Clinical overview
Cervical cerclage is a surgical procedure in which a non-absorbable suture is placed around the cervix to provide mechanical support and resistance against premature cervical shortening and dilatation. It is one of the few interventions we have for the syndrome of preterm birth that targets a specific, identifiable mechanism: cervical insufficiency, in which the cervix shortens, effaces and dilates painlessly, often in the mid-trimester, leading to second-trimester loss or very preterm birth.
The exam framing of this objective is important. Cerclage is not a treatment for "threatened preterm labour" in general, and it is not a substitute for progesterone. It is a targeted intervention for a minority of women, and choosing the right woman is harder than performing the operation. Most registrars who get this objective wrong do so by over-applying cerclage (placing it where vaginal progesterone or expectant management would do as well or better) or by missing the small group in whom it genuinely changes the outcome. The three classic clinical situations — history-indicated, ultrasound-indicated, and physical-examination-indicated ("rescue") cerclage — each have different evidence, different urgency, and different risk profiles, and you must be able to distinguish them.
In the South African context, preterm birth is a major contributor to perinatal mortality, and the Saving Mothers / Saving Babies (NCCEMD) reporting consistently identifies immaturity-related deaths as a leading avoidable category. Cerclage is a procedure for the regional or tertiary level; the relevant decisions for a district-level practitioner are recognising the at-risk woman, referring appropriately, and not delaying when a cervix is found open. See preterm-birth-and-pprom and progesterone-in-pregnancy for the wider preterm-prevention picture, and recurrent-pregnancy-loss for the differential of mid-trimester loss.
Core knowledge
What "cervical insufficiency" is — and is not
Cervical insufficiency (the older term "incompetence" is discouraged) describes the inability of the cervix to retain a pregnancy in the absence of contractions or labour, classically presenting as painless cervical dilatation in the mid-trimester. It is best understood as one end of a continuum of cervical function rather than a discrete on/off defect. Cervical competence is a function of structural integrity (collagen, smooth muscle, the internal os), and is influenced by prior cervical trauma, congenital factors, and the same inflammatory and infective pathways that drive spontaneous preterm labour. This matters because cerclage addresses only the mechanical component; where infection or inflammation is the driver, a suture alone will not help and may even be harmful.
Recognised risk factors and associations include:
- Prior cervical surgery — large or repeated LLETZ/cone biopsy, mechanical dilatation; this is increasingly relevant in SA given the cervical-screening and CIN-treatment workload (see cin-management and cervical-screening-sa).
- Prior mid-trimester loss or spontaneous preterm birth, particularly with a history of painless dilatation.
- Congenital factors — Müllerian anomalies, in-utero DES exposure (now rare), collagen disorders (e.g. Ehlers–Danlos).
- Prior obstetric trauma to the cervix.
The three types of cerclage by indication
| Type | Trigger | Timing | Evidence strength |
|---|---|---|---|
| History-indicated (elective) | Past history (≥3 mid-trimester losses / preterm births classically) | ~12–14 weeks, after viability scan | Benefit in a defined high-risk group |
| Ultrasound-indicated | Short cervix on TVS (singleton + prior spontaneous PTB) | When short cervix detected (typically 16–24 wk) | Reduces PTB in this subgroup |
| Physical-examination-indicated ("rescue") | Dilatation with exposed membranes on exam | Emergency, individualised | Lower-quality evidence; may prolong gestation |
Figure J15.1 — The three cervical cerclage indication pathways, with timing, triggers and inappropriate-use traps.
Surgical approaches
- Transvaginal cerclage is the default. Two techniques are described: the McDonald (a purse-string suture at the cervicovaginal junction, no bladder dissection, easily removable) and the Shirodkar (a higher suture placed after dissecting and reflecting the bladder, with the suture buried submucosally). The McDonald is technically simpler and is the more commonly used; there is no robust evidence that the Shirodkar is superior, and the standard teaching is that the two are broadly equivalent in outcome.
- Transabdominal cerclage places the suture at the level of the internal os via laparotomy or laparoscopy. It is reserved for women in whom a transvaginal suture has failed or is anatomically impossible (e.g. a very short or absent vaginal cervix after trachelectomy or extensive surgery). It commits the woman to caesarean delivery and to a second procedure (or leaving the suture in situ) for removal, and carries the additional morbidity of abdominal surgery.
A non-absorbable tape or braided suture is used. The suture aims to support the cervix mechanically; it does not (and is not expected to) restore a cervix that has already substantially failed with established infection.
Assessment
Selecting the right woman is the core skill, and it begins long before the operation.
