Preterm birth: the value of sonographic measurement of cervical length
Prof KH Nicolaides, Harris Birthright Research Centre for Fetal Medicine, King’s College Hospital Medical School, Denmark Hill, London SE5 8RX, UK. Email email@example.com
Transvaginal ultrasound scanning of cervical length at approximately 20 weeks of gestation in women attending for routine antenatal care is useful for predicting the likelihood of spontaneous early preterm birth. The risk of early birth increases exponentially with decreasing cervical length in both singleton and multiple pregnancies. In such women, individualisation of risk would lead to rationalisation of antenatal care, including frequency of visits, patient education in recognising and reporting symptoms of spontaneous preterm labour and timely administration of steroids. It is also possible that in women identified as being at high risk, the rate of preterm birth might be reduced by the prophylactic use of progesterone. In women presenting with threatened spontaneous preterm labour, transvaginal measurement of cervical length provides a useful distinction between those who are likely to deliver within the subsequent 7 days and those who are not. Since only 10–20% of such women are in true spontaneous preterm labour, the cervical length measurement in rational care can avoid the current practice of hospitalisation and administration of steroids and tocolytics to all. This article reviews the evidence in support of the clinical introduction of transvaginal sonography for both the prediction and management of spontaneous preterm labour.
Transvaginal ultrasound scanning (TVUSS) of cervical length at mid-gestation provides a useful method for predicting the likelihood of subsequent preterm birth. In women who present with threatened spontaneous preterm labour, TVUSS of cervical length can help distinguish between true and false spontaneous preterm labour. This article reviews the evidence that supports the clinical introduction of TVUSS in both the prediction and management of spontaneous preterm labour.
Technique of sonographic measurement of cervical length
The best approach for measurement of cervical length is by TVUSS. With the transabdominal approach, the cervix may not be visualised in up to 50% of cases unless the bladder is full, but bladder filling significantly increases the length of the cervix.1 The transperineal route is limited both by the inconsistency in correlation between transvaginal and transperineal measurements and the inadequate visualisation of the cervix in up to 25% of cases.1 For TVUSS, the woman is asked to empty her bladder, placed in the dorsal lithotomy position and the transvaginal probe is placed in the anterior fornix of the vagina. A sagittal view of the cervix is obtained, and the calipers are used to measure the distance between the triangular area of echodensity at the external os and the V-shaped notch at the internal os.1 Each examination should be performed over a period of about 3 minutes. In about 1% of cases, dynamic cervical changes, due to uterine contractions, are observed. In such cases, the shortest measurement is recorded. TVUSS of cervical length is highly reproducible, and in 95% of occasions, the difference between two measurements by the same observer and by two observers is 3.5 mm or less and 4.2 mm or less, respectively.2
TVUSS assessment in asymptomatic pregnancies
The traditional method of antenatal screening for spontaneous preterm labour and preterm birth is based on maternal characteristics, such as age, race and smoking status, and obstetric history. Risk scoring systems, which attempt to define women as being at high or low risk according to these maternal factors, have been shown to have a low detection rate and high false-positive rate. Data extracted from a recent systematic review of the literature showed that with the most commonly used risk scoring system, the detection rate of spontaneous preterm labour and preterm birth was 38% for a false-positive rate of 17%.3 An alternative method to identify high-risk women is TVUSS of cervical length at 20–24 weeks of gestation, and several studies have showed that the risk of preterm birth is inversely related to the length of the cervix.4–6 Combined data from the three largest studies involving a total of 7861 women showed that the detection rate of birth before 35 weeks was 34% for a false-positive rate of about 5%.4–6
To et al.7 conducted a population-based prospective multicentre study in 39 284 women with singleton pregnancies attending for routine hospital antenatal care in London, UK. The cervical length, measured by TVUSS at 22–24 +6/7 days, was normally distributed with a mean of 36 mm. The length was 15 mm or less in about 1% of women. The probability of spontaneous preterm labour and preterm birth was influenced by maternal age and obstetric history and was inversely related to cervical length. The detection rate of spontaneous preterm labour and preterm birth before 32 weeks, for a fixed false-positive rate of 5%, when screening by maternal factors alone, cervical length alone and by the combination of maternal factors and cervical length was 29, 48 and 57%, respectively, and the respective values for a fixed false-positive rate of 10% were 38, 55 and 69%, respectively.
The concept of combining factors from the maternal history with the findings of special investigations in the current pregnancy is increasingly being applied in many aspects of obstetric care. When screening for fetal trisomy 21, maternal age and history of previously affected pregnancies is combined with sonographic findings from examination of the fetus and maternal serum biochemical testing. Similarly, in screening for early-onset pre-eclampsia, the addition of sonographic measurement of impedance to flow in the uterine arteries to the traditional approach of eliciting factors in the maternal history substantially improves the detection rate.
