Cervical Incompetence: What It Is, and What It Isn't

Cervical insufficiency — the older name is cervical incompetence — describes a cervix that opens in the middle of pregnancy without contractions and without warning, so that the pregnancy is lost or the baby is born far too early. It is a real thing that happens to real women. It is also one of the most over-diagnosed conditions in obstetrics, and understanding why matters, because the treatment is an operation.

This page is about the condition and how it is actually established. A companion page covers the stitch itself — what the operation involves, when it helps and when it does not.

What actually happens

In a normal pregnancy the cervix stays long and closed until the very end, then softens, shortens and opens in response to labour. In cervical insufficiency that sequence starts far too early, and crucially it starts without the contractions that would tell a woman something was happening. The classical story is a woman at 18 to 22 weeks who feels pressure, or notices discharge, or feels nothing at all, and is found to be several centimetres dilated with the membranes bulging through.

Two definitions are worth having in front of you, because both are quoted constantly and neither is as solid as it sounds. The American College describes it as the inability of the cervix to retain a pregnancy in the absence of contractions or labour in the second trimester — and then says, in the same document, that the diagnosis is challenging because of a lack of objective findings and clear diagnostic criteria. The Royal College is blunter still: it calls cervical insufficiency an imprecise clinical diagnosis applied to women with that history, in which it is assumed the cervix is weak.

That is not a technicality. It is the reason so much of what follows is uncertain, and the reason a second opinion is reasonable rather than rude.

The "painless" problem

Every textbook and every website will tell you the hallmark is painless dilatation. It is worth knowing two things about that.

First, it is often not painless. Pelvic pressure, backache, an increase in discharge and spotting are all commonly described in retrospect, and women frequently say afterwards that they did mention something and were reassured.

Second, and more importantly: the Royal College states that it is unknown whether the specific characteristics of the previous event — painless dilatation, waters breaking, previous cervical surgery — are helpful in deciding whether to place a stitch. The feature that defines the condition in every description of it has never been shown to help make the decision that follows from it.

How common is it?

Published estimates run from 1 in 100 to 1 in 2,000 pregnancies. A twentyfold spread is not measurement error; it is disagreement about what counts. Under the traditional definition it is said to affect up to 1% of pregnancies. Under a broader definition that includes a previous preterm birth plus a short cervix, it becomes 3 to 4% of all pregnancies. One frequently quoted figure attributes 8% of second-trimester losses to it, and that rests on a single cohort of 158 women.

The honest summary is that nobody knows how common it is, because nobody agrees what it is. Treat any confident single number with suspicion.

The tests that do not work

This is the section that matters most to a woman who has been investigated between pregnancies, and it is the one almost nobody is told.

Over the years several tests were devised to prove a cervix was weak while a woman was not pregnant: passing a Hegar dilator and seeing whether it met resistance, hysterosalpingography, pulling on a balloon catheter and measuring the traction needed to dislodge it, inflating a balloon inside the uterus and measuring the pressure, and a calculated cervical resistance index.

If you were told between pregnancies that a dilator passed too easily, or that an HSG showed a wide internal os, and that this proved your cervix is weak — you were given information that no guideline supports. It does not mean the diagnosis is wrong. It means that particular test did not establish it.

What does carry information: cervical length

Measuring the cervix on a transvaginal scan during pregnancy is the one assessment with real predictive value. A short cervix is usually defined as 25 mm or less, measured between about 16 and 24 weeks.

The relationship is continuous rather than a cliff edge. In a study of 2,915 women the average cervical length at 24 weeks was about 35 mm, and the relative risk of spontaneous preterm birth before 35 weeks rose steadily as the measurement fell — roughly 2.4 times at 35 mm, 3.8 at 30 mm, 6.2 at 26 mm, 9.5 at 22 mm, and 14 times at 13 mm. There is no single number below which you are in trouble and above which you are safe.

And here is the sentence that keeps the whole thing honest, from the American College: a short cervical length is a marker of preterm birth in general rather than a specific marker of cervical insufficiency. A short cervix tells you this pregnancy is at risk. It does not tell you the cervix is structurally weak, and it does not by itself mean an operation.

What raises the risk, with actual numbers

Most risk factors for this condition are asserted rather than quantified. Two are properly measured, and both are things that were done to the cervix.

• Previous treatment for abnormal cervical cells. Pooled across 71 studies and more than six million women, any local excision of the cervix raised preterm birth before 37 weeks with a relative risk of 1.78 (1.60–1.98), birth before 32 to 34 weeks 2.40 (1.92–2.99), and birth before 28 to 30 weeks 2.54 (1.77–3.63).
• Depth matters, and it is a gradient. A shallow excision of 10 to 12 mm carried a risk of 1.54; 15 to 17 mm, 2.77; and 20 mm or more, 4.91 (2.06–11.68). Having more than one excision carried 3.78 (2.65–5.39) against 1.75 for a single treatment.
• By technique, cold-knife conisation carried the highest risk (14% versus 5%, RR 2.59), LEEP an intermediate one (11% versus 7%, RR 1.70), and laser conisation was not associated with an increase at all. If you have had treatment, it is worth knowing which you had — our page on abnormal smears and CIN explains the difference.
• Repeated mechanical dilatation of the cervix carries a modest association, around 1.36 (1.24–1.50). Modest is the correct word; this is not the dominant factor it is often presented as.
• Previous second-trimester loss or preterm birth is the commonest reason the question arises at all — but see the section below, because it is also where the most unnecessary surgery happens.

