Obstetrics & Gynecology Track the topics, authors and articles important to you
HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS
QUICK SEARCH:   [advanced]


     

Obstetrics & Gynecology 2000;96:554-558
© 2000 by The American College of Obstetricians and Gynecologists
This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by CARR, D. B.
Right arrow Articles by SHIELDS, L. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by CARR, D. B.
Right arrow Articles by SHIELDS, L. E.

ORIGINAL RESEARCH

Ultrasonography for Cervical Length Measurement: Agreement Between Transvaginal and Translabial Techniques

DARCY B. CARR, MD, KARIN SMITH, RDMS, LISA PARSONS, RDMS, KARI CHANSKY, MS and LAURENCE E. SHIELDS, MD

From the Department of Obstetrics and Gynecology, Division of Perinatal Medicine and Diagnostic Imaging at the University of Washington School of Medicine, and the Division of Public Health Sciences, Fred Hutchinson Cancer Research Center, Seattle, Washington.

Address reprint requests to: Laurence E. Shields, MD Division of Perinatal Medicine University of Washington Box 35-6460 Seattle, WA 98195-6460 E-mail: lshields{at}u.washington.edu


   Abstract
Top
Abstract
Materials and Methods
 Results
 Discussion
 References
 
Objective: To assess agreement between transvaginal and translabial ultrasonography for measuring cervical length in the second and third trimesters.

Methods: Eighty-four obstetric patients at 14–40 weeks’ gestation were enrolled. Three images of the cervix were obtained by translabial (3.5-MHz curvilinear transducer) and transvaginal (6.5-MHz endovaginal probe) ultrasonography. The mean cervical length obtained by using each method was used for data analysis. The McNemar {chi}2 test was used to assess the difference between techniques in their ability to obtain a measurement. The paired t-test was used to evaluate the differences between the measurements. The mean difference and SD for the differences were used to calculate the limits of agreement. An acceptable difference was defined as less than 0.5 cm.

Results: Cervical length measurements were obtained in 84 patients (100%) by using the transvaginal technique and in 80 patients (95%) by using the translabial technique (P = .1). Eighty patients had both transvaginal and translabial measurements for comparison. The mean difference in cervical length was 0.37 cm (P < .001; 95% confidence interval [CI] 0.21, 0.52). The upper and lower limits of agreement were 1.75 cm (95% CI 1.48, 2.02) and -1.01 cm (95% CI -0.74, -1.28), respectively. The differences between the two measurements were within these limits 95% of the time. These limits of agreement were greater than the acceptable difference of 0.5 cm.

Conclusion: Transvaginal and translabial techniques should not be used interchangeably for clinical assessment of cervical length because agreement between the methods is not within an acceptable range.

Premature delivery is a leading cause of perinatal morbidity and mortality in the United States.1 Recent studies have demonstrated that cervical length early in pregnancy is an important predictor of subsequent preterm delivery.2,3 Ultrasonographic assessment of cervical length is becoming an increasingly popular component of prenatal care, particularly in high-risk patients.4 Ultrasonographic assessment of cervical length is more objective,5 has greater accuracy,5 and provides additional information about cervical competency6,7 compared with digital examination.

The cervix can be imaged by using transvaginal,3,8–10 transabdominal,8,10,11 and translabial ultrasonography.12–14 Many centers use transvaginal imaging as their accepted standard.8 Translabial ultrasonography is an alternative method of evaluating the cervix that does not require vaginal instrumentation. Several studies have demonstrated that translabial ultrasonographic measurements of cervical length correlate well with those obtained by using digital examination.12,13 However, adequate visualization of the cervix by translabial imaging is not always possible because of maternal obesity or acoustic shadowing from rectal gas or the symphysis pubis.12,14,15 Acoustic shadowing from the rectum or the symphysis pubis during translabial imaging may also limit the accuracy of the measurement by falsely shortening the cervical length. We sought to assess agreement between transvaginal and translabial ultrasonographic measurements of cervical length obtained in the second and third trimesters of pregnancy.


