Methods

Design

The C-STICH trial was a pragmatic, open, parallel, multicentre, randomised, superiority trial of monofilament versus braided suture type at CC to prevent pregnancy loss. The trial had a favourable ethical opinion from the National Research Ethics Service (NRES) Committee Cambridgeshire and Hertfordshire obtained in February 2015. The final protocol version was v9.0 dated 25 March 2020. The protocol for the trial has been published,18 and the full protocol is available at the C-STICH website.

Participants

Some text in this chapter has been reproduced from our study protocol,18 published under the Creative Commons Attribution License (CC BY 4.0), which permits its unrestricted use, distribution and reproduction in any medium, provided the original work is properly cited (https://creativecommons.org/licenses/by/4.0/).

The participants in the C-STICH trial were recruited from secondary and tertiary care NHS hospitals located across the UK. Women were eligible for the trial if they required a vaginal CC as part of their routine care within their current pregnancy and they fulfilled the following eligibility criteria:

Inclusion

• Singleton pregnancy.
• Indication for CC for either:
  • a history of three or more previous midterm losses or premature births (≤ 28 weeks), OR
  • insertion of cervical sutures in previous pregnancies, OR
  • a history of mid-trimester loss or premature birth with a shortened (≤ 25 mm) cervix, OR
  • women whom clinicians deemed to be at risk of preterm birth either because of their history or because of the results of an ultrasound scan

Exclusion

• Had taken part in C-STICH previously.
• Aged < 18 years old at the time of presentation.
• Those requiring a rescue cerclage [for the purposes of the study, a rescue cerclage was defined as an emergency cerclage where stitches are inserted in women who have had their preterm labours (e.g. uterine contractions, progressive cervical dilatation, bulging membranes) sufficiently halted by tocolysis or other means between 15 and 28 weeks].
• Unwilling or unable to give informed consent.
• Those in whom a cerclage was to be placed by any route other than vaginally (e.g. via an abdominal route).
• Immediate need for insertion of a suture [immediate need for the insertion of a suture should not have been delayed by the trial (thus, if giving information about the trial and waiting for the participant to decide upon whether or not she wants to participate would delay the insertion of an urgently needed suture, then treatment should have gone ahead, and the woman should have been excluded from the trial)].
• Membranes that had ruptured or were surfacing (Women with membranes that are ruptured or bulging through the external OS should have a rescue cerclage and be excluded from trial participation).

Recruitment

Women were identified in antenatal clinics, in delivery suites or in dedicated preterm birth prevention clinics. Women requiring a vaginal CC as part of their standard care were approached by doctors, research nurses and midwives. The decision about CC had already been made by the woman and clinicians and ideally the woman should have already completed the consent process for cerclage. Following consent for cerclage, women were approached to consider taking part in C-STICH. The trial was explained to women and a patient information leaflet was provided. To promote equity in access to the trial, written information was available in five additional languages. All staff approaching women regarding the trial were trained in ‘Good Clinical Practice’ (GCP) and had received dedicated trial-specific training. Participation in the trial was entirely voluntary, and women were informed that participation or non-participation would not affect their usual care. If an eligible woman agreed to participate, written consent was obtained, and both the clinician and the woman signed the consent form. The original copy was kept in the investigator site file, one copy was given to the woman, one copy was retained in the woman’s hospital records and one copy was sent to the clinical trials unit (CTU). Women declining participation were recorded in screening logs and sites were also asked to record women undergoing CC who were not approached about the trial.

Randomisation

Once final eligibility was confirmed and consent obtained, women were randomised to the C-STICH trial by the research staff at sites using a secure online, central randomisation service provided by the Birmingham Clinical Trials Unit. Additionally, there was a central telephone back-up service available.

Randomisation minimisation

Women were randomised on a 1 : 1 basis to either a monofilament or braided suture for the CC. A minimisation procedure, incorporating a random element, using a computer-based algorithm, was used to avoid chance imbalances in important prognostic variables. Strata used in the minimisation were:

• planned bladder dissection (yes, no)
• intention to commence patient on progesterone (yes, no)
• indication for cerclage [a history of three or more previous mid-term losses or premature births (≤ 28 weeks), insertion of cervical sutures in previous pregnancies, a history of mid-trimester loss or premature birth with a (current) shortened (≤ 25 mm) cervix, women whom clinicians deem to be at risk of preterm birth either because of their history or because of the results of an ultrasound scan] (cerclage categories were developed to reflect the varying background risk of preterm birth)
• randomising centre.

