Ultrasonographic Evaluation of the Cervix

Transcription

1 Ultrasonographic Evaluation of the Cervix Transperineal Versus Endovaginal Imaging Barbara S. Hertzberg, MD, Elizabeth Livingston, MD, David M. DeLong, PhD, Patricia J. McNally, RDMS, Constance K. Fazekas, RDMS, RVT, Mark A. Kliewer, MD Objective. This study compares transperineal and endovaginal ultrasonography of the gravid cervix to evaluate image quality and assess for a systematic difference in cervical lengths measured by the 2 techniques. Methods. Transperineal and endovaginal ultrasonography of the cervix was performed on 64 pregnant women. Two physicians reviewed the images and rated the relative diagnostic value of the techniques for assessing the cervix and for evaluating for placenta previa. Cervical length was measured prospectively in both techniques. Data were analyzed to determine if there is a systematic difference in length using the 2 approaches and if length differences are dependent on gestational age. Results. There was a strong reviewer preference for endovaginal ultrasonographic images over transperineal images for both assessing the cervix (P <.001) and evaluating for placenta previa (P <.001). Despite this, transperineal and endovaginal ultrasonographic images were frequently rated as similar in diagnostic quality by both reviewers for depicting the cervix (35.9% of patients) and evaluating for placenta previa (57.8% of patients). The mean length of the cervix was slightly shorter at transperineal ultrasonography (28.4 mm) than at endovaginal ultrasonography (30.1 mm). When cervical lengths were subdivided by gestational age, however, a significant length discrepancy was found only in the 14- to 20-week gestational age range. In this age range, mean cervical length at transperineal ultrasonography (28.6 mm) averaged 5.5 mm less than at endovaginal ultrasonography (34.1 mm). Conclusions. Both transperineal and endovaginal ultrasonography can provide satisfactory images of the cervix, but endovaginal images are frequently superior to transperineal images. Endovaginal ultrasonography should be considered the optimal method for imaging the cervix in most situations. Transabdominal or transperineal ultrasonography can also be used, but if the cervix is not adequately depicted from these perspectives, endovaginal ultrasonography is indicated. Transperineal measurements of cervical length can be significantly shorter than endovaginal measurements, particularly before 20 weeks; therefore, short cervical lengths documented at transperineal ultrasonography before 20 weeks should be confirmed by endovaginal ultrasonography. Key words: pregnancy, ultrasonography; cervix, gravid; endovaginal ultrasonography; transperineal ultrasonography; placenta previa; pregnancy, complications. Received May 29, 2001, from the Departments of Radiology (B.S.H., D.M.D., P.J.M., C.K.F., M.A.K.) and Obstetrics and Gynecology (E.L.), Duke University Medical Center, Durham, North Carolina. Manuscript accepted for publication June 11, We thank Susan Murray for assistance with manuscript preparation. Address correspondence and reprint requests to Barbara S. Hertzberg, MD, Duke University Medical Center, Department of Radiology, Box 3808, Room 2526 Blue Zone S, Durham, NC Ultrasonographic evaluation of the gravid cervix is important for the assessment of cervical incompetence, preterm labor, and placenta previa Although historically the gravid cervix was initially examined from the transabdominal perspective, 1 17 more recently this perspective has been supplemented or even superseded by the transperineal and endovaginal ultrasonographic approaches, both of which have been shown to more consistently visualize the cervix than transabdominal ultrasonography by the American Institute of Ultrasound in Medicine J Ultrasound Med 20: , /01/$3.50

2 Transperineal Versus Endovaginal Imaging of the Cervix There are, however, relatively few studies comparing the transperineal and endovaginal approaches Consequently, the choice between the 2 approaches tends to be based on customary practice, rather than on considered weighing of the diagnostic value of the 2 approaches. This study evaluates the subjective quality of the images obtained by the 2 approaches and assesses if there is a systematic difference in cervical lengths obtained by transperineal and endovaginal ultrasonography to assess the relative roles of transperineal and endovaginal ultrasonography in the obstetrical ultrasonographic examination. Materials and Methods A total of 64 pregnant women underwent ultrasonographic evaluation of the cervix using both transperineal and endovaginal ultrasonography. Criteria for inclusion in the study were initial transabdominal images that failed to adequately demonstrate the cervix or that suggested an abnormality such as cervical shortening, funneling, or placenta previa. Gestational age ranged from 14.0 to 38.2 weeks. All sonograms were obtained by sonographers and sonologists experienced in both transperineal and endovaginal ultrasonography of the cervix. Indications for ultrasonography were as follows: estimate gestational age or evaluate fetal anatomy (n = 29), assess the cervix (n = 