   Materials and Methods
Top
Abstract
 Materials and Methods
Results
 Discussion
 References
 
Study approval was obtained from the Institutional Review Board of the University of Washington Medical Center. Patients presenting to the University of Washington Medical Center after 14 weeks’ gestation were invited to participate in the study. Women were recruited from the radiology department and the Maternal and Infant Care Clinic if they were undergoing ultrasonographic measurement of cervical length for clinical indications. Women were also recruited from the antepartum and the labor and delivery units. The success rate for recruitment was high because most women (91.7%) were already undergoing ultrasonographic measurement of cervical length for clinical indications. Consent for study participation was obtained before ultrasonographic cervical length measurements were performed. Gestational age was determined by menstrual dating and was confirmed by using first-or early second-trimester ultrasonography.

All examinations were performed by registered diagnostic medical sonographers who had at least 3 years of experience in measuring cervical length by using both transvaginal and translabial ultrasonography. Translabial ultrasonography was performed by using a gloved 3.5-MHz curvilinear transducer (Siemens, Issaquah, WA). The gloved transducer surface was covered with a thin layer of lubricant, positioned sagittally against the labia, and rotated until the cervix was visualized. The cervical length was measured by placing the calipers at the internal and external os along the canal of the cervix. Transvaginal ultrasonography was performed by using a 6.5-MHz endovaginal probe (Siemens). The endovaginal probe was covered with a condom and lubricant. The probe was inserted into the vagina with care to avoid compression of the cervix, since cervical compression can falsely increase the cervical length measurement.3 The cervical length was measured by placing the calipers along the length of the canal from the internal to the external cervical os. For each technique, three images were obtained and the mean of the three measurements was used for data analysis. Although the sonographer was not formally masked during the examination, the measurements were not recorded until both techniques had been completed. Occasionally, the examiner could not adequately visualize the entire cervix with the ultrasonographic technique. Patients for whom visualization was inadequate remained in the study to evaluate potential limitations of the technique.

The McNemar {chi}2 test was used to evaluate the frequency of inadequate visualization of the cervix for each technique. The reproducibility of each technique was assessed by using the intraclass correlation coefficient. To determine the degree of correlation between the two measurement techniques, the paired data were analyzed by using the Pearson correlation coefficient. To determine the mean difference between the two techniques in cervical length measurements, the data were analyzed by using a paired two-tailed t-test. Agreement between the techniques was assessed by using the methods described by Bland and Altman.16 These methods involve plotting the difference in cervical length obtained by transvaginal and translabial ultrasonography against the average cervical length as measured by both techniques. Limits of agreement are calculated from the mean differences and standard deviation (SD) for the differences (mean difference ± 2 SD). The differences between the two techniques are within the limits of agreement 95% of the time. We considered a difference of 0.5 cm or greater in cervical length measurement between the two techniques to be clinically unacceptable on the basis of the findings of Iams et al,3 who demonstrated an increase in the relative risk (RR) of preterm delivery as cervical length decreased (RR increased approximately 1.5 times as cervical length decreased by 0.5 cm). Paired t-tests, McNemar {chi}2 tests, and intraclass correlation tests were performed by using SPSS 6.1.1 software (SPSS, Inc., Chicago, IL).


   Results
Top
Abstract
 Materials and Methods
 Results
Discussion
 References
 
Eighty-four pregnant women at 14–40 weeks’ gestation were enrolled in the study between 1995 and 1998. Demographic data and pregnancy characteristics are shown in Table 1Go. Seventy-seven women (91.7%) had one or more of the following clinical indications for cervical ultrasonography: previous preterm delivery (23.8%), symptoms of preterm labor (48.8%), assessment of possible abnormal placentation (10.7%), multifetal gestation (23.8%), or suspected abnormal cervical anatomy (13.1%).


View this table:
[in this window]
[in a new window]
  		Table 1. Demographic Data and Pregnancy Characteristics
 
Cervical length measurements were successfully obtained in 84 patients (100%) by using the transvaginal technique and in 80 patients (95%) by using the translabial technique. Although this finding suggests that transvaginal ultrasonography is superior to translabial ultrasonography, the difference was not significant according to the McNemar {chi}2 test (P = .1). The reproducibility of each technique was calculated by using the intraclass correlation coefficient. The intraclass correlation coefficients for transvaginal and translabial ultrasonographic measurements of cervical length were 0.94 and 0.91, respectively, indicating that both methods had good reproducibility when repeated measurements were performed by the same sonographer. Cervical length measurements obtained by using both ultrasonographic techniques were available in 80 patients. Although correlation between the two techniques was good (r = .77; P < .001) (Figure 1Go), the mean difference between the paired measurements was 0.37 cm (P < .001; 95% confidence interval [CI] 0.21, 0.52). Seventy-one percent of the cervical length measurements obtained by using translabial ultrasonography were smaller than those obtained by using transvaginal ultrasonography (mean difference, 0.6 cm). Twenty-four percent of the cervical length measurements obtained by using translabial ultrasonography were greater than those obtained by using transvaginal ultrasonography (mean difference, 0.4 cm) (Table 2Go).