Interventions

Both monofilament and braided suture threads were standard surgical materials already in use. Different brands of suture material were available for both suture threads. Prior to the C-STICH trial commencing, the brands of suture threads used within potential sites were established. The commonest types were Ethilon for a monofilament suture and Mersilene^® for a braided suture. Ethilon is a non-absorbable, sterile surgical suture composed of the long-chain aliphatic polymers of nylon. Mersilene is a non-absorbable, sterile surgical suture composed of poly-ethylene terephthalate. Due to the pragmatic nature of the trial, the brand of suture thread was left to the discretion of the clinician inserting the suture. However, Ethilon^® and Mersilene sutures were encouraged. An absorbable suture thread was not permitted.

The vaginal CC technique (i.e. with or without bladder dissection) was at the discretion of the surgeon. However, the surgeon’s surgical intention (with or without bladder dissection) was expressed prior to randomisation and used as a minimisation variable. The trial was pragmatic in nature and surgeons were allowed to use antibiotics, tocolytics and to perform the CC according to their local practices. The surgical procedure could include an occlusion stitch to close the external OS with the stipulation that the same suture type as the randomised allocation be used.

Co-enrolment

Women already participating in another study [observational or a clinical trial of an investigational medicinal product (CTIMP)] prior to their confirmed eligibility for C-STICH were permitted to co-enrol into C-STICH. After a woman had been randomised to the C-STICH trial, she could participate in further CTIMP trials, although trials for the prevention of second trimester miscarriage or preterm birth, or any trials with the same primary outcome as C-STICH were excluded (such as trials of tocolytic agents). Women participating in C-STICH were permitted to be recruited into observational studies.

Blinding

The clinicians performing the vaginal CC could not be blinded to the suture thread and the surgical record documented the type of cerclage inserted, the type of suture used and the placement of any knots to facilitate removal. As far as possible, we requested that women and other research staff remained blinded to the suture-thread allocation, including microbiologists and outcome assessors. The duration of blinding continued until both mother and baby had reached their respective trial end points (i.e. final follow-up).

Adherence monitoring

Adherence to the allocated intervention was monitored on the cerclage placement form, where the suture material used to place the primary cerclage was recorded. We defined adherence as those participants in the monofilament suture group who received a monofilament suture for their primary cerclage and those participants in the braided suture group who received a braided suture for their primary cerclage. Subsequent suture replacements following the initial placement were permitted as necessary. For subsequent sutures, we encouraged clinicians to use the same suture thread as per the initial cerclage; however this was not mandated. The type of suture used for subsequent sutures was not considered as part of the adherence definition.

Withdrawal from the trial

Women could voluntarily withdraw from the trial at any time. The reasons for withdrawal were captured where possible. If a participant explicitly withdrew consent to have any further data recorded, their decision was respected and recorded on the electronic data capture system. All communication surrounding the withdrawal was noted in the patient’s medical notes, and no further data were collected for that participant. In rare instances where a participant was randomised and was later found to violate the inclusion criteria, they remained in the trial unless they had a multiple pregnancy or there was no informed consent.

Outcomes and assessment

The outcomes for the C-STICH trial were developed in collaboration with our patient advisory group and followed on from the non-randomised pilot study, COTS.16 Patient input into the trial outcomes informed the primary outcome for this trial. Pregnancy loss was viewed as the most important outcome for women and their families, with prevention of prematurity considered as a secondary outcome, rather than the originally planned primary outcome.

Early in the trial, the core outcome set for preterm birth was developed and published.19 The trial management group (TMG) recognised the importance of collecting all core outcomes within this preterm birth trial. Therefore, the outcome set in C-STICH was expanded in February 2016 to include all outcomes in the core outcome set for preterm birth.19 No outcomes were removed in incorporating the core outcome set (C-STICH protocol amendment 3.0–4.0). In February 2016, the trial had commenced recruitment but the outcomes from the women recruited had not yet been collected and, therefore, all women in the trial had a complete preterm-birth core outcome set collected.19

Primary outcome

Pregnancy loss (miscarriage and perinatal mortality, including any stillbirth or neonatal death in the first week of life).

Secondary outcomes

Key secondary outcome

• Time from conception to pregnancy end (any reason).

Maternal outcomes

• Miscarriage and previable neonatal death (defined as delivery < 24 weeks).
• Stillbirth (defined as intrauterine death ≥ 24 weeks).
• Gestation at delivery (in live births ≥ 24 weeks).
• Gestational age < 28/< 32/< 37 weeks at delivery (in live births ≥ 24 weeks).
• Time from conception to onset of spontaneous vaginal delivery (in live births ≥ 24 weeks).
• Sepsis (at any time in pregnancy and until 7 days postnatal).
• PPROM.
• Mode of initiation of labour (spontaneous or induced).
• Mode of delivery (vaginal, operative vaginal or caesarean).
• Cerclage placement complications (cervical laceration/bleeding from cervix/ruptured membranes/bladder injury).
• Cerclage removal complications (cervical tears/need for anaesthetic/difficult to remove).
• Other maternal complications: vaginal bleeding/steroid use/chorioamnionitis/maternal pyrexia of 38°C (intrapartum/postnatal)/admission to HDU or ITU (pre/postdelivery).

