Why are proponents of the obstetric paradigm so concerned with controlling the capricious cervix?
We might have moved on from the old notion that a woman’s uterus is able to wander around her body and wreak havoc at random, but the uterus in general and the cervix in particular are still deemed rather unpredictable and potentially naughty parts of a woman’s body. For most of pregnancy, good cervical etiquette dictates that it should remain long, thick and closed. Indeed, Western medicine defines the cervix as incompetent and quickly offers intervention if it looks like it will open too soon. Yet, as soon as that arbitrary point called ‘established labour’ has been reached, expectations change and the cervix is expected to thin, shorten and open according to the schedule of the partogram. If it then falls short of these requirements, it is deemed to have failed to progress and other kinds of intervention are proposed.
The last decade has seen the growth of a new method of assessing the state and behaviour of the cervix in pregnancy; the transvaginal ultrasound (TVU). The increasing popularity of this intervention seems to be based on the combination of a few studies showing that the transvaginal route is a more accurate way of measuring the cervix than digital examination (Jackson et al 1992, To et al 2000) and other studies showing a correlation between shorter cervixes (generally defined as less than 15mm in length) and premature labour (Heath et al 1998a, 1998b, Rozenberg et al 2002). Bearing in mind the enormous risks attached to preterm labour and birth, this might not, on the surface, seem too problematic. But I believe there are a number of questions and issues which should be considered in relation to this invasive procedure.
Routine Cervical Screening by TVU
Some of the most oft-cited studies which have been set up to explore questions around measuring cervical length by TVU and then offering interventions to women who are deemed to have short cervixes are those conducted by a group of British researchers. Heath et al (1998a) measured the cervices of 2567 women at 23 weeks gestation and found that 1.7 per cent of women had a cervix less than 15mm long. In their abstract, the authors state that, “…this group contained 86%, 58% and 20% of pregnancies that delivered spontaneously at 28, 32 and 36 weeks, respectively” (312) and conclude that, “measurement of cervical length provides accurate prediction of risk for early preterm delivery” (312).
After they had measured all these women’s cervices, Heath et al (1998b) then identified 43 women whose cervices were less than 15mm long. These women were then randomised into two groups: 21 women were “managed expectantly” (312), while 22 women underwent cervical cerclage (Heath et al 1998b). In this study, around half of the women whose cervices were less than 15mm but who didn’t have a Shirodkar suture went into labour before 32 weeks, while none of the women in the cerclage group went into labour before this point. All but one of the babies born to women in this study survived; the baby who did not survive was born to a woman who did not have cervical cerclage. As a consequence of studies like this one, the notion has grown that routine cervical length measurement (CLM) by TVU, coupled with treatment (which may include cerclage) for those women who have short cervixes, might be advisable on a routine basis in order to reduce the incidence of preterm labour and birth.
The Other Side of the Story
However, not everybody agrees with the idea that routine CLM by TVU is either useful or acceptable as a routine intervention, and one of the main reasons for this concerns the question of how good CLM is as a screening test. In Heath et al’s (1998a) study, although 86 per cent of the women who gave birth before 28 weeks had been found to have a cervix less than 15mm on TVU, 14 per cent of the women who gave birth before 28 weeks had a cervix which was defined as of ‘normal’ length on TVU, and some of the women who had a short cervix did not go into premature labour. Hoesli et al (2003) looked at a number of studies reviewing the use of CLM as a screening test, and concluded that, while CLM may be useful as a screening test for women who are already at risk of premature labour, they would not recommend it for all women.
One of Hoesli et al’s (2003) concerns is that, when it comes to using CLM as a population screening test, the sensitivity (or true positive rate, which, in this instance, shows how accurately CLM identifies the women at risk of preterm labour) and specificity (or true negative rate, which shows how accurately CLM can identify the women who are not at risk of preterm labour) are not within acceptable limits. CLM has low sensitivity as a screening test when used on the population of women as a whole, partly because preterm labour occurs in only a small percentage of women. As demonstrated by Heath et al’s (1998a) results, CLM will only identify a proportion of the women who are at risk of preterm labour.