View larger version (15K):
[in this window]
[in a new window]
  		Figure 1. Mean cervical length measurements obtained by using transvaginal and translabial ultrasonography. The Pearson correlation coefficient was used for comparison.

 

View this table:
[in this window]
[in a new window]
  		Table 2. Discordance in Cervical Length Measurements Between Translabial and Transvaginal Ultrasonography
 
To assess agreement between the techniques, the difference in transvaginal and translabial ultrasonographic measurements of cervical length was plotted against the average cervical length measured by using both techniques (Figure 2Go). A regression line, represented by the solid line in Figure 2Go, had a negative slope, indicating that the transvaginal measurements tended to be greater than the translabial measurements at shorter cervical lengths. However, the slope of the regression line did not statistically differ from zero (P = .2). The mean difference and SD for the differences in cervical length were used to calculate the limits of agreement (mean difference ± 2 SD), represented by the dashed lines in Figure 2Go. Assuming that the differences were normally distributed, differences in cervical length obtained by using the two ultrasonographic techniques would be within this interval 95% of the time. The upper and lower limits of agreement were 1.75 cm (95% CI 1.48, 2.02) and -1.01 cm (95% CI -0.74, -1.28), respectively. Therefore, the differences between the two measurements would be between 1.75 cm and -1.01 cm 95% of the time. The interval was not symmetrical around zero, reflecting the bias toward shorter cervical length when translabial ultrasonography was used.



View larger version (17K):
[in this window]
[in a new window]
  		Figure 2. Bland-Altman plot of the agreement between transvaginal and translabial ultrasonography for cervical length measurement.

 
We considered a 0.5 cm difference or greater to be clinically significant. The limits of agreement were outside of this acceptable range. Forty-two percent of the cervical length measurements disagreed by 0.5 cm or more, and cervical length measurements obtained by using transvaginal ultrasonography were greater than those obtained by using translabial ultrasonography. Ten percent of the measurements disagreed by 0.5 cm or more, and cervical lengths obtained by using translabial ultrasonography were greater than those obtained by using transvaginal ultrasonography.


   Discussion
Top
Abstract
 Materials and Methods
 Results
 Discussion
References
 
Cervical length measurements in pregnancy have gained clinical importance for predicting preterm delivery2,3 and modifying obstetric management.4,17 Cervical length can be measured either digitally or by using transabdominal, transvaginal, or translabial ultrasonography; many investigators favor ultrasonography because of its improved accuracy and reproducibility.5,7,12,13 Jackson et al5 evaluated the accuracy of digital examination and ultrasonography for measuring cervical length in nonpregnant patients before hysterectomy. The investigators concluded that ultrasonography is more accurate because it allows complete visualization of the entire cervical canal; in contrast, part of the cervical canal may not be palpable during digital examination.

Several investigators have evaluated the use of translabial ultrasonography for measurement of cervical length.12–15,18,19 We, like most other investigators,12,13,15,18 noted that the cervix could be adequately imaged in most (86%–95%) patients. Fifty percent of the patients in whom we were unable to adequately visualize the cervix by using translabial ultrasonography were at fewer than 17 weeks’ gestation. Acoustic shadowing from the rectum or symphysis pubis may cause difficulty in using translabial ultrasonography to measure cervical length.12,14,15 This shadowing can artificially shorten the cervix or obscure visualization of the cervix. Techniques have been suggested for reducing these shadowing artifacts, such as placing the patient in a left decubitus position, elevating the patient’s hips, or rotating the transducer obliquely.12,14,15 However, these maneuvers were not always successful in our study population.

