusters of size 1, 2 or 3. Ignoring the clustering may lead to incorrect inferences. The aim of this study was to compare statistical methods which can be used to analyse such data: generalised estimating equations, multilevel models, multiple linear regression and logistic regression. Four datasets which differed in total size and in percentage of multiple births ( n  = 254, multiple 18%; n  = 176, multiple 9%; n  = 10 098, multiple 3%; n  = 1585, multiple 8%) were analysed. With the continuous outcome, two‐level models produced similar results in the larger dataset, while generalised least squares multilevel modelling (ML GLS ‘xtreg’ in Stata) and maximum likelihood multilevel modelling (ML MLE ‘xtmixed’ in Stata) produced divergent estimates using the smaller dataset. For the dichotomous outcome, most methods, except generalised least squares multilevel modelling (ML GH ‘xtlogit’ in Stata) gave similar odds ratios and 95% confidence intervals within datasets. For the continuous outcome, our results suggest using multilevel modelling. We conclude that generalised least squares multilevel modelling (ML GLS ‘xtreg’ in Stata) and maximum likelihood multilevel modelling (ML MLE ‘xtmixed’ in Stata) should be used with caution when the dataset is small. Where the outcome is dichotomous and there is a relatively large percentage of non‐independent data, it is recommended that these are accounted for in analyses using logistic regression with adjusted standard errors or multilevel modelling. If, however, the dataset has a small percentage of clusters greater than size 1 (e.g. a population dataset of children where there are few multiples) there appears to be less need to adjust for clustering.

Abstract: Cervical cerclage is a recognised treatment to prevent late miscarriage and pre-term birth (PTB). Emergency cervical cerclage (ECC) for cervical dilatation with exposed unruptured membranes is less common and the potential benefits of cerclage are less certain. A randomised control trial is needed to accurately assess the effectiveness of ECC in preventing pregnancy loss compared to an expectant approach. Methods C-STICH2 is a multicentre randomised controlled trial in which women presenting with cervical dilatation and unruptured exposed membranes at 16 + 0 to 27 + 6 weeks gestation are randomised to ECC or expectant management. Trial design includes 18 month internal pilot with embedded qualitative process evaluation, minimal data set and a within-trial health economic analysis. Inclusion criteria are ≥16 years, singleton pregnancy, exposed membranes at the external os, gestation 16 + 0–27 + 6 weeks, and informed consent. Exclusion criteria are contraindication to cerclage, cerclage in situ or previous cerclage in this pregnancy. Randomisation occurs via an online service in a 1:1 ratio, using a minimisation algorithm to reduce chance imbalances in key prognostic variables (site, gestation and dilatation). Primary outcome is pregnancy loss; a composite including miscarriage, termination of pregnancy and perinatal mortality defined as stillbirth and neonatal death in the first week of life. Secondary outcomes include all core outcomes for PTB. Two-year development outcomes will be assessed using general health and Parent Report of Children’s Abilities-Revised (PARCA-R) questionnaires. Intended sample size is 260 participants (130 each arm) based on 60% rate of pregnancy loss in the expectant management arm and 40% in the ECC arm, with 90% power and alpha 0.05. Analysis will be by intention-to-treat. Discussion To date there has been one small trial of ECC in 23 participants which included twin and singleton pregnancies. This small trial along with the largest observational study ( n = 161) found ECC to prolong pregnancy duration and reduce deliveries before 34 weeks gestation. It is important to generate high quality evidence on the effectiveness of ECC in preventing pregnancy loss, and improve understanding of the prevalence of the condition and frequency of complications associated with ECC. An adequately powered RCT will provide the highest quality evidence regarding optimum care for these women and their babies. Trial registration ISRCTN Registry ISRCTN12981869 . Registered on 13th June 2018.