Neonatal outcomes

• Early neonatal death (defined as a death within 7 days after delivery) (in live births ≥ 24 weeks).
• Late neonatal death (defined as a death beyond 7 days and before 28 days after delivery) (in live births ≥ 24 weeks).
• Birth-weight centile adjusted for gestational age and sex (in live births ≥ 24 weeks).
• Small for gestational age and sex (< 10th centile, in live births ≥ 24 weeks).
• Resuscitation at birth (in live births ≥ 24 weeks).
• Additional care required [special care baby unit (SCBU)/neonatal intensive care unit (NICU)/HDU/transitional care] and length of stay in additional care (in live births ≥ 24 weeks).
• Antibiotics within 72 hours (in live births ≥ 24 weeks).
• Sepsis (clinically diagnosed/proven) (in live births ≥ 24 weeks).
• Early neurodevelopmental morbidity (severe abnormality on cranial ultrasound scan) (in live births ≥ 24 weeks).
• Respiratory support and days on respiratory support (in live births ≥ 24 weeks).
• Supplementary oxygen requirements at 36 weeks post menstrual age (in live births ≥ 24 weeks).
• Necrotising enterocolitis (Bell’s stage 2 or 3) (in live births ≥ 24 weeks).
• Retinopathy of prematurity requiring laser treatment (in live births ≥ 24 weeks)
• Disabilities (live births ≥ 24 weeks).
• Congenital anomalies (in live births ≥ 24 weeks).

Outcome generation

There were no extra visits beyond the standard care required for women recruited into the C-STICH trial. All outcome data were collected from the medical records. The outcomes were collected either on electronic case report forms (CRFs) and entered directly into the database or on paper CRFs and sent to the trial office. Details of how the outcome measures were generated are given in Table 2.

Definition of the end of the trial

The end of trial was defined as the time when each participant had reached their defined end point. The maternal end point for the trial was 28 days post delivery or discharge to home (whichever occurred sooner).

For surviving babies, the definition of the end of the trial was different, depending on if the baby was born preterm (< 37 weeks) or at term. For preterm babies, the end of neonatal follow-up was the estimated date of delivery or discharge to home (whichever occurred sooner). For term babies, the end of the trial was 28 days post delivery or discharge to home (whichever occurred sooner). The primary analysis was scheduled to occur after all the corresponding outcome data had been entered into the study database and validated as being ready for analysis.

Serious adverse events

Serious adverse events (SAEs) were recorded on a purpose-designed SAE form and notified by local investigators to the CTU within 24 hours of the local investigators becoming aware of these events. Local investigators were responsible for additionally reporting SAEs to their host institutions, according to local regulations.

Full details of the SAEs reporting within C-STICH can be found in the published protocol.18 SAEs categorised by the chief investigator as both suspected to be related to the trial intervention and unexpected were classified as suspected unexpected serious adverse reactions (SUSARs) and were subject to expedited reporting.

Sample size

The original sample size for C-STICH was informed by a non-randomised meta-analysis of audit data.15 The pregnancy loss rate observed was approximately 7% in women who received a monofilament suture and 19% in women who received a braided suture [risk ratio (RR) 0.34, 95% confidence interval (CI) 0.18 to 0.63]. A total sample of 326 women would have been sufficient to detect a difference of this size with 90% power (alpha = 0.05). However, we inflated this to a total sample target of 900 women (which included inflation for an attrition rate of 2.5%) and which enabled us to detect a more plausible relative reduction of 41% (19% with braided sutures to 11.2% with monofilament sutures) with 90% power (alpha = 0.05).

Given that there was uncertainty around the estimates used for the rates of pregnancy loss, it was agreed that the data monitoring and ethics committee (DMEC) would monitor the overall (pooled) event rate throughout the study to assess any deviation from the original sample size assumptions. In July 2017, the DMEC disclosed that the current estimate of the pooled event rate was lower than anticipated and may affect the trial’s ability to detect a difference between groups, should one exist. They advised that to maintain 90% power (alpha = 0.05) to detect the same relative reduction of 41% the sample size should be increased. A final sample size of 2050 women was agreed, which assumed a pooled pregnancy loss rate of 7.4% [9.3% in the braided suture group and 5.5% in the monofilament group (RR 0.59)], with 90% power, an alpha error rate of 5% and allowing for 2.5% attrition. The overall pooled event rate and calculation was not disclosed to the TMG.

Statistical methods

A comprehensive statistical analysis plan (SAP) was drawn up prior to any analysis and provided to the independent DMEC and trial steering committee (TSC) for review. Full details of the statistical analysis can be found in the SAP, which is available on request from Birmingham Clinical Trials Unit.

