Cervical Cerclage: The Stitch, Explained

A cervical cerclage is a stitch placed around the cervix to hold it closed during pregnancy. It is one of the oldest operations in obstetrics and one of the most argued about. It genuinely saves pregnancies, and it is also performed on a great many women who would have been fine without it.

This page is about the operation. The companion page covers cervical incompetence itself — what the condition is, how it is established, and why surveillance rather than a reflex stitch is usually the right starting point.

The three operations

The ultrasound-indicated stitch: the conditional bit

This is the most common situation and the one where the evidence is most often misrepresented, so it is worth being precise.

Pooling individual data from five randomised trials, 504 women with a singleton pregnancy, a previous spontaneous preterm birth, and a cervix under 25 mm before 24 weeks: preterm birth before 35 weeks was 28.4% with a stitch against 41.3% without, a relative risk of 0.70 (0.55–0.89). The combined rate of death and serious illness in the babies was 15.6% against 24.8%, a relative risk of 0.64 (0.45–0.91). That is a real and worthwhile benefit.

Now the same analysis in women with a short cervix but no previous preterm birth — 419 pregnancies across five trials: preterm birth before 35 weeks was 21.9% against 27.7%, a relative risk of 0.88 (0.60–1.23). The confidence interval crosses one. There is no demonstrated benefit.

So the same scan finding justifies an operation in one woman and does not in another, and the thing that separates them is her history. The Royal College accordingly does not recommend a stitch for a short cervix found incidentally in a low-risk pregnancy, and specialist guidance grades that recommendation firmly. If you are being offered a stitch for a short cervix and you have never had a preterm birth, that is the question to ask.

McDonald or Shirodkar?

Two vaginal techniques, and most women are never told which they had.

• McDonald — a purse-string suture placed at the junction of cervix and vagina, without moving the bladder, with the knot left where it can be seen and reached. Straightforward to place and straightforward to remove.
• Shirodkar — the vaginal lining is incised, the bladder is lifted out of the way, and the suture is placed higher, closer to the internal os, then buried under the lining that is stitched back over it.

Do outcomes differ? The available comparison found no significant difference in birth before 33 weeks, with an odds ratio of 0.55 (0.2–1.3). That is not randomised evidence, but there is no good reason to believe one is superior.

What does differ is the practical consequence: a high, buried Shirodkar suture usually requires an anaesthetic to remove, while a McDonald normally comes out with a speculum and some pain relief. That is worth knowing in advance rather than discovering at 36 weeks. Ask which you have, and ask for it in writing.

The abdominal stitch, and the trial that settled a real question

A transabdominal cerclage is placed through the abdomen rather than the vagina, around the very top of the cervix. It is reserved for women in whom a vaginal stitch has failed, or where there is too little cervix left to stitch — after extensive cervical surgery, for example.

The obvious question — if a low vaginal stitch failed, should the next one be abdominal, or just placed higher? — was actually tested. Women with a previous failed low vaginal cerclage were randomly assigned to an abdominal stitch, a high vaginal stitch, or another low vaginal one.

• Abdominal against low vaginal: preterm birth before 32 weeks 8% against 38%, a relative risk of 0.23 (0.07–0.76). About four women need the abdominal operation for one additional pregnancy to get past 32 weeks.
• Pregnancy loss was 3% against 21%, a relative risk of 0.12.
• High vaginal against low vaginal: no better, a relative risk of 1.15 (0.62–2.16). Simply raising the stitch is not the answer, which is exactly what was being done before this trial.

Two practical points. First, the abdominal stitch can be placed before pregnancy or in the first trimester, and it can be left in place between pregnancies. Second — and this is not optional — every woman with an abdominal cerclage must be delivered by caesarean. It cannot be removed vaginally.

Laparoscopic and open placement give equivalent baby survival, around 91%. The keyhole route loses less blood and gets you home faster: significant blood loss in 0.0% against 1.9%, wound infection 0.1% against 1.5%, hospital stay 1.3 days against 6.1. It takes longer in theatre. In the trial above, incidentally, every abdominal stitch was placed by the open route — so the evidence is for the operation, and the laparoscopic route is a reasonable refinement of it rather than a separately proven thing.

Rescue cerclage

The frightening situation: you are 20 or 22 weeks pregnant, the cervix is found to be several centimetres open, and the membranes are bulging. Doing nothing here usually ends badly.

Across 10 studies and 757 women, a rescue stitch against expectant management gave neonatal survival of 71% against 43%, a relative risk of 1.65 (1.19–2.28). Pregnancy was prolonged by a mean of 34 days, and delivery happened on average 4.6 weeks later. In a situation with no good options, those are large numbers.