History
The most useful single piece of information is a careful obstetric history. Ask specifically about the gestation and manner of any prior loss or preterm birth: was it painless dilatation with bulging membranes (suggestive of cervical insufficiency), or did it follow contractions, bleeding, ruptured membranes, or systemic illness (pointing elsewhere)? Document prior cervical surgery (number and depth of LLETZ/cone procedures), prior dilatation and curettage, and any known uterine anomaly. A history of three or more consecutive mid-trimester losses or early preterm births is the classic profile for which history-indicated cerclage has the best supporting evidence.
Examination and investigations
- Confirm viability and gestation with ultrasound before any elective cerclage; date accurately (see gestational-age-assessment).
- Exclude fetal anomaly — there is no point supporting a pregnancy with a lethal malformation; arrange the anomaly scan appropriately, recognising that history-indicated cerclage is often placed before the routine anomaly scan window.
- Transvaginal ultrasound (TVS) of the cervix is the key surveillance tool for the ultrasound-indicated pathway. In a woman with a singleton pregnancy and a prior spontaneous preterm birth, serial TVS cervical-length measurement (commonly from ~16 weeks) allows a cerclage to be offered if the cervix shortens — typically to <25 mm before 24 weeks (standard threshold) — rather than committing every such woman to surgery. This is a more selective, evidence-aligned strategy than blanket elective cerclage.
- Exclude infection and labour before a rescue cerclage. Look for clinical chorioamnionitis (maternal pyrexia, tachycardia, uterine tenderness, offensive discharge, raised inflammatory markers, fetal tachycardia). Assess for contractions. A cerclage placed over an infected, labouring uterus is harmful.
- Speculum and gentle assessment in suspected physical-examination-indicated cerclage to define dilatation and whether membranes are visible or prolapsing.
Who should NOT have cerclage
This list is examinable. Cerclage is generally contraindicated or inappropriate where there is active vaginal bleeding suggestive of abruption, established preterm labour, clinical chorioamnionitis, ruptured membranes, lethal fetal anomaly, or fetal death. Multiple pregnancy is a particular trap: in twins, cerclage placed for a short cervix has not been shown to help and some older data suggested possible harm, so it is not routinely recommended (see multiple-pregnancy). The decision in twins should be individualised at tertiary level.
Management
History-indicated (elective) cerclage
Offered to the woman whose history defines a high a-priori risk (classically ≥3 previous mid-trimester losses or preterm births). It is placed electively, usually around 12–14 weeks, after a viability scan and ideally after excluding gross anomaly. Pre-operative counselling must cover the realistic expectation of benefit, the procedural risks, and the plan for removal.
Ultrasound-indicated cerclage
The preferred strategy for the woman with a singleton pregnancy and a prior spontaneous preterm birth who does not automatically meet history criteria: place her under TVS cervical-length surveillance and offer cerclage if the cervix shortens (e.g. <25 mm before 24 weeks). This spares women whose cervix remains long an unnecessary operation, and it is the approach favoured by current guidance over routine elective cerclage in this group.
Physical-examination-indicated ("rescue") cerclage — the emergency
RESCUE CERCLAGE IS A TIME-CRITICAL, INDIVIDUALISED DECISION. THE MOMENT YOU FIND A DILATED CERVIX WITH EXPOSED MEMBRANES IN THE MID-TRIMESTER, THINK IN THIS ORDER:
- STOP and EXCLUDE the contraindications first — active labour, clinical chorioamnionitis, ruptured membranes, abruption (bleeding), lethal anomaly, fetal death. If any are present, rescue cerclage is OFF.
- CALL FOR SENIOR / SPECIALIST HELP IMMEDIATELY — this is a regional/tertiary-level decision. A district practitioner's job is to recognise it, avoid digital interference that could rupture membranes, and refer urgently. Do not vacillate; the window is short.
- COUNSEL the woman honestly — rescue cerclage may prolong the pregnancy and improve the chance of a viable baby in selected cases, but the evidence is weaker than for the elective and ultrasound-indicated groups, and the procedure can precipitate membrane rupture or loss.
- TREAT the whole picture, not just the suture — consider corticosteroids for fetal lung maturity if near viability, infection screening, and tocolysis only if appropriate; bladder filling or amnioreduction may be used to reduce membranes back behind the os as an adjunct at the time of surgery.
The success of a rescue cerclage falls as dilatation increases and as membranes prolapse further; advanced dilatation with hourglassing membranes is a poor prognostic situation.
Figure J15.2 — Rescue cerclage emergency pathway: exclude contraindications, call senior help, counsel honestly and treat the whole picture.
Peri-operative care and removal
- Anaesthesia is usually regional. Antibiotic and tocolytic use around insertion is selective and not uniformly evidence-supported; follow local protocol.
- Removal timing is examinable. An elective transvaginal cerclage is typically removed at around 36–37 weeks, before the onset of labour, to avoid cervical laceration. If labour establishes earlier, remove the suture promptly. If the membranes rupture preterm (PPROM), the suture is generally removed because of the infection risk, individualised against the gestation and the wish to gain steroid cover — discuss at senior level (see preterm-birth-and-pprom).
