- Top of page
Ultrasound assessment of the cervix, in recent decades, has become an important part of obstetric diagnostic imaging, especially since the development of transvaginal probes and the increasing acceptance by patients of transvaginal sonography during pregnancy1, 2.
Transvaginal sonographic measurement of the cervix has emerged as an alternative method for the assessment of cervical length as it allows better quality and more accurate visualization of the uterine cervix and has fewer limitations than does the transabdominal approach3, 4.
Several studies on pregnancy have reported that cervical assessment may provide a useful tool for the prediction of preterm delivery5–8. However, there is no consensus as to what constitutes the normal pattern of change in cervical length during pregnancy. Very few studies have examined cervical changes between the first and second trimesters of pregnancy in relation to preterm delivery. In one study, 89 women at high risk for premature delivery had transvaginal sonographic measurements of the cervical length taken at least twice between the 15th and 24th weeks of gestation. The authors found that the incompetent cervix can begin to shorten at as early as 15 weeks' gestation and that the rates of cervical shortening were greater in the group with an incompetent cervix than in that with a normal cervix9.
The aim of this study was to compare cervical length measurements at the 11–14-week and 22–24-week scans in an unselected group of pregnant women and correlate these measurements with the gestational age at delivery.
- Top of page
This was a prospective study involving 641 pregnant women attending the antenatal clinic of the Obstetric Department of São Paulo University between March 1999 and April 2000. Patients who had an elective preterm delivery or missing outcomes were excluded from the study. A total of 529 women who underwent a routine ultrasound scan at 11–14 weeks and at 22–24 weeks, and for whom the outcome of pregnancy was fully known, were included in the analysis. The scans included fetal examination and the option of having a transvaginal scan to measure cervical length as a screening test for spontaneous preterm delivery. The study had been approved by the Hospital Ethics Committee, and written informed consent was obtained from those who agreed to participate.
For the cervical examination, women were asked to empty their bladders and were placed in the dorsal lithotomy position. Transvaginal sonography with a 5-MHz transducer (Toshiba Sonolayer 77A and Toshiba Eccocce, Toshiba, Tokyo, Japan) was carried out by a sonographer who had received The Fetal Medicine Foundation Certificate of Competence in Cervical Assessment (www.fetalmedicine.com). The probe was placed in the anterior fornix of the vagina and a sagittal view of the cervix as well as visualization of the endocervical mucosa along the length of the canal were obtained. Care was taken to avoid exerting undue pressure on the cervix. The calipers were used to measure the distance between the triangular area of echodensity at the external os and the V-shaped notch at the internal os. Each examination lasted about 3 min to permit the observation of any cervical changes; in such cases, the shortest measurement was recorded. When the cervix was measured in the first trimester of pregnancy, care was taken not to include the lower uterine segment as part of the length of the cervix.
Patients' characteristics, including demographic data and previous obstetric and medical history, were obtained from the patients at their first ultrasound scan and entered into a computer database. Similarly, the ultrasound findings were recorded in the database at the time of the scans. However, the results of the transvaginal scan were not recorded in the patients' and doctors' reports. A routine antenatal care examination was performed by the patients' original doctors. Gestational age was determined from the date of the last menstrual period and confirmed by the measurement of the crown–rump length at the first-trimester scan. Data on pregnancy outcome were obtained from the fetal ultrasound database, hospital charts and patients themselves by telephone contact. Delivery occurring before 37 completed weeks of pregnancy was defined as preterm and that occurring before 33 weeks as extremely preterm.
The mean cervical length was calculated at the 11–14-week and the 22–24-week scans. Student's t-test was used to determine the differences in the cervical lengths at the first and second scans for the group of patients who delivered either at term or preterm. Multiple regression was used to determine the independent contribution of the shortening of the cervix, the cervical length at the 11–14 week and the 22–24 week scans, the past obstetric history and the demographic characteristics, in predicting preterm delivery in women at low and high risk. Wilks's statistics were used to test the differences between the groups. The sensitivity, specificity and positive and negative predictive values for a cervical length of ≤ 20 mm were calculated for spontaneous preterm delivery at 33 and 35 weeks of pregnancy.
- Top of page
The median maternal age for the women involved in this study was 26.2 (range, 13–48) years. There was no significant difference in the mean maternal age, ethnic group or educational level between the group which delivered at term and the preterm group (Table 1).