We found a correlation (r = .77; P < .001) between transvaginal and translabial ultrasonography for cervical length measurements. This is not surprising, since correlation assesses the linear relation between the measurements obtained by the two techniques. Most reports comparing transvaginal and translabial ultrasonography have also demonstrated good correlation.14,19 However, a recent study by Owen et al18 reported poor correlation (r = .38) between transvaginal and translabial ultrasonographic cervical length measurements. The differences in the findings between these studies may be related to the different gestational ages that were primarily evaluated. Owen et al evaluated women at an earlier gestational age (mean 19.6 weeks, range 15–24 weeks), whereas Kurtzman et al14 evaluated women at 14–34 weeks’ gestation and found good correlation (r = .96). Both investigating centers have expertise with ultrasonographic measurement of cervical length; however, only Kurtzman et al conducted a pilot study of 200 patients before initiating the trial.14

Although we found a correlation between translabial and transvaginal ultrasonographic measurements of cervical length, our analysis revealed disagreement of 0.5 cm or more in 52% of measurements. Our results are similar to the those of Owen et al,18 who found that one third of measurements differed by 20% or more. Poor agreement between transvaginal and translabial ultrasonographic measurements of cervical length may be related to artificial shortening of the cervix by acoustic shadowing during translabial ultrasonography. Our findings differed from those of Kurtzman et al,14 who demonstrated good agreement between the two techniques. They compared a single, most optimal image of cervical length instead of the mean of cervical lengths obtained by using each technique. The image chosen for the transvaginal technique was often the shortest measurement.14 Because of the shortening artifact that can occur with translabial ultrasonography, the optimal cervical length as determined by using that technique was usually the longest.14 We analyzed our data using the means of three images obtained by using each technique. Perhaps averaging all measurements instead of comparing the single best image accounts for the difference in conclusions about the agreement between these techniques. Although we used the average of cervical length measurements for each technique, the variability between measurements for each technique was small (intraclass correlation coefficients were greater than 0.9 for each technique). Furthermore, our findings are similar to those of Owen et al,18 who used the method described by Kurtzman et al14

Our results suggest that translabial ultrasonography is not interchangeable with transvaginal ultrasonography for measurement of cervical length. Translabial ultrasonography should still be considered an imaging option, particularly when vaginal instrumentation is unacceptable or tolerated poorly by the patient. However, clinicians should be aware of the potential pitfalls of this procedure, especially when experience in translabial ultrasonography is limited. Translabial ultrasonography appears be biased toward shorter cervical lengths. However, some of the translabial measurements in our study were longer than the transvaginal measurements, making it difficult to rely on a consistent bias for which the sonographer could adjust when using translabial ultrasonography. More agreement between translabial and transvaginal ultrasonography is needed before these techniques can be used interchangeably to measure cervical length.


   Footnotes
 
PII S0029-7844(00)00973-X

Received January 10, 2000. Received in revised form April 24, 2000. Accepted May 3, 2000.


   References
Top
Abstract
 Materials and Methods
 Results
 Discussion
 References
 
1. Guyer B, Hoyert D, Martin J, Ventura S, MacDorman M, Strobino D. Annual summary of vital statistics—1998. Pediatrics 1999;104: 1229–46.[Abstract/Free Full Text]

2. Andersen H, Nugent C, Wanty S, Hayashi R. Prediction of risk for preterm delivery by ultrasonographic measurement of cervical length. Am J Obstet Gynecol 1990;163:859–67.[Medline]

3. Iams J, Goldenberg R, Meis P, Mercer B, Moawad A, Das A, et al The length of the cervix and the risk of spontaneous premature delivery. National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network. N Engl J Med 1996;334:567–72.[Abstract/Free Full Text]

4. Zalar R. Transvaginal ultrasound and preterm prelabor: A nonrandomized intervention study. Obstet Gynecol 1996;88:20–3.[Abstract]

5. Jackson G, Ludmir J, Bader T. The accuracy of digital examination and ultrasound in the evaluation of cervical length. Obstet Gynecol 1992;79:214–8.[Abstract/Free Full Text]

6. Berghella V, Tolosa J, Kuhlman K, Weiner S, Bolognese R, Wapner R. Cervical ultrasonography compared with manual examination as a predictor of preterm delivery. Am J Obstet Gynecol 1997;177: 723–30.[Medline]

