Aims and objectives
The aims and objectives of the C-STICH2 trial are presented in Table 1.
Table 1
Description of the aims and objectives of C-STICH2 internal pilot and full trial
|
Aim: To evaluate whether ECC can improve outcomes for mothers and babies’ in women who present with cervical dilatation and exposed unruptured fetal membranes.
|
|
Pilot Objectives
|
• To ascertain if the trial and trial processes are acceptable to women, including the ability to recruit and randomise women.
• To assess whether the event rate of the primary outcome is compatible with the estimate used in the sample size calculation.
• To explore if clinicians are in equipoise and willing to randomise to an RCT.
|
|
Full Trial Objectives
|
• To determine whether an ECC reduces pregnancy loss (miscarriage, termination of pregnancy, stillbirth or neonatal death within 7 days of delivery) in women who present with cervical dilatation and exposed, unruptured fetal membranes between 16 + 0 and 27 + 6 weeks.
• To follow up all surviving babies to 2 years of corrected age to determine their general health and medium term neurodevelopmental outcomes.
• To determine the complication rates of ECC: iatrogenic rupture of membranes during the procedure; insertion failure; predictors of successful ECC placement such as magnitude of dilatation
|
Study design and setting
C-STICH2 is an open, multicentre, superiority, RCT in the United Kingdom. There is an internal pilot with an embedded qualitative process evaluation. The full trial includes a within-trial economic analysis and a minimal dataset of anonymised demographic and pregnancy outcome data from all women presenting with the condition who do not consent to participate in the randomised arms of C-STICH2.
Participants
Women will be considered eligible for C-STICH2 if they are aged 16 years or over, with a singleton pregnancy and in whom the cervix is found to be dilated with the fetal membranes intact and exposed, at or below the level of the external os, based on judgement. This may be diagnosed on speculum examination or by ultrasound finding, if performed by a suitably trained practitioner. The gestational age of the pregnancy must be 16 + 0 to 27 + 6 weeks based on the best available estimate and the woman must be able to provide written informed consent. The exclusion criteria are as follows: contra-indication to ECC as judged by the responsible clinician; if the participant has already received cervical cerclage of any type in this pregnancy, or has a cerclage in situ from a previous pregnancy.
Participant enrolment
Participants will be identified following clinical presentation to maternity assessment units, or women may be identified through screening in a specialist PTB clinic setting (Fig. 1). Women who meet the eligibility criteria should be approached by the most senior clinician available about participating in the trial. It is recognised that this may be a very difficult time for women and their families and sufficient time will be given for women to consider participation and explore options. The clinician should be familiar with counselling women at risk of preterm delivery and have received training from the trial team or principal investigator. A detailed participant information sheet (PIS) will be provided. If the woman agrees to participate, written consent must be obtained prior to randomisation. Participants must understand that they are free to withdraw from the trial at any time and that this will not affect their subsequent care.
In women who decline to participate in the randomised arms of C-STICH2 or who are not approached regarding the study, a non-consented anonymised minimal dataset will be collected. This dataset includes basic demographics and pregnancy outcome data to inform overall prevalence of the condition and incidence of pregnancy loss.
Qualitative process evaluation
All women approached about the trial should be informed about the qualitative process evaluation, even if they have chosen to decline trial participation. They will receive written information about participation in the qualitative study either at the time of presentation and trial approach, shortly after at a time considered appropriate by their healthcare professionals (HCPs), at a follow up appointment, or via a follow up letter. Women who express an interest in participating will, where feasible, be asked to complete and sign a consent to contact form giving written permission to be contacted by the qualitative research team. Women who are sent a follow up letter about the qualitative study will be asked to return a completed consent to contact form or will be able to contact the research team directly. Following receipt of the consent to contact form or direct contact, the qualitative team will assess suitability of contact timing with the recruiting site. If this is deemed to be appropriate, the team will then discuss participation with the woman and separate informed consent (written, electronically completed or verbal) for the qualitative study will be obtained. The qualitative process evaluation will use semi-structured interviews to explore women’s experiences of diagnosis and care, and their views of the trial processes irrespective of their decision to participate in C-STICH2.
HCPs named on the site delegation logs will also be approached directly by the qualitative research team to participate in semi-structured interviews to help inform our understanding of the following: issues approaching women to participate, views on barriers and facilitators to trial recruitment, thoughts and experiences on the use of ECC including personal and community equipoise. All participants will complete a demographic questionnaire to inform sampling and subsequent description of participant characteristics.