Cervical length and FFN
Two screening studies in asymptomatic women with singleton pregnancies have examined the value of a short cervix and positive fetal fibronectin (FFN) test at 22–24 weeks of gestation in the prediction of spontaneous preterm labour and preterm birth before 32 and 33 weeks, respectively.8,9 Goldenberg et al.8 examined 2915 pregnancies at 24–26 weeks, and Heath et al.9 examined 5146 women at 22–24 weeks. A positive FFN test was reported in 6.6 and 3.5% of women, respectively, and cervical length of 25 mm or less in 9.1 and 8.4%, respectively. Heath et al.9 reported that FFN positivity increased exponentially with decreasing cervical length, from 3 to 19% and 57% for respective cervical lengths of 31–40 mm, 11–15 mm and 0–5 mm. Similarly, in Goldenberg et al.,8 the incidence of a positive FFN test was 6% in those women with a cervical length of more than 25 mm and 16% for a cervical length of 25 mm or less. In these studies, a short cervix and a positive FFN test provided significant independent contributions for the prediction of preterm birth with similar odds ratios (12.5 versus 12.2, and 8.7 versus 9.8 in nulliparous women and 10.0 versus 4.6 for multiparous women).
In twin pregnancies, the rate of early preterm birth is 5–10%, compared with 1–2% in singletons.10 Furthermore, due to delayed childbirth and widespread uptake of assisted conception techniques, the rate of twin pregnancies is increasing rapidly, and it is now about 2%.
Several studies have showed that in twin pregnancies, the rate of spontaneous preterm labour and preterm birth can be predicted from TVUSS of cervical length at 21–24 weeks. In the largest study, cervical length was measured at 22–24 weeks in 1163 twin pregnancies attending for routine antenatal care.11 The rate of delivery before 32 weeks was inversely related to cervical length, being 66% for 10 mm, 24% for 20 mm, 12% for 25 mm and less than 1% for 40 mm. The rate of preterm birth was also strongly influenced by the obstetric history, such that, in women with a previous preterm birth, the risk is twice as high as that in nulliparous women, whereas in those with a previous term birth, the risk is half. The median cervical length was 35 mm and the length was 25 mm or less in about 16% of women, 20 mm or less in 8% and 15 mm or less in 5%. Using these cutoff values, the respective detection rates of spontaneous preterm labour and preterm birth before 32 weeks were 67, 49 and 35%. A single measurement of cervical length at mid-gestation provides sensitive prediction of spontaneous preterm labour and preterm birth and permits individualisation of risk for this pregnancy complication.
Cervical length and FFN
A study of 101 asymptomatic women with twin pregnancies examined the value of a short cervix and FFN positivity at 24 weeks of gestation in the prediction of spontaneous preterm labour and preterm birth before 35 weeks of gestation, which occurred in 22% of the women. FFN was present in 15% of the women. The incidence of spontaneous preterm labour and preterm delivery was 54% in the FFN-positive group and 16% in those with a negative result. Cervical length was a poor predictor of preterm birth, and a cervical length of less than 25 mm had no predictive value for preterm birth.12 Gibson et al.13 published a study of 91 twin pregnancies in which a short cervix of less than 22 mm at 24 weeks increased the risk of preterm birth prior to 35 weeks by ten times. In contrast, FFN at 24 weeks was not useful in predicting preterm birth.
TVUSS assessment in women in spontaneous preterm labour
In pregnancies presenting with spontaneous preterm labour, the outcome measure of relevance to clinical management is delivery within the subsequent 7 days, rather than preterm birth as such. The clinical dilemma revolves around the issue of whether the woman is in true spontaneous preterm labour, requiring hospitalisation in a unit with facilities for neonatal intensive care and administration of tocolytics with the objective of prolongation of pregnancy for the effective administration of corticosteroids to improve fetal lung maturity.
It could be argued that, in the absence of reliable methods to distinguish between true and false spontaneous preterm labour, the high mortality and handicap rates associated with preterm birth justify treatment of all women with spontaneous preterm labour because such risks outweigh the economic cost of hospitalisation, the maternal risks associated with tocolytics and the potential fetal risks associated with corticosteroids.
This concept of ‘treatment for all’ has been challenged by evidence that TVUSS of cervical length at presentation can help distinguish between those women who do and those who do not deliver within the subsequent 7 days. In the combined data from three sonographic studies in a total of 510 singleton pregnancies presenting with spontaneous preterm labour, preterm birth within 7 days occurred in 49% of those women with a cervical length less than 15 mm and in 1% of those with cervical length of 15 mm or more.14 Hospitalisation and the administration of tocolytics and steroids should be reserved for women who are truly in labour, and these women can be identified by TVUSS of cervical length at presentation.
Cervical length and FFN
Two studies investigated the effect of combining cervical length and FFN in the prediction of delivery within 7 days of presentation.15,16 Tsoi et al.15 examined 195 singleton pregnancies with spontaneous preterm labour at 24–36 weeks of gestation. FFN positivity in cervicovaginal secretions was determined, and TVUSS of cervical length was carried out. Birth within 7 days occurred in 51% of the 35 women with a cervical length of less than 15 mm and 0.6% of the 160 women with a cervical length of 15 mm or more, in 21% of the 85 women with a FFN-positive result and in 1% of the 110 women with a FFN-negative result. Gomez et al.16 examined 215 pregnancies. Birth within 7 days occurred in 57% of the 30 women with a cervix of less than 15 mm and in 5.9% of the 185 women with a long cervix. The respective values for a positive and negative FFN test were 35 and 6.2%.