It might be possible to increase the sensitivity and therefore correctly identify more women who are at risk of preterm labour by increasing the cut-off point at which women are deemed to be at risk. In this instance, this might mean that we increase the definition of a short cervix from 15mm to, say, 18mm or 20mm, but the effect of this is only to decrease the specificity. This would result in a higher false positive rate, where more women will be told they have a problem when they do not, with all of the anxiety and unnecessary intervention that this brings (Rozenberg et al 2002). Because sensitivity and specificity are mathematically inter-related, higher accuracy on one almost always means lower accuracy on the other. Where CLM is offered as a screening test to all women, the cut-off point chosen locally will either mean that a number of women who are at risk of preterm labour will not be correctly identified, or that a number of women who are not at risk will be told that they are.
The Bigger Picture
Hoesli et al (2003) also point out that screening is only useful when effective preventative therapy is available, and argue that the debates regarding tocolysis and cervical cerclage are not yet concluded. Certainly the Heath et al (1998b) study, which included only 43 women, may not be large or rigorous enough on which to base decisions about offering the rather significant intervention of cervical cerclage. Some midwives believe that many cases of preterm labour can be prevented by improving women’s nutrition, or by reducing stress, or by exploring women’s lifestyle choices with them. It seems pretty clear from the literature that cervical “incompetence” is only one factor in preterm labour, and that many, many questions exist around this area.
More questions are raised when we consider how pregnant women feel about having to remove their clothing and having a probe inserted into their vagina. What about the stress that this test and the anxiety that any possible conclusions and recommendations which arise from it might create? What might this intervention represent when considered from postmodernist and feminist perspectives? Although a couple of early studies showed that transperineal ultrasound is just as effective at measuring cervical length as TVU (e.g. Ozdemir et al 2005), we are still left with the fact that any kind of routine cervical assessment will have to be applied to several hundred women in order to identify one woman with a shorter-than-average cervix. That risks exacerbating the belief that technology is deemed to supersede the wisdom of women’s own bodies, knowledge and experience. Again, I would argue that lots of issues need further exploration.
Although some authors are promoting the idea of routine antenatal CLM screening, along with interventions such as cervical cerclage for women deemed to have short cervices (e.g. Heath et al 1998), others are advising caution. For example, Hoesli et al (2003) argue that the sensitivity and specifity of this test are not within acceptable limits and that many questions remain about whether and how we can prevent preterm labour.
TVU as a Diagnostic Tool
In some areas, TVU is offered to women who are admitted to hospital with threatened preterm labour. Although the use of CLM by TVU has more support in the medical literature as a diagnostic (rather than screening) test, one midwife suggested to me that this “seems a bit like closing the proverbial stable door”.
Where a woman with threatened preterm labour considers whether to undergo CLM by TVU, the most important questions relate to that woman’s particular story. Will the outcome of the test change the options available to the woman, or will it provide her with information that she wishes to have before making a decision about possible treatments? For instance, are the treatments and drugs that are offered to women experiencing threatened preterm labour acceptable to her? Will the findings of TVU change the options that are offered to her, or will she be advised to undergo treatment regardless? Of course, even if she would not consent to medical treatment, she may still like to have as much information as possible.
It is also important to consider any possible risks of TVU as a procedure. In many areas, midwives do not digitally examine women who present with threatened preterm labour; partly because this is deemed to be ‘abnormal’ and thus is the remit of obstetrics, and partly because unnecessary cervical stimulation could cause prostaglandin release and thus intensify the problem. But isn’t there also a risk that the insertion of an ultrasound transducer into a woman’s vagina might introduce a degree of stimulation and exacerbate the very problem that it is being used to diagnose?