Women and HCPs will be able to choose how they participate in an interview (e.g. face to face, via phone, or via video calling). Interviews will be digitally-audio recorded with data collection and initial analysis taking place iteratively (Dicciccio-Bloom 2006). We anticipate undertaking up to 50 interviews in total but the numbers will remain flexible to ensure that we collect sufficiently rich data to address the aims and objectives of the study (Malterud 2015). Audio files will be transcribed clean verbatim by a specialist transcription company and the framework approach (Gale 2013) used to facilitate a systematic and flexible approach to the analysis.
The findings from the qualitative process evaluation will dynamically inform trial design both during and subsequent to the pilot (O'Cathain 2018). The embedded qualitative findings will influence trial processes moving beyond the pilot study, inform trial steering committee (TSC) recommendations and help to maximise recruitment and retention in this challenging setting.
Randomisation and blinding
After eligibility has been confirmed and informed consent obtained, women can be randomised into the C-STICH2 trial (Fig. 1). Randomisation will be provided by a secure 24/7 online randomisation system and a back-up free telephone randomisation service (available during working hours). Randomisation can be completed by any staff member on the delegation log assigned to do so. Participants will be randomised in a 1:1 ratio to either ECC or expectant management. A minimisation algorithm will be used to ensure balance in the treatment allocation by site, gestation (16 + 0–19 + 6 / 20 + 0–23 + 6 / 24 + 0–27 + 6 weeks) and cervical dilatation (≤3cm / ≥4cm / fully dilated-minimal cervix felt). A ‘random element’ will be included in the algorithm, so that each participant has a probability (unspecified here) of being randomised to the opposite treatment that they would have otherwise received.
It is not possible to blind women or clinicians to the intervention they have received as the comparison is between a surgical intervention and expectant management. The primary outcome (miscarriage, stillbirth, termination of pregnancy and neonatal death) is, however, objective and thus should minimise bias.
Trial interventions
This trial will be pragmatic in nature allowing the use of treatment adjuncts such as antibiotics or indomethacin in both the ECC and expectant management arms following randomisation.
For women allocated to ECC, this should be delivered at a time that is felt to be clinically appropriate and within 72 hours of randomisation. An ECC will involve, under appropriate anaesthesia, the replacement of the fetal membranes into the uterine cavity and the placement of a “purse string” cervical cerclage around the body of the cervix, aiming to occlude the cervical canal and prevent further prolapse of the fetal membranes. Pre-operative, operative, and post-operative management is at the discretion of the clinician responsible for the woman.
Women allocated to expectant management will be will be managed pragmatically according to local protocols. If an ECC is placed this will be considered a protocol deviation.
Follow-up
Surviving children will be followed up at two years of corrected age. On discharge from hospital, consent for long-term follow-up will be confirmed and contact details for the main carer will be collected. The research team will maintain contact with families of surviving children at intervals to minimise loss to follow-up. At two years corrected age, the main caregiver will be sent questionnaires, this will include general health, Parent Report of Children's Abilities-Revised (PARCA-R ) questionnaire (Johnson 2008, Martin 2013) and an assessment of motor outcomes.
Adverse Event Reporting
Adverse outcomes, both maternal and neonatal, are common in this high-risk obstetric population. Adverse events (AEs) will be reported via case report forms (CRF) and captured via pre-defined outcome measures. Serious adverse events (SAEs) including but not limited to maternal admission requiring care in high dependency units (HDU) or intensive treatment units (ITU), life-threatening maternal conditions and maternal complications associated with ECC will be reported directly to the trial office and may require expedited reporting. It is recognised that some SAEs are to be expected in these women either as an outcome or known risk of the intervention and while it is important that they are reported, this can be done within the relevant CRF. Examples include but are not limited to miscarriage, preterm delivery, admission to hospital for delivery or removal of suture, damage to cervix during ECC insertion or admission to a neonatal unit.
Outcome measures
The primary outcome is pregnancy loss defined as miscarriage, termination of pregnancy and perinatal mortality defined as stillbirth and neonatal death in the first week after birth. Secondary outcomes are detailed in Table 2. All the core outcomes for PTB are included (van 't Hooft 2016).
Table 2
Full description of all secondary outcomes including complete core outcome set for preterm birth
|
Secondary Outcomes – Maternal, Neonatal and Paediatric
|
|
Maternal
|
|
Pregnancy loss (miscarriage, termination of pregnancy and perinatal mortality, including any stillbirth or neonatal death in the first week of life). Excluding those due to congenital anomalies (chromosomal and/or structural) assessed via death certification.