What the operation actually involves

1. Where. In an operating theatre, at hospital. This is not a clinic-room procedure anywhere.
2. Anaesthetic. Usually a spinal — awake, numb from the waist down — or a general anaesthetic. A catheter may be placed for a short time.
3. How long. Under an hour. A speculum is placed, the cervix is held, and the suture is passed around it and tied.
4. Going home. Usually the same day, though you may be kept in longer depending on the indication and how things go. A rescue stitch generally means a longer stay.
5. Afterwards. Some vaginal bleeding or brownish discharge for a day or two is expected. The stitch itself should not cause ongoing discomfort. Period-type cramping for a short time is common.
6. Activity. Resting in bed is not routinely recommended, and neither is abstaining from sex. Both are widely advised anyway. Bed rest in particular has been tested after a cerclage and made no difference — 7.9% against 8.5% preterm birth.

Taking it out — and the danger of leaving it in

A vaginal cerclage is removed between 36+1 and 37+0 weeks, unless a caesarean is already planned, in which case it can come out at the operation. Removal takes a few minutes, usually needs no anaesthetic, and pain relief is available. A high Shirodkar suture is the exception and generally does need an anaesthetic.

The point that matters most on this whole page: if labour starts with the stitch still in, it must be removed. A cervix contracting against a fixed suture can tear, and reported consequences include cervical laceration and, rarely, the cervix being torn away entirely. If you think you are in labour with a stitch in place, that is a same-day call to the maternity unit, not something to sleep on.

An abdominal stitch is not removed vaginally at all. It comes out at the caesarean, or is left in place for a future pregnancy.

The risks, quantified

• Fever after delivery is roughly doubled — an odds ratio of 2.35 (1.37–4.05) pooled across seven randomised trials. This is the most consistently demonstrated harm.
• Infection of the membranes is not increased — an odds ratio of 0.73 (0.36–1.46). Worth knowing, because it is the risk women most fear.
• Damage during the operation itself — bladder injury, cervical trauma, rupturing the membranes — is rare, under 1%.
• Bleeding and cervical trauma rise with urgency: bleeding 0.9% for a history-indicated stitch, 1.4% ultrasound-indicated, 2.3% for a rescue stitch; cervical trauma 0.2%, 0.6% and 1.3% respectively.
• Tearing of the cervix at delivery is reported at 11 to 14% against about 2% without a stitch — retrospective figures, so probably overstated, but not nothing.
• Caesarean is modestly more likely — a relative risk of 1.19 (1.01–1.40).
• Waters breaking early after a rescue stitch occurs in about 2.5%.
• Fistula — an abnormal connection to the bladder or bowel — exists only as isolated case reports. There is no meaningful rate to quote, which is itself reassuring.

What a stitch cannot do

A cerclage is a mechanical solution. Much of preterm birth is not a mechanical problem — it is inflammation and infection, and a suture does nothing about either. That is the reason a stitch is removed rather than relied upon when labour actually starts.

The honest numbers: even in the group where it works best, 28.4% of women still delivered before 35 weeks with a stitch in place. In an Indian tertiary series of 129 women the failure rate was 15.6% for planned stitches and 18.7% for emergency ones. The Royal College's own patient information says it plainly: even successfully placed, a stitch may not prevent late miscarriage or preterm birth.

Adding progesterone on top of a stitch does not help either — compared directly, stitch plus progesterone against stitch alone gave no difference in preterm birth before 37 weeks.

Twins

The guidance is explicit: a history-indicated or ultrasound-indicated cerclage is not recommended in a multiple pregnancy. Perinatal death in twins with a stitch was 19.2% against 9.5% without — a point estimate that favours harm, with a confidence interval that crosses one. Cochrane found no evidence of benefit and an association with lower birthweight and more breathing problems.

A rescue stitch in twins with an already-open cervix is a separate and genuinely unresolved question, and it belongs in a specialist fetal medicine unit rather than in a general discussion. Our page on twin pregnancy care covers what is recommended in twins, which is cervical length measurement and vaginal progesterone rather than surgery.

How it works out in Indian practice

A series of 129 women at a major Indian centre, 81 planned stitches and 48 emergency ones: delivery at or beyond 34 weeks in 71% of the planned group and 53.3% of the emergency group; live birth 84.4% and 81.3%; serious neonatal problems 14% against 35.5%.

The pattern is worth absorbing. An emergency stitch still produces a live baby in about four cases out of five — but the babies are smaller and sicker, and the difference between the two columns is essentially the difference between catching the problem on a scan and catching it when the cervix is already open. That is the argument for surveillance, made in outcomes.

What happens at the clinic

Consultation, deciding whether a stitch is appropriate for you, the rest of the pregnancy assessment, blood tests, prescribing progesterone, and all the antenatal care before and after the operation happen at the clinic. Cervical length scans, like all scans, are arranged by referral. The cerclage itself is an operation carried out at hospital, with the removal booked in advance.

If you have been offered a stitch and have not been told which of the three it is, or why it applies to you rather than to someone with a different history, that is a reasonable second opinion to seek. Bring any scan reports and the notes from previous pregnancies.