Abstract: Evidence from the Birthplace in England Research Programme supported a policy of offering ‘low risk’ women a choice of birth setting, but a number of unanswered questions remained. Aims This project aimed to provide further evidence to support the development and delivery of maternity services and inform women’s choice of birth setting: specifically, to explore maternal and organisational factors associated with intervention, transfer and other outcomes in each birth setting in ‘low risk’ and ‘higher risk’ women. Design Five component studies using secondary analysis of the Birthplace prospective cohort study (studies 2–5) and ecological analysis of unit/NHS trust-level data (studies 1 and 5). Setting Obstetric units (OUs), alongside midwifery units (AMUs), freestanding midwifery units (FMUs) and planned home births in England. Participants Studies 1–4 focused on ‘low risk’ women with ‘term’ pregnancies planning vaginal birth in 43 AMUs ( n  = 16,573), in 53 FMUs ( n  = 11,210), at home in 147 NHS trusts ( n  = 16,632) and in a stratified, random sample of 36 OUs ( n  = 19,379) in 2008–10. Study 5 focused on women with pre-existing medical and obstetric risk factors (‘higher risk’ women). Main outcome measures Interventions (instrumental delivery, intrapartum caesarean section), a measure of low intervention (‘normal birth’), a measure of spontaneous vaginal birth without complications (‘straightforward birth’), transfer during labour and a composite measure of adverse perinatal outcome (‘intrapartum-related mortality and morbidity’ or neonatal admission within 48 hours for 〉  48 hours). In studies 1 and 3, rates of intervention/maternal outcome and transfer were adjusted for maternal characteristics. Analysis We used (a) funnel plots to explore variation in rates of intervention/maternal outcome and transfer between units/trusts, (b) simple, weighted linear regression to evaluate associations between unit/trust characteristics and rates of intervention/maternal outcome and transfer, (c) multivariable Poisson regression to evaluate associations between planned place of birth, maternal characteristics and study outcomes, and (d) logistic regression to investigate associations between time of day/day of the week and study outcomes. Results Study 1 – unit-/trust-level variations in rates of interventions, transfer and maternal outcomes were not explained by differences in maternal characteristics. The magnitude of identified associations between unit/trust characteristics and intervention, transfer and outcome rates was generally small, but some aspects of configuration were associated with rates of transfer and intervention. Study 2 – ‘low risk’ women planning non-OU birth had a reduced risk of intervention irrespective of ethnicity or area deprivation score. In nulliparous women planning non-OU birth the risk of intervention increased with increasing age, but women of all ages planning non-OU birth experienced a reduced risk of intervention. Study 3 – parity, maternal age, gestational age and ‘complicating conditions’ identified at the start of care in labour were independently associated with variation in the risk of transfer in ‘low risk’ women planning non-OU birth. Transfers did not vary by time of day/day of the week in any meaningful way. The duration of transfer from planned FMU and home births was around 50–60 minutes; transfers for ‘potentially urgent’ reasons were quicker than transfers for ‘non-urgent’ reasons. Study 4 – the occurrence of some interventions varied by time of the day/day of the week in ‘low risk’ women planning OU birth. Study 5 – ‘higher risk’ women planning birth in a non-OU setting had fewer risk factors than ‘higher risk’ women planning OU birth and these risk factors were different. Compared with ‘low risk’ women planning home birth, ‘higher risk’ women planning home birth had a significantly increased risk of our composite adverse perinatal outcome measure. However, in ‘higher risk’ women, the risk of this outcome was lower in planned home births than in planned OU births, even after adjustment for clinical risk factors. Conclusions Expansion in the capacity of non-OU intrapartum care could reduce intervention rates in ‘low risk’ women, and the benefits of midwifery-led intrapartum care apply to all ‘low risk’ women irrespective of age, ethnicity or area deprivation score. Intervention rates differ considerably between units, however, for reasons that are not understood. The impact of major changes in the configuration of maternity care on outcomes should be monitored and evaluated. The impact of non-clinical factors, including labour ward practices, staffing and skill mix and women’s preferences and expectations, on intervention requires further investigation. All women planning non-OU birth should be informed of their chances of transfer and, in particular, older nulliparous women and those more than 1 week past their due date should be advised of their increased chances of transfer. No change in the guidance on planning place of birth for ‘higher risk’ women is recommended, but research is required to evaluate the safety of planned AMU birth for women with selected relatively common risk factors. Funding The National Institute for Health Research Health Services and Delivery Research programme.