In the first instance, participants were analysed in the treatment group to which they were randomised [intention to treat (ITT)], irrespective of adherence to the treatment protocol. All estimates of the differences between groups are presented with 95%, two-sided CIs, and were adjusted for the minimisation variables where possible. The time from conception to pregnancy end, gestational age at delivery and time from conception to onset of spontaneous vaginal delivery were further adjusted for gestational age at randomisation. The categorical data were summarised with frequencies and percentages. The normally distributed continuous variables were summarised with means and standard deviations (SDs); otherwise, medians and interquartile ranges (IQRs) were presented.

The primary outcome was analysed using a mixed-effects log-binomial model to generate an adjusted RR and risk difference (RD) (the latter using an identity link function), including the centre as a random effect. The statistical significance of the treatment group parameter was determined (p-value generated) through an examination of the associated chi-squared statistic (obtained from the log-binomial model which produces the RR). A hierarchical approach to testing was planned to control for the overall rate of type I error; if the primary outcome was to meet a conclusion of superiority, then the key secondary outcome (time from conception to pregnancy end) was also to be tested in this manner.

Other binary maternal outcomes and the following neonatal outcomes – early neonatal death, late neonatal death and small for gestational age – were analysed as per the primary outcome but were not subjected to hypothesis testing. For continuous secondary outcome measures (birthweight centile and gestational age at delivery), the adjusted mean differences were estimated using a linear regression model, including the centre as a random effect. Time-to-event outcomes (time from conception to pregnancy end, time from conception to onset of spontaneous vaginal delivery) were summarised using medians and IQRs. A Cox regression model was fitted to generate adjusted hazard ratios; Kaplan–Meier plots were produced to assess the data visually. Cerclage placement complications, cerclage removal complications and all other neonatal outcomes (excluding those previously listed) were analysed descriptively only. Numbers and percentages were provided for binary data and means (or medians) and standard deviations (or IQRs) for continuous normal (or non-normal) data. If the overall event rates exceeded 3%, then formal analysis was undertaken as per the methods outlined above.

Sensitivity and supportive analyses were limited to the primary outcome only. These included a per-protocol analysis (including only those participants who received their randomised allocation), an as-treated analysis and an assessment of missing data by means of a ‘tipping point’ approach, which explored the possibility that missing responses were ‘missing not at random’ (MNAR). Unadjusted models were utilised. Firstly, all women with missing outcome data were considered as having not met the primary outcome (i.e. pregnancy loss was no). Two scenarios were then considered. In the first scenario, in women who originally had missing data in the monofilament suture group, ‘events’ (i.e. pregnancy loss was changed from no to yes) were sequentially added to this group until the number of events added was equal to the number of women with missing outcome data in that group. With the addition of each event, an unadjusted model was run and the RR stored. The tipping point for the monofilament suture group occurred when enough events had been added such that the upper/lower limit of the CI from the corresponding model differed from that of the primary ITT finding (in regard to whether the CI crosses the null value of one), should the tipping point exist. A second scenario was then considered which repeated this process in the braided arm.

Pre-planned subgroup analyses (limited to the primary outcome measure only) were completed for the following: (1) planned bladder dissection (yes/no), (2) intention to commence patient on progesterone (yes/no) and (3) indication for cerclage (a history of three or more previous mid-term losses or premature births (≤ 28 weeks)/insertion of cervical sutures in previous pregnancies/a history of mid-trimester loss or premature birth with a (current) shortened (≤ 25 mm) cervix/women whom clinicians deem to be at risk of preterm birth either because of their history or because of the results of an ultrasound scan). The effects of these subgroups were examined by adding the subgroup-by-treatment-group interaction parameters to the regression model. p-values from the tests for statistical heterogeneity were presented alongside the effect estimate and 95% CI within each subgroup. In addition to this, a ratio(s) [and 95% CI(s)] was provided to quantify the difference between the treatment effects estimated within each subgroup.

Interim analyses of effectiveness and safety end points were performed on behalf of the DMEC on an approximately annual basis during the period of recruitment. These analyses were performed with the use of the Haybittle–Peto principle and, hence, no adjustment was made in the final p-values to determine significance.20

All analyses were performed in SAS (version 9.4) or Stata (version 17.0).

Trial oversight

Study oversight was provided by a TSC (chaired by Professor Harry Gee, University of Warwick) and a DMEC (chaired by Professor Marian Knight, National Perinatal Epidemiology Unit).

The TSC provided independent supervision for the trial, providing advice to the chief and co-investigators and the sponsor on all aspects of the trial throughout the study. The DMEC adopted the DAMOCLES charter to define its terms of reference and operation in relation to oversight of the C-STICH trial.21