Table 1. Patients' characteristics in those with preterm and term delivery
|Characteristic||Delivery ≥ 37 weeks||Delivery < 37 weeks||P|
|Maternal age||27.2 ± 6.89||26.3 ± 6.1||0.5002|
| (years, mean ± SD)|| || || |
|Race (n (%))|| || || |
| White||175 (34.6)|| 8 (34.8)||0.837|
| Non-white||331 (65.4)||15 (65.2)|| |
|Educational level (n (%))|| || || |
| High school/superior||116 (23.1)|| 8 (34.8)||0.209|
| Other||390 (76.9)||15 (65.2)|| |
|Cigarette (n (%))|| || || |
| Smoker|| 97 (19.2)|| 2 (8.9)||0.085|
| Non-smoker||409 (80.8)||21 (91.1)|| |
The minimum gestational age at delivery was 26 + 6 weeks and the maximum was 42 weeks; 23 (4.3%) women had a spontaneous preterm delivery. The mean cervical length at the first examination (11–14 weeks) was 42.4 mm and at the second examination (22–24 weeks) it was 38.6 mm (P = 0.0001).
Cervical length at 11 to 14 weeks was not statistically different between the group of women who delivered at term (42.7 mm) and the preterm group (40.6 mm, P = 0.2459). However, at the 22–24-week evaluation, cervical length was significantly different between the two groups of patients (39.3 mm and 26.7 mm, respectively; P = 0.0001).
In the group of patients (n = 10) who had an extremely preterm delivery (before 33 weeks' gestation), the cervix was significantly shorter at the 11–14-week and 22–24-week examinations: 37.6 mm and 22.4 mm, respectively (P = 0.0007).
There was spontaneous shortening of the cervix from the first to the second scan. This shortening was more pronounced in the group of women that delivered prematurely (Figure 1).
Figure 1. Shortening of cervical length from the 11–14-week scan (1) to the 22–24-week scan (2) in the group of pregnant women who had a term (▪——▪) and preterm (◊- - - -◊) delivery.
Download figure to PowerPoint
Forty pregnant women had a previous history of one spontaneous preterm delivery or more. Cervical changes were also evaluated in this group between the first and second examination and compared to those of the group of women with no previous history of preterm delivery (Figure 2). The mean cervical length at the first examination was 42.6 mm and at the second it was 39.1 mm in the group without previous preterm delivery, and 42.8 mm and 35.1 mm in the group with previous preterm delivery, at the 11–14-week and 22–24-week scans, respectively. Therefore, there was a significantly faster decrease in cervical length between the first and the second examinations in the group of patients with a previous history of spontaneous preterm delivery than in that with no previous preterm delivery (P = 0.0001).
Figure 2. Cervical length changes from the first (11–14 weeks) to the second (22–24 weeks) trimester evaluation in a group of patients with previous preterm (◊) and previous term (▪) deliveries.
Download figure to PowerPoint
Multiple regression analysis demonstrated that, in relation to the rate of spontaneous preterm delivery, there were significant independent contributions from cervical length at the 22–24-week scan (P < 0.0001; odds ratio, 0.823) and previous history of preterm delivery (P < 0.0283; odds ratio, 3.45), whereas for an extremely preterm delivery (< 33 weeks) only cervical length was significant. When the shortening of the cervix was considered as a variable, it was statistically significant in predicting preterm delivery (P < 0.0001; odds ratio, 21.3).
Using a cut-off of 35 weeks, a cervical length of ≤ 20 mm had sensitivity, specificity and positive and negative predictive values of 42.3%, 96.7%, 37.9% and 97.2%, respectively, for predicting preterm delivery. For an extremely preterm delivery (< 33 weeks), the same cut-off cervical length had sensitivity, specificity and positive and negative predictive values of 40%, 97%, 23.5% and 98.8%, respectively (odds ratio, 25.9; 95% confidence interval, 4.69–122.67).
- Top of page
We evaluated cervical length in the first (11–14 weeks) and second (22–24 weeks) trimesters of pregnancy in an unselected population of pregnant women. The mean cervical length at 11 to 14 weeks was 42.4 mm and in the second trimester it was 38.6 mm.
In a study by Guzman et al.10, cervical length assessment was performed on 469 women at high risk for preterm delivery between 14 and 24 weeks' gestation on 1265 occasions with serial sonography. The cervical measurement was shorter at 21–24 weeks' gestation than at 15–20 weeks. The mean cervical length observed by Zalar11 in the first trimester (11.3 ± 1.9 weeks) of pregnancy was 46 mm in a low-risk group of pregnant women. In a study involving 166 women at low risk for preterm delivery, cervical length at 8–13 weeks of pregnancy was 43 mm12. These findings are consistent with those of our study.