|
|
Time from conception to pregnancy end (any reason)
|
|
Miscarriage & pre-viable neonatal death (defined as delivery < 24 weeks)
|
|
Stillbirth (defined as intrauterine death ≥ 24 weeks)
|
|
Gestation at delivery
|
|
Pre-term delivery (pre-specified groups of ≤ 28/≤32/≤37 weeks)
|
|
Maternal sepsis (at any time in pregnancy and until discharge from hospital postnatally)
|
|
Preterm (< 37 weeks) pre labour rupture of membranes (> 24 hours prior to delivery) (PPROM) adjusting for gestational age at occurrence of membrane rupture
|
|
Mode of initiation of birth (spontaneous or iatrogenic)
|
|
Indication for iatrogenic delivery (maternal and/or fetal)
|
|
Mode of delivery (vaginal or operative vaginal or caesarean)
|
|
Cerclage placement complications assessed as a composite and individually: cervical laceration; bleeding from cervix; ruptured membranes; bladder injury
|
|
Cerclage removal complications assessed as a composite and individually: cervical tears; difficulty in removal defined as requiring unexpected anaesthesia or unexpected dissection of suture
|
|
Suspected or confirmed chorioamnionitis (during pregnancy and up to 7 days postnatally)
|
|
Maternal admission to HDU or ITU pre-delivery
|
|
Maternal admission to HDU or ITU post-delivery
|
|
Serious adverse events
|
|
Neonatal
|
|
Early neonatal death (defined as a death within 7 days after delivery)
|
|
Late neonatal death (defined as a death beyond 7 days and before 28 days after delivery)
|
|
Early neonatal death (defined as a death within 7 days after delivery excluding those secondary to congenital anomalies)
|
|
Late neonatal death (defined as a death beyond 7 days and before 28 days after delivery excluding those secondary to congenital anomalies)
|
|
Birth weight adjusted for gestational age and sex
|
|
Small for gestational age (< 10th centile)
|
|
Advanced resuscitation at birth (assisted ventilation and/or drug administration and/or cardiac compressions)
|
|
Admission to specialist care (SCBU/NICU/HDU/transitional care)
|
|
Length of stay in each additional specialist care setting
|
|
Suspected sepsis (clinically diagnosed defined as commenced on intravenous antibiotics for > 48 hours after birth)
|
|
Confirmed sepsis (positive microbiology)
|
|
Brain injury (defined as any intraventricular haemorrhage (IVH) (excludes subependymal haemorrhages), parenchymal cystic or haemorrhagic lesion or persistent ventriculomegaly (VI > 97th percentile)
|
|
Respiratory support (ventilation/CPAP)
|
|
Days on respiratory support
|
|
Supplementary oxygen requirements at 36 weeks corrected gestational age
|
|
Necrotising enterocolitis (Bell’s stage 2 or 3)
|
|
Retinopathy of prematurity requiring laser treatment
|
|
Disabilities
|
|
Congenital abnormalities
|
|
Serious adverse events
|
|
Paediatric Outcomes
|
|
Death at greater than 28 days until 2 years
|
|
At two years corrected age, outcomes obtained from questionnaires, which includes general health, PARCA-R questionnaire and an assessment of motor outcomes (further details given in the statistical analysis plan).
|
Sample size estimates
Published literature (Althusius 2003, 2015) noted large effect sizes in favour of cervical cerclage. We accept these effect sizes are likely to be exaggerated due to methodological limitations of these studies and so have opted for a more conservative target difference of a 33% relative risk reduction from 60% pregnancy loss in the expectant management group to 40% in the ECC group. To detect this difference with 90% power (α = 0.05) we require 260 women in total (130 in each arm). Smaller differences are also likely to be clinically relevant. The pregnancy loss rate in the expectant management arm is uncertain; however, 260 women would still achieve high levels of power (≥ 80%) in scenarios where the event rate approaches high levels (Fig. 2).
Statistical analysis
A separate statistical analysis plan for the quantitative analysis of the C-STICH2 study will provide a detailed description of the planned statistical analyses. A brief outline is given below.
Pregnancy loss will be summarised by treatment arm using frequencies and percentages. A log-binomial model will be used to generate relative risks (and 95% confidence intervals (CI)), adjusting for the minimisation variables.
Secondary outcomes which are binary will be analysed as per the primary outcome. Continuous outcomes which are deemed to be normally distributed will be summarised using means and standard deviations and a linear model will be fitted to generate adjusted mean differences (and 95% CIs). Continuous outcomes which are not deemed to be normally distributed will be summarised using medians and interquartile ranges and unadjusted differences in medians will be produced with 95% CIs. Time to event data will be summarised using medians and interquartile ranges. A cox regression model will be fitted to generate adjusted hazard ratios (and 95% CIs) and a Kaplan Meier plot will be produced to assess the data visually. All analyses will be adjusted for the minimisation variables (where possible).
All analyses will be by intention to treat. Every attempt will be made to collect pregnancy outcome data on all women, and it is anticipated that missing data will be minimal. Women with missing primary outcome data will not be included in the first instance. Sensitivity analyses will be conducted to assess the impact of missing data. Subgroup analyses will be limited to gestational age at randomisation, cervical dilatation and use of adjuvant therapies.