Cervical Stories
In an ideal world, the diagnosis of preterm labour arises as the logical conclusion of a number of different pieces of information rather than from one factor. I believe we should be wary about any diagnosis of preterm labour made from the results of TVU alone, not least because women can have very different cervical stories, which can provide more insight into the issue of false positive results. For example:
- Saskia became pregnant as a student midwife, and hadn’t yet perfected the art of vaginal examination, so decided to learn on herself. She soon figured out what she was feeling, and it was fascinating to hear how her cervix changed every few days during the last weeks of her pregnancy; not only shortening, thinning and dilating, but sometimes becoming longer and less dilated than before.
- Sunita, who was both curious and well acquainted with her midwife, was examined every week during the last month of her first pregnancy. Her cervix was 4-5cm dilated every time, and perhaps had been for a while before this. Her baby did not fall out, and she went on to have a normal labour and home birth at term. I often wonder what might have happened had Sunita been “booked” in one of those trusts where cervical length measurement by TVU is becoming routine.
As most midwives will know from their own experience, the behaviour of women’s cervices is somewhat mysterious, marvellously erratic and essentially unpredictable. This is supported by the research of Bergelin and Valentin (2001), whose longitudinal study identified several normal, but different, patterns of change in cervical length and other attributes during the pregnancies of the women they studied. As their research and women’s stories show, no woman’s cervix is a static, solid, immovable organ. Even in non-pregnant women, the cervix dips, rises, opens and closes according to the different phases of the menstrual cycle. The cervix alters position and texture when we make love, and probably does all sorts of other marvellous things that we don’t even know about. What on earth makes us think that pregnancy would suddenly make women’s cervices decide to behave in the ordered ways described by textbooks and graphs?
As far as the prevention of preterm labour is concerned, there is much we need to learn and explore before we can draw any firm conclusions. If I am honest, I am not sure whether we will ever know everything there is to know about the behaviour of the cervix in relation to preterm birth. I do, however, believe that we need to at least attempt to broaden our knowledge in as many areas as possible before we conclude that the cervix is the villain of this picture and routinely use TVU to monitor and control its activity.
This is an older article. For up-to-the-minute updates on new birth-related research and thinking, jump on my newsletter list.
A version of this article was originally published as Wickham S (2005). Controlling the Capricious Cervix Part 1. TPM 9(4):36-37 & Wickham S (2005). Controlling the Capricious Cervix Part 2. TPM 9(5):41-42.
References
Bergelin I and Valentin L (2001) Patterns of normal change in cervical length and width during pregnancy in nulliparous women: a prospective, longitudinal ultrasound study. Ultrasound in Obstetrics & Gynecology, 18: 217
Heath VCF, Southall TR, Souka AP et al (1998a) Cervical length at 23 weeks of gestation: predicttion of spontaneous preterm delivery. Ultrasound in Obstetrics and Gynecology, 12(5): 312 – 317
Heath VC, Souka AP, Erasmus I, et al (1998b). Cervical length at 23 weeks of gestation: the value of Shirodkar suture for the short cervix. Ultrasound Obstet Gynecol 12(5):318-22
Hoesli I, Tercanli S and Holzgreve W (2003) Cervical length assessment by ultrasound as a predictor of preterm labour – is there a role for routine screening? BJOG 110(s20): 61
Jackson GM, Ludmir J, and Bader TJ (1992) The accuracy of digital examination and ultrasound in the evaluation of cervical length Obstetrics & Gynecology 79:214-218
Ozdemir I, Demirci F and Yucel O (2005) Transperineal versus transvaginal ultrasonographic evaluation of the cervix at each trimester in normal pregnant women Australian and New Zealand Journal of Obstetrics and Gynaecology 45(3): 191
Rozenberg P, Gillet A and Ville V (2002) Transvaginal sonographic examination of the cervix in asymptomatic pregnant women: review of the literature. Ultrasound in O & G 19(3): 302
Taipale P, Hilesma V. Sonographic measurement of uterine cervix at 18–22 weeks’ gestation and the risk of preterm delivery. O & G 92: 902–907.
To MS, Skentou C, Cicero S et al (2002) Cervical assessment at the routine 23-weeks’ scan: problems with transabdominal sonography. Ultrasound in O & G 15(4): 292 – 296.