12), evaluate fetal growth (n = 11), evaluate placental location (n= 7), advanced maternal age (n = 2), abnormal α-fetoprotein (n = 2), and rule out placental abruption (n = 1). Transperineal sonograms were obtained with 2.5- to 5-MHz multifrequency vector transducers (128XP; Acuson Corporation, Mountain View, CA). With the patient in a supine position, hips elevated above the scanning table by a 6- to 8-in cushion under the buttocks, 33 the transducer was positioned on the labia minora and oriented to optimize sagittal images of the cervix. Endovaginal sonograms were obtained using 5- to 7-MHz single-frequency and multifrequency endovaginal transducers (Acuson 128XP). The transducer was introduced into the vagina with the patient in a supine position and adjusted to optimally demonstrate the cervix. Two physicians experienced in obstetrical ultrasonography reviewed the transperineal and endovaginal images on each patient. The physicians independently rated the diagnostic value of the 2 approaches for assessing cervical length and morphology and for evaluating for placenta previa. The techniques were rated as similar in quality for assessing the cervix if the physician had equal confidence that the entire cervical canal was depicted and accurately measured and that funneling would be evident if present. If instead 1 view inspired greater confidence in the assessment of cervical length or funneling, then that view was rated as superior for assessing the cervix in that patient. Similarly, transperineal and endovaginal images were rated as similar in quality for evaluating for placenta previa if there were comparable levels of confidence that the inner surface of the cervix was demonstrated, the structures overlying the cervix (such as placenta, fetal parts, and amniotic fluid) could be distinguished, and placenta previa could be diagnosed or excluded. If instead a more confident assessment for placenta previa was possible by one of the techniques, that approach was scored as superior in evaluating for placenta previa. In cases in which the physician preferred 1 technique over the other for evaluating the cervix or assessing placenta previa, the reason for the discrepant performance of the 2 techniques was recorded. Interobserver variability in preference ratings for transperineal versus endovaginal ultrasonography for evaluation of the cervix and assessment for placenta previa was analyzed by the κ statistic. The possibility of a relationship between gestational age and observer preference for transperineal versus endovaginal images was evaluated with the Kendall correlation. Signed rank tests were used to test for a correlation between reviewers ratings of the 2 perspectives and the amount of variation in the cervical length measurements between those views. Cervical length was measured prospectively at the time of both transperineal and endovaginal ultrasonography by positioning electronic calipers on the internal and external cervical os (Fig. 1). The internal os was examined for funneling. When funneling was observed, the caliper at the internal os was placed at the caudal limit of the funneled area (Fig. 1) to assess the effective intact cervical length. The cervix was observed during real-time evaluation for at least 2 minutes in each of the 2 perspectives J Ultrasound Med 20: , 2001

3 Hertzberg et al A B Figure 1. Twenty-year-old pregnant woman with funneling of the internal cervical os at 26 weeks 2 days of gestation. Endovaginal (A) and transperineal (B) ultrasonographic images of the cervix demonstrate the technique for measuring cervical length. Electronic calipers are positioned on the external cervical os (curved arrow) and on the internal cervical os at the caudal limit of the funneled segment (straight arrow). Response to transfundal pressure was also documented in both views Previous studies have demonstrated that the minimum length of the cervix obtained over the course of imaging correlates best with the likelihood of preterm delivery. 6,8,10 Therefore, if changes in the length and configuration of the cervix were observed during the examination, the cervical length measurement used for data analysis was the value obtained when the length of the cervix was shortest and the funneling most profound. The average cervical length at both transperineal and endovaginal ultrasonography was calculated. The mean absolute difference in cervical length between the 2 views was determined by averaging the absolute values of the differences in length obtained by the 2 techniques for each patient. Paired t tests were used to assess if there was a systematic difference in cervical length measured from the 2 perspectives. The possibility that length differences between the 2 techniques was dependent on gestational age was investigated with the Kendall correlation. Paired t tests were also performed within each of 4 gestational age ranges (14 20, , , and weeks) to more specifically identify how cervical length discrepancies might change through gestation. The McNemar sign test assessed for a difference in identification of funneling by the 2 techniques. All statistical tests were considered significant at P.05. Results Table 1 shows