Health economics
The principal economic analysis will be based on the data collected within the trial and will relate only to the initial period of assessment and the principal outcome of the trial at 7 days. The results of the economic evaluation will be expressed in terms of major outcomes averted (MOA), where MOA represents a composite outcome based on miscarriage, termination of pregnancy, neonatal death and PTB.
A further analysis based on all data up to the infant reaching two years of corrected age will be carried out; the outcome of this second analysis will be based on the parent report and neurodevelopment at two years.
The economic evaluation will primarily take the perspective of the NHS and Personal Social Services (PSS) and as far as possible, depending on available data in the literature, will also be analysed from the societal perspective. Data will be collected prospectively on NHS resource from all participating centres for both arms of the trial and follow up care. Unit costs from routine sources will be attached to resource use to estimate overall costs for each trial arm.
A bootstrapping approach will be used to calculate confidence intervals around the difference in mean costs. Initially, the base-case analysis for the within trial analysis will be framed in terms of cost-consequences, reporting data in a disaggregated manner on the incremental cost and the important consequences as assessed in the trial arms. An incremental economic analysis will be conducted on the primary outcome and other secondary outcomes. The results of these economic analyses will be presented using cost-effectiveness acceptability curves to reflect sampling variation and uncertainties in the appropriate threshold cost-effectiveness value. Simple and stochastic cost effectiveness analyses will explore the robustness of the results to plausible variations in key assumptions and variations in the analytical methods used, and to consider the broader issue of the generalisability of the results.
Trial management and oversight
The trial is registered with the ISRCTN registry - trial reference ISRCTN12981869.
The trial is funded by NIHR HTA (project number 16/151/01). The sponsor is Birmingham Women’s and Children’s NHS Foundation Trust. Neither the funder nor sponsor are involved in data collection or analysis.
The trial will be administered by a clinical trials unit with extensive experience. Data will be kept in accordance with General Data Protection Regulations 2018. Completed CRFs will be reviewed by the clinical trials unit and missing or ambiguous data queried. The sponsor will ensure data integrity through quality assurance processes and audit at participating sites.
Any changes to the protocol will be agreed by the TSC prior to implementation and these will be disseminated to individual sites by the trial management group (TMG) subject to research ethics committee approval.
The TMG are responsible for the day to day running of the trial. The TSC and data monitoring committee (DMC) provide independent oversight of the trial including an assessment of the pilot study at the end of year two in line with the pre-specified objectives. TSC members include a majority of members who are independent of the investigators, their employers, institutions and the funding body. The DMC comprises three independent members (two obstetricians and a statistician with extensive trial experience) who are responsible for reviewing interim analyses. Responsibility for continuation or modification of the trial is held by the TSC and will include guidance from the DMC. The terms of reference and charter for this DMC will be guided by the DAMOCLES project, and we anticipate the DMC and TSC will meet biannually.
Pilot evaluation
The internal pilot comprises the first 18 months of recruitment to C-STICH2. Following this, the feasibility of the trial will be assessed on its ability to screen, randomise and follow-up women. To support the quantitative assessment of feasibility a qualitative process evaluation is fully embedded in this pilot phase and will explore the feasibility, acceptability and appropriateness of the trial and intervention for HCPs and women (as described previously).
We have limited the use of fixed stop/go criteria for the pilot aiming to evaluate the strengths and weaknesses of the proposed trial qualitatively proposing solutions to the challenges experienced. This is partly because so little is known about the prevalence of the condition and its sequelae. Further information will be obtained from the minimal dataset.
Following the pilot phase of the trial the TMG and oversight committees will review the study progress and make recommendations to the funder on how the project should proceed. We anticipate multiple possible scenarios for how the project could proceed following the internal pilot, whilst maintaining the central research question. These are as follows:
(1) The RCT is deemed feasible and remains unchanged (as detailed previously).
(2) The RCT is deemed semi-feasible, and a change is required. Recruitment is considered feasible; however, recruitment rates are not as anticipated and it is considered unlikely we will reach the original sample size target of 260 women within the propose timelines. Thus, the RCT could be potentially underpowered within the original sample size parameters. Alternative analysis methods will be considered. In this scenario, a prospective observational cohort study (POS) would be implemented to run in parallel to the trial, aiming to collect detailed information from eligible women who have declined participation in the randomised cohort or where the randomised trial was not offered to the women (Fig. 3A). Ineligible women or those who do not consent to the POS will be considered for the minimal dataset only (as per methods described previously). The POS data will aim to support findings from the RCT.
(3) The RCT is considered not feasible and is discontinued. In this scenario the project would proceed with a POS and minimal dataset only (Fig. 3B).