7. Onderglu L. Digital examination and transperineal ultrasonographic measurement of cervical length to assess risk of preterm delivery. Int J Gynaecol Obstet 1997;59:223–8.[Medline]

8. Brown J, Thieme G, Shah D, Fleischer A, Boehm F. Transabdominal and transvaginal endosonography: Evaluation of the cervix and lower uterine segment in pregnancy. Am J Obstet Gynecol 1986; 155:721–6.[Medline]

9. Okitsu O, Mimura T, Nakayama T, Aono T. Early prediction of preterm delivery by transvaginal ultrasonography. Ultrasound Obstet Gynecol 1992;2:402–9.[Medline]

10. Andersen H. Transvaginal and transabdominal ultrasonography of the uterine cervix during pregnancy. J Clin Ultrasound 1991;19: 77–83.[Medline]

11. O’Leary J, Ferrell R. Comparison of ultrasonographic and digital cervical evaluation. Obstet Gynecol 1986;68:718–9.[Medline]

12. Mahony B, Nyberg D, Luthy D, Hirsch J, Hickok D, Petty C. Translabial ultrasound of the third-trimester uterine cervix. Correlation with digital examination. J Ultrasound Med 1990;9:717–23.[Abstract]

13. Richey S, Ramin K, Roberts S, Ramin S, Cox S, Twickler D. The correlation between transperineal sonography and digital examination in the evaluation of the third-trimester cervix. Obstet Gynecol 1995;85:745–8.[Abstract]

14. Kurtzman J, Goldsmith L, Gall S, Spinnato J. Transvaginal versus transperineal ultrasonography: A blinded comparison in the assessment of cervical length at midgestation. Obstet Gynecol 1998; 179:852–7.

15. Hertzberg B, Bowie J, Weber T, Carroll B, Kliewer M, Jordan S. Sonography of the cervix during the third trimester of pregnancy: Value of the transperineal approach. AJR Am J Roentgenol 1991; 157:73–6.[Abstract/Free Full Text]

16. Bland J, Altman D. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1(8476):307–10.[Medline]

17. Heath V, Souka A, Erasmus I, Gibb D, Nicolaides K. Cervical length at 23 weeks of gestation: The value of Shirodkar suture for the short cervix. Ultrasound Obstet Gynecol 1998;12:318–22.[Medline]

18. Owen J, Neely C, Northern A. Transperineal versus endovaginal ultrasonographic examination of the cervix in the midtrimester: A blinded comparison. Am J Obstet Gynecol 1999;181:780–3.[Medline]

19. Berghella V, Kuhlman K, Weiner S, Texeira L, Wapner R. Cervical funneling: Sonographic criteria predictive of preterm delivery. Ultrasound Obstet Gynecol 1997;10:161–6.[Medline]




This article has been cited by other articles:


Home page
 Journal of Diagnostic Medical SonographyHome page
A. M. A. Madsen
Sonographic Methods of Evaluating Cervical Length: A Literature Review
Journal of Diagnostic Medical Sonography, May 1, 2003; 19(3): 157 - 165.
[Abstract] [PDF]

Home page
 J Ultrasound MedHome page
M. Zilianti, B. S. Hertzberg, and M. A. Kliewer
Transperineal Versus Transvaginal Sonography During Pregnancy
J. Ultrasound Med., January 1, 2002; 21(1): 101 - 102.
[Full Text] [PDF]

This Article
Right arrow Abstract Freely available
Right arrow Full Text (PDF)
Right arrow Alert me when this article is cited
Right arrow Alert me if a correction is posted
Services
Right arrow Similar articles in this journal
Right arrow Similar articles in PubMed
Right arrow Alert me to new issues of the journal
Right arrow Download to citation manager
Citing Articles
Right arrow Citing Articles via HighWire
Right arrow Citing Articles via Google Scholar
Google Scholar
Right arrow Articles by CARR, D. B.
Right arrow Articles by SHIELDS, L. E.
Right arrow Search for Related Content
PubMed
Right arrow PubMed Citation
Right arrow Articles by CARR, D. B.
Right arrow Articles by SHIELDS, L. E.

HOME HELP FEEDBACK SUBSCRIPTIONS ARCHIVE SEARCH TABLE OF CONTENTS