reviewer preference ratings for the assessment of the cervix. Reviewers strongly preferred endovaginal images over transperineal images in assessing the cervix (P <.001). Endovaginal ultrasonography was considered better for imaging the cervix by at least 1 reviewer in 41 patients (64.1%); the 2 techniques were rated the same for depicting the cervix in 23 patients (35.9%); and transperineal ultrasonography was not preferred by either reviewer for depicting the cervix for any patient (0%). Subdividing reviewer preference ratings by gestational age revealed that endovaginal ultrasonography was preferred over transperineal ultrasonography by at least 1 reviewer in a larger percentage of cases (80%) in the 14- to 20-week gestational age group than in any other age group. Interobserver agreement in reviewer preference ratings for assessing the cervix was Table 1. Endovaginal Versus Transperineal Ultrasonography for Assessing the Cervix: Reviewer Preference Reviewer Preference, n (%) Gestational Both 1 EV, 1 No Both No Age, wk n* EV Preference Preference (60.0) 4 (20.0) 4 (20.0) (35.0) 5 (25.0) 8 (40.0) (45.0) 2 (10.0) 9 (45.0) (25.0) 1 (25.0) 2 (50.0) All gestational ages (45.3) 12 (18.8) 23 (35.9) EV indicates endovaginal. *Number of patients. J Ultrasound Med 20: ,

4 Transperineal Versus Endovaginal Imaging of the Cervix very good (κ = 0.62). The most common reasons cited by reviewers for preferring endovaginal ultrasonography over transperineal ultrasonography for assessing the cervix were that bowel gas obscured the external os and that the location of the cervix, the endocervical canal, or both was not as clearly depicted by transperineal ultrasonography. Less commonly, reviewers noted that shadowing from the pubic symphysis obscured part of the cervix at transperineal ultrasonography. There was likewise a strong reviewer preference for endovaginal images in evaluating for placenta previa (P <.001; Table 2). Endovaginal ultrasonography was preferred by at least 1 reviewer for assessing for placenta previa in 23 cases (35.9%); the 2 techniques were judged to be of similar diagnostic quality for demonstrating the presence or absence of placenta previa in 37 cases (57.8%); and transperineal ultrasonography was preferred by both reviewers in 4 cases (6.3%). There was a statistically significant association between gestational age and the likelihood of a reviewer preference for endovaginal ultrasonography to assess for placenta previa (P <.01): reviewers were more likely to prefer endovaginal ultrasonography earlier in pregnancy than later (Table 2). Interobserver agreement in reviewer preference ratings for evaluating for placenta previa was very good (κ = 0.61). Reasons reviewers commonly cited for preferring endovaginal images over transperineal images for evaluating for placenta previa were improved visualization of the internal surface of the cervix, better discrimination between fetal parts, placental tissue, and amniotic fluid overlying the cervix, and more definitive depiction of the location of the cervix or endocervical canal at endovaginal ultrasonography. In each of the 4 cases in which reviewers preferred transperineal images over endovaginal Table 2. Endovaginal Versus Transperineal Ultrasonography for Evaluating for Placenta Previa: Reviewer Preference Reviewer Preference, n (%) Gestational Both 1 EV, 1 No Both No Both Age, wk n* EV Preference Preference TP (30.0) 5 (25.0) 8 (40.0) 1 (5.0) (10.0) 5 (25.5) 12 (60.0) 1 (5.0) (10.0) 2 (10.0) 16 (80.0) 0 (0) (25.0) 0 (0) 1 (25.0) 2 (50.0) All gestational ages (17.2) 12 (18.8) 37 (57.8) 4 (6.3) EV indicates endovaginal; and TP, transperineal. *Number of patients. images for assessing for placenta previa, the reason cited was improved visualization of the internal surface of the cervix. The mean length of the cervix was slightly shorter at transperineal ultrasonography (28.4 mm) than at endovaginal ultrasonography (30.1 mm) for the entire study group. When length differences were calculated for the pair of measurements in each patient, however, the difference in lengths measured between the 2 techniques was more dramatic: the mean absolute difference in length between transperineal and endovaginal ultrasonography was 5.5 mm. Cervical lengths at transperineal ultrasonography were shorter than the lengths at endovaginal ultrasonography in 36 of 64 cases (56.3%), longer than endovaginal lengths in 27 cases (42.1%), and the same in 1 case (1.6%). The paired t test used to assess for a significant difference in lengths by transperineal and endovaginal ultrasonography across all gestational ages suggested a trend toward shorter lengths at transperineal ultrasonography but failed to reach statistical significance (P =.059). However, when the influence of gestational age was considered, the difference in length at transperineal versus endovaginal ultrasonography was significant (P <.02). Table 3 shows mean length differences between transperineal and endovaginal ultrasonography stratified by gestational age. This format demonstrates that the length discrepancy between the techniques was attributable to the 14- to 20-week gestational age group (P <.006). During the 14- to 20-week