Cervical length and microbial invasion of the amniotic fluid
In a study by Gomez et al.,17 amniocentesis and TVUSS were carried out in singleton pregnancies admitted in spontaneous preterm labour to examine the association between short cervix and microbial invasion in the amniotic cavity as an underlying mechanism for chorioamnionitis and fetal inflammatory response syndrome. Microbial invasion, mainly Ureaplasma urealyticum, was found in 26% of the women with a cervical length of less than 15 mm, compared with 4% in those with a cervix of more than 15 mm and 1.9% when the cervix length was more than 30 mm. Hassan et al.18 confirmed this finding in asymptomatic women. In a retrospective study of 57 asymptomatic pregnancies at 14–24 weeks with a short cervix of less than 25 mm and amniocentesis, microbial invasion was found in 9% of the women. Of those, 40% delivered before 32 weeks.
Cervical length and ruptured membranes
In women with preterm prelabour rupture of the membranes (PPROM), approximately one-third will have evidence of intrauterine infection, and the incidence is inversely related to the gestation at PPROM. The interval between PPROM and birth is substantially shorter in women with evidence of intrauterine infection than in those with no infection. In women with no infection, there is an inverse correlation between gestation at membrane rupture and the interval to delivery.19,20 Delivery within 7 days of PPROM occurs in about 60% of women.21
Three studies have examined the relationship between TVUSS of cervical length and the latency period in PPROM. Rizzo et al.22 examined 92 women and reported that the median interval to delivery was 2 days in those with a cervical length below 20 mm, compared with 6 days in those with a longer cervix. In contrast, Carlan et al.23 examined 45 women and found no significant difference in the latency period between those with a cervical length of more than 30 mm and those with a shorter cervix at presentation. Tsoi et al.24 measured the cervical length in 101 women with singleton pregnancies presenting with PPROM at 24–36 weeks of gestation. Delivery within 7 days of presentation occurred in 57% of the pregnancies, and a significant independent contribution in the prediction of such delivery was provided by cervical length, gestation and presence of contractions at presentation.
In women presenting with PPROM, clinical examination is directed at the confirmation of membrane rupture but is not useful in assessing the likelihood of delivery within the next few days.25 Furthermore, there is evidence to suggest that digital vaginal assessment of the cervix shortens the latency period and is therefore contraindicated in the absence of active labour. In contrast, TVUSS has been proven to be safe in women with PPROM.23 In the management of such women, prediction of the risk of delivery within the subsequent 7 days can help optimise the neonatal care for the potentially preterm infant, through referral to a specialist centre. In PPROM, unlike spontaneous preterm labour with intact membranes, measurement of cervical length needs to be combined with other parameters in order to derive a reliable risk.
Cervical length in twin pregnancies
Crane et al.26 examined TVUSS of cervical length after spontaneous or tocolytic-induced arrest of contractions in 26 twin pregnancies presenting with spontaneous preterm labour at 23–33 weeks of gestation. The outcome measure of the study was preterm birth before 34 weeks. This occurred in 5 of the 14 women (36%) with a cervical length of less than 30 mm and in none of the 12 women with a cervical length of 30 mm or more.
Fuchs et al.27 examined 87 twin pregnancies presenting with spontaneous preterm labour and preterm birth at 24–36 weeks of gestation. Delivery within 7 days of presentation occurred in 22% of the pregnancies, and the incidence of such delivery was inversely related to cervical length. The threshold in cervical length that distinguishes between false and true labour in twin pregnancies presenting with spontaneous preterm labour was 25 mm, compared with 15 mm in singleton pregnancies.
The findings of these studies suggest that TVUSS of cervical length in twin pregnancies presenting with spontaneous preterm labour can define the risk of preterm birth within 7 days and may help distinguish between true and false spontaneous preterm labour.
TVUSS of cervical length is highly reproducible and is acceptable to women. Routine measurement of cervical length at 20–24 weeks of gestation provides a sensitive prediction of preterm birth. The risk of such preterm birth increases exponentially with decreasing cervical length. In singleton pregnancies, the threshold cervical length for the exponential increase in risk is 15 mm and in twins it is 25 mm.
In pregnancies presenting with spontaneous preterm labour, the outcome measure of relevance to clinical management is delivery within the subsequent 7 days, rather than preterm birth as such. Measurement of cervical length at presentation provides a useful distinction between those who deliver within 7 days and those who do not. Furthermore, the prediction is better with measurement of cervical length than that with assessment of cervicovaginal FFN. The threshold in cervical length that distinguishes between false and true spontaneous preterm labour is 15 mm in singleton pregnancies and 25 mm in twin pregnancies.
This study was supported by a grant from the Fetal Medicine Foundation (Charity No. 1037116).