Uterine anomalies, connective tissue disorders such as Ehlers-Danlos and Marfan, and historical DES exposure appear on every list. They are plausible and they are poorly quantified — no reliable effect estimate exists for cervical insufficiency specifically, and anyone quoting you one is quoting something that was not measured.

The case against a reflex stitch

If you have had one mid-trimester loss, you will very likely be offered a cerclage in the next pregnancy. Before agreeing, these are the numbers.

None of that means a stitch is never right. It means the sensible default after one loss is surveillance rather than surgery — serial cervical length measurements, and a stitch if and when the cervix actually shortens. That approach spares an operation to most of the women who would never have needed it, and it catches the ones who do.

The exception is a woman with three or more previous preterm births or second-trimester losses. That is the one group in which the benefit was clear and consistent, and in whom a stitch placed on history alone is genuinely justified.

Vaginal progesterone

Less known than it should be in India, and the evidence is good.

An analysis pooling individual data from five randomised trials and 974 women with a cervix of 25 mm or less found that vaginal progesterone reduced preterm birth before 33 weeks with a relative risk of 0.62 (0.47–0.81) — graded as high-quality evidence. Neonatal death was 1.4% against 3.2%, and there was no difference in how the children were developing at two years.

Current specialist guidance recommends it for a cervix of 20 mm or less before 24 weeks, and suggests considering it between 21 and 25 mm after a proper conversation. The older injectable progesterone should not be used for a short cervix. It is a pessary you use yourself, it is inexpensive, and it is not an operation — which is why it deserves to be discussed before a stitch is, and frequently is not.

What does not work

• Bed rest. The American College says activity restriction has not been proved effective and its use is discouraged. Specialist guidance recommends against it. In women with a cervix under 30 mm, activity restriction was associated with more preterm birth, with an adjusted odds ratio of 2.28 (1.36–3.80). Bed rest of three days or more raises the risk of clots from 0.8 to 15.6 per 1,000 women. Lying down is not a neutral intervention; it is one with a harm profile and no proven benefit.
• Bed rest after a stitch has been tested directly and made no difference: 7.9% versus 8.5% preterm birth.
• The cervical pessary. A large trial of 544 women with a cervix of 20 mm or less was stopped early — for futility and for concern about deaths. Preterm birth or fetal death was 45.5% against 45.6%, and fetal or infant death was 13.3% against 6.8%. Specialist guidance now says it should not be used in singleton pregnancy.
• Abstaining from sex after a stitch is not routinely recommended, contrary to what most women are told.

An Indian complication worth naming

India records more preterm births than any other country — around 3.02 million in 2020, roughly 13% of Indian births and more than a fifth of the world's total. So the question of who to screen matters here more than almost anywhere.

And there is a real divergence. The Indian federation's practice algorithm suggests universal cervical length screening even in low-risk women. The Royal College explicitly does not recommend a stitch for a short cervix found incidentally in a low-risk singleton pregnancy, and specialist guidance grades that recommendation firmly.

The practical consequence, and it happens in Gurugram routinely: a low-risk woman is scanned, a cervix of 22 mm is found, she is frightened, and a stitch is discussed — for a finding that no guideline would treat with surgery. Progesterone is a reasonable conversation in that situation. An operation, in the absence of a previous preterm birth, is not.

A pregnancy after a late loss

If you are reading this while planning or early in the next pregnancy, this is roughly what a sensible plan looks like.

• A proper history first, including exactly what happened and when, and what was found at the time. The details of the loss matter more than any test.
• Other causes excluded, because a late loss is not always the cervix. Our page on recurrent miscarriage covers the tests that are worth doing and the many that are not.
• Serial cervical length measurement from about 16 weeks, usually fortnightly to 24 weeks. These are arranged by referral, as are all scans, and reviewed with you here.
• Vaginal progesterone if the cervix shortens, started promptly.
• A stitch considered if the cervix shortens despite that and you have had a previous preterm birth — a conversation, not a reflex.
• No bed rest. Carry on with your life.

What happens at the clinic

Consultation, going through the notes from the loss, the rest of the recurrent-loss assessment, blood tests, prescribing progesterone and organising the surveillance schedule all happen at the clinic. Cervical length scans, like all scans, are arranged by referral — we do not run imaging here. A cerclage is an operation and is done at hospital.

If you have been told you need a stitch on the strength of one loss and nothing else, that is a reasonable thing to want a second opinion about. Bring the discharge summary and any scan reports from the pregnancy you lost.