In another study, cervical length at 15 weeks' gestation was 37 mm among a high-risk group of pregnant women with competent cervices; this is shorter than that observed in our study9. The possible explanation for this is that even in the presence of a competent cervix, the values of cervical length obtained by transvaginal ultrasound are lower in a group of women at high risk for premature delivery.
The length of the cervix in the first trimester observed in our study (42–46 mm) and that observed by Zalar11 are longer than the 38 mm that was the mean cervical length on transvaginal ultrasound examination of 20 premenopausal non-pregnant women13. The longer cervix in the first trimester could be explained by the role of the lower uterine segment in playing the part of the functional cervix in the normal first trimester of pregnancy.
Preterm delivery rates have been stable over the last three decades despite improvements in antenatal clinical care14. The measurement of the cervix by transvaginal ultrasound examination has been shown to be a good marker for the identification of patients at increased risk of premature birth. The shortest cervical length measurement obtained by transvaginal ultrasound should have a better predictive value in establishing the risk for preterm delivery than the average cervical length obtained5, 6.
In the general population the cervical length scan should be performed between 22 and 24 weeks as part of the anomaly scan. Heath et al.6 measured cervical length in 2567 singleton pregnancies at 23 weeks. The risk for delivery before 32 weeks decreased from 78% at a cervical length of 5 mm to 0.5% at 50 mm. The cut-off cervical length for considering elective cerclage was 15 mm.
It might be helpful to perform the cervical examination earlier or even weekly in cases of suspected clinically severe cervical incompetence or in women at high risk of preterm birth, such as those with a history of preterm delivery, late miscarriage, conization, maternal diethylstilbestrol exposure, uterine malformation or multiple pregnancies15, 16.
However, there are few reports on cervical length in the first trimester of pregnancy in relation to preterm delivery. Some studies have begun cervical assessment at 15–18 weeks of pregnancy; they were, however, performed on pregnancies with an increased risk of preterm delivery9, 10, 17.
In our study the mean cervical length of 38.6 mm at the 22–24-week examination is in accordance with the results of other series1, 6, 17. We noticed that the length of the cervix decreased from the first to the second trimester in our unselected study group; this shortening was, however, more pronounced in the group of pregnant women who underwent preterm delivery. Cervical length in the group of patients who delivered at term diminished from 42.7 mm to 39.3 mm. However, in the group of patients who delivered prematurely, the shortening of the cervix was from 40.6 mm to 26.7 mm between the 11–14-week and the 22–24-week evaluations. These findings show that in a group of women who will deliver prematurely, the decrease in cervical length might occur between the first and second trimesters of pregnancy and the rate of shortening might be related to the degree of prematurity. That normal and abnormal rates of cervical shortening may help in the early identification of women who are studied longitudinally for ultrasound signs of cervical incompetence has also been suggested by Guzman et al.9. They described weekly rates of cervical shortening in high-risk women between 15 and 24 weeks of gestation. The ultrasound diagnosis of cervical incompetence was defined as a progressive shortening of the endocervical canal length to ≤ 20 mm before 24 weeks' gestation. Competent cervices had a non-significant rate of endocervical canal length shortening (− 0.3 mm/week) while incompetent cervices had significantly greater endocervical canal length shortening (− 4.1 mm/week; P < 0.001).
In the study presented by Murakawa et al.18 cervical length was measured in 32 women with threatened preterm delivery and in 177 normal singleton pregnancies between 18 and 37 weeks' gestation. Normal cervical shortening during pregnancy was demonstrated in the control group. In the group with threatened preterm labor, the mean cervical length at the time of admission was shorter in the group which had preterm deliveries than in that which had term deliveries (23.2 mm and 31.7 mm, respectively).
In our study there was a spontaneous shortening of the pregnant cervix from the first to the second trimester of pregnancy. This shortening was more rapid in the group of pregnant women who delivered prematurely and who had a previous history of preterm delivery. The evaluation of cervical measurements at 22–24 weeks by transvaginal ultrasound was good for identifying women at increased risk of preterm delivery when the cervix appeared short. However, as cervical length assessment in the first trimester of pregnancy does not seem to be an efficient method for the identification of women at increased risk for preterm delivery, an earlier and serial evaluation of the cervix may be proposed for the high-risk group, to identify increased cervical shortening and therefore, the need of intervention to avoid prematurity.
There are few longitudinal studies of cervical assessment in the literature as a screening test for preterm delivery. Further studies should be undertaken to establish whether the best management procedure for screening for prematurity is a single examination in the second trimester or follow-up scans from the first trimester of pregnancy.