- A transabdominal cerclage is left in situ and the baby delivered by caesarean; the suture may remain for a future pregnancy.
Figure J15.3 — Cerclage technique comparison, removal rules and complication watchpoints.
South African service context
Cerclage is a regional/tertiary-level procedure requiring ultrasound, theatre access and neonatal back-up. The NDoH National Integrated Maternal and Perinatal Care Guideline (NDoH, 2024) structures the levels of care and referral pathways that govern who does what and where. Practically:
- A woman with a suggestive history or a prior mid-trimester loss should be identified at booking and referred for specialist antenatal care and cervical-length surveillance where available (see antenatal-booking and high-risk-pregnancy-risks).
- TVS cervical-length screening may be resource-limited at district level; the high-risk woman is best surveilled where the imaging and the operator are available.
- In HIV-positive women (a large proportion of SA antenatal populations), there is no contraindication to cerclage; ensure she is on suppressive ART (TLD per SA HIV guidelines, see hiv-in-pregnancy) and that infection is excluded, given the interaction between genital-tract infection/inflammation and preterm birth.
- The realistic constraint is that rescue presentations may arrive too late at facilities without the means to operate; the discipline is early recognition and timely referral rather than improvised intervention.
Throughout, remember that cerclage is one tool. Vaginal progesterone is the comparator and sometimes the better choice for the short cervix; do not present cerclage as the only option (see progesterone-in-pregnancy).
Red flags / pitfalls
- Treating cerclage as a cure-all for preterm birth. It targets cervical insufficiency, not the inflammatory/infective pathway that drives most spontaneous preterm labour. Misapplied, it exposes women to surgical risk without benefit.
- Placing a suture over infection or labour. Cerclage in the presence of clinical chorioamnionitis, ruptured membranes, active bleeding or established labour is harmful — you may worsen sepsis or precipitate loss. Exclude these first, every time.
- Forgetting to exclude lethal anomaly and confirm viability before an elective procedure.
- Routine cerclage in twins. Not recommended for a short cervix in multiple pregnancy; individualise at tertiary level, do not apply the singleton logic.
- Digital or rough vaginal examination of a dilated mid-trimester cervix, risking iatrogenic membrane rupture, when a gentle speculum assessment and urgent referral were what was needed.
- Leaving an elective suture in past 36–37 weeks or through established labour, risking cervical laceration, annular detachment or uterine rupture. Have a clear removal plan documented in the notes.
- Missing complications: cerclage can cause cervical trauma/laceration, suture migration or erosion, bladder injury (especially Shirodkar/abdominal), bleeding, infection, membrane rupture, and — rarely with retained sutures in labour — cervical dystocia or uterine rupture. A buried Shirodkar or an abdominal suture must be actively planned for at delivery.
- Confusing the three indications. The history-indicated, ultrasound-indicated and rescue groups are not interchangeable; their evidence, timing and counselling differ. Be explicit about which one you are doing.
Evidence anchors
- RCOG Green-top Guideline No. 75 — Cervical Cerclage (current; replaces the archived GTG 60). The primary specialty reference for indications (history-, ultrasound- and examination-indicated), technique (McDonald vs Shirodkar, transabdominal), contraindications, the non-recommendation of routine cerclage in twins, and removal timing. Use this as the source of truth for this objective.
- NDoH National Integrated Maternal and Perinatal Care Guideline (NDoH, 2024) — the South African obstetric source of truth; governs levels of care, high-risk antenatal identification, and referral pathways relevant to who is screened, surveilled and operated on, and where.
- Saving Mothers / Saving Babies reports (NCCEMD), latest triennium — frame the burden of preterm birth and immaturity-related perinatal death in South Africa that this intervention sits within.
- RCOG GTG 74 — Antenatal corticosteroids — relevant to peri-viable rescue presentations where steroid cover is being weighed alongside cerclage decisions.
- RCOG GTG 73 — Preterm prelabour rupture of membranes (PPROM ≥24 weeks) — informs management of the cerclage when membranes rupture preterm (suture removal vs retention, infection risk).
- South African HIV / ART Consolidated Guidelines (2023; first-line TLD) and SAHCS 2023 Adult ART Guidelines — relevant to the large HIV-positive antenatal population in whom infection control and ART suppression matter around any cerclage decision.
Note on hedging: where this chapter gives numerical thresholds (e.g. a short-cervix threshold of <25 mm before 24 weeks, elective placement around 12–14 weeks, suture removal around 36–37 weeks, the ≥3 prior losses defining the classic history-indicated group), these reflect standard teaching and current guidance and should be confirmed against the current RCOG GTG 75 text and the NDoH 5th-edition guideline before being quoted as exact protocol numbers in an individual unit.