Abstract: In women with late preterm pre-eclampsia (i.e. at 34 +0 to 36 +6 weeks’ gestation), the optimal delivery time is unclear because limitation of maternal–fetal disease progression needs to be balanced against infant complications. The aim of this trial was to determine whether or not planned earlier initiation of delivery reduces maternal adverse outcomes without substantial worsening of perinatal or infant outcomes, compared with expectant management, in women with late preterm pre-eclampsia. Methods We undertook an individually randomised, triple non-masked controlled trial in 46 maternity units across England and Wales, with an embedded health economic evaluation, comparing planned delivery and expectant management (usual care) in women with late preterm pre-eclampsia. The co-primary maternal outcome was a maternal morbidity composite or recorded systolic blood pressure of ≥ 160 mmHg (superiority hypothesis). The co-primary short-term perinatal outcome was a composite of perinatal deaths or neonatal unit admission (non-inferiority hypothesis). Analyses were by intention to treat, with an additional per-protocol analysis for the perinatal outcome. The primary 2-year infant neurodevelopmental outcome was measured using the PARCA-R (Parent Report of Children’s Abilities-Revised) composite score. The planned sample size of the trial was 900 women; the trial is now completed. We undertook two linked substudies. Results Between 29 September 2014 and 10 December 2018, 901 women were recruited; 450 women [448 women (two withdrew consent) and 471 infants] were allocated to planned delivery and 451 women (451 women and 475 infants) were allocated to expectant management. The incidence of the co-primary maternal outcome was significantly lower in the planned delivery group [289 (65%) women] than in the expectant management group [338 (75%) women] (adjusted relative risk 0.86, 95% confidence interval 0.79 to 0.94; p  = 0.0005). The incidence of the co-primary perinatal outcome was significantly higher in the planned delivery group [196 (42%) infants] than in the expectant management group [159 (34%) infants] (adjusted relative risk 1.26, 95% confidence interval 1.08 to 1.47; p  = 0.0034), but indicators of neonatal morbidity were similar in both groups. At 2-year follow-up, the mean PARCA-R scores were 89.5 points (standard deviation 18.2 points) for the planned delivery group (290 infants) and 91.9 points (standard deviation 18.4 points) for the expectant management group (256 infants), both within the normal developmental range (adjusted mean difference –2.4 points, 95% confidence interval –5.4 to 0.5 points; non-inferiority p  = 0.147). Planned delivery was significantly cost-saving (–£2711, 95% confidence interval –£4840 to –£637) compared with expectant management. There were nine serious adverse events in the planned delivery group and 12 in the expectant management group. Conclusion In women with late preterm pre-eclampsia, planned delivery reduces short-term maternal morbidity compared with expectant management, with more neonatal unit admissions related to prematurity but no indicators of greater short-term neonatal morbidity (such as need for respiratory support). At 2-year follow-up, around 60% of parents reported follow-up scores. Average infant development was within the normal range for both groups; the small between-group mean difference in PARCA-R scores is unlikely to be clinically important. Planned delivery was significantly cost-saving to the health service. These findings should be discussed with women with late preterm pre-eclampsia to allow shared decision-making on timing of delivery. Limitations Limitations of the trial include the challenges of finding a perinatal outcome that adequately represented the potential risks of both groups and a maternal outcome that reflects the multiorgan manifestations of pre-eclampsia. The incidences of maternal and perinatal primary outcomes were higher than anticipated on the basis of previous studies, but this did not limit interpretation of the analysis. The trial was limited by a higher loss to follow-up rate than expected, meaning that the extent and direction of bias in outcomes (between responders and non-responders) is uncertain. A longer follow-up period (e.g. up to 5 years) would have enabled us to provide further evidence on long-term infant outcomes, but this runs the risk of greater attrition and increased expense. Future work We identified a number of further questions that could be prioritised through a formal scoping process, including uncertainties around disease-modifying interventions, prognostic factors, longer-term follow-up, the perspectives of women and their families, meta-analysis with other studies, effect of a similar intervention in other health-care settings, and clinical effectiveness and cost-effectiveness of other related policies around neonatal unit admission in late preterm birth. Trial registration The trial was prospectively registered as ISRCTN01879376. Funding This project was funded by the National Institute for Health and Care Research (NIHR) Health Technology Assessment programme and will be published in Health Technology Assessment . See the NIHR Journals Library website for further project information.