gestational age range, the mean cervical length at endovaginal ultrasonography (34.1 mm) was 5.5 mm greater than the mean length at transperineal ultrasonography (28.6 mm). In contrast, there were no significant differences in the mean lengths obtained at transperineal versus endovaginal ultrasonography in the other gestational age groupings, in which mean length differences were less than 1 mm ( weeks, P =.9; weeks, P =.8; and weeks, P =.2). Signed rank tests revealed a strong correlation (P <.0007) between the likelihood that reviewers rated the techniques as different in diagnostic quality and the amount of variation between the cervical length measurements obtained in the 2 views; that is, when reviewers rated endovaginal and transperineal ultrasonography as different in diagnostic quality for evaluating the cervix, there was more variation in the cervical length measurements J Ultrasound Med 20: , 2001

5 Hertzberg et al Cervical funneling was identified in 20 patients (31.3%). Funneling was identified by both the endovaginal and transperineal techniques in 15 patients but was seen only at endovaginal ultrasonography in the remaining 5 patients. Statistical analysis using the McNemar sign test suggested a trend toward the superior identification of funneling by the endovaginal approach, but this difference failed to reach conventional statistical significance (P <.06). Table 3. Comparison of Cervical Lengths at Transperineal Versus Endovaginal Ultrasonography by Gestational Age Gestational Mean Length, mm Mean Absolute Length Age, wk n* TP EV Difference (EV TP), mm EV indicates endovaginal; and TP, transperineal. *Number of patients. Discussion Evaluation of the cervix is an important component of the antenatal obstetrical ultrasonographic examination. Depiction of the cervix is crucial in the ultrasonographic diagnosis of a number of potentially catastrophic conditions, including cervical incompetence, preterm labor, and placenta previa Because transabdominal ultrasonography inconsistently demonstrates the cervix, transperineal and endovaginal ultrasonography are often needed to augment the examination. Both of these approaches have been shown to be superior to the transabdominal assessment and to compare favorably with digital cervical examination, 4,17 28 but which is to be preferred, or can the techniques be used interchangeably? Clearly, each of these ultrasonographic approaches has unique strengths and limitations. The resolution of endovaginal ultrasonographic images tends to be superior to that of transperineal images, because the transducer is closer to the cervix, and higher ultrasonic frequencies can be used. 17,24 The external os may be easier to demonstrate using endovaginal techniques if there is extensive shadowing from bowel gas or pelvic bones from the transperineal perspective. 1,22,35 It has also been suggested that the cervix is particularly difficult to image from the transperineal perspective early in pregnancy. 23,31,32 The transperineal approach may, however, be preferable to endovaginal ultrasonography in patients in whom digital examination of the cervix would be of questionable safety, such as after rupture of the membranes or in patients with active bleeding and suspected placenta previa. 1,3,23 The endovaginal assessment of the cervix can also potentially be compromised by distortion of the cervix due to direct pressure from the probe, 1,24,36 a factor that is not pertinent to transperineal ultrasonography. This study demonstrates that in the majority of patients, endovaginal ultrasonography produces images of superior or equivalent diagnostic quality compared with transperineal ultrasonography. This statement should not, however, be misconstrued to indicate that all transperineal images are inadequate or that the transperineal imaging approach has no place in the evaluation of the gravid cervix. Indeed, in many cases, the 2 approaches afforded similar levels of diagnostic confidence. Despite this, endovaginal images were preferred by at least 1 reviewer for evaluating the cervix in 64.1% of cases and for assessing for placenta previa in 36% of cases. This preference for endovaginal ultrasonography was most pronounced before 20 weeks gestation. Transperineal images were never rated as superior in diagnostic quality for evaluating the cervix and were only rarely preferred in the assessment for placenta previa. Reviewers cited a range of features that accounted for the superior diagnostic quality of endovaginal ultrasonography. Endovaginal ultrasonography provided clearer definition of the cervical boundaries and the endocervical canal, more certain discrimination between fetal parts, placental tissue, and amniotic fluid overlying the cervix, and superior visualization of the internal surface of the cervix, and it was less likely to be compromised by bowel gas obscuring the external os of the cervix. Regardless of the imaging approach used, it is important that the length of the cervix be measured accurately. The length of the cervix is a strong predictor of the likelihood of preterm delivery: the shorter the cervix, the more likely the patient will deliver prematurely. 1,2,4 8,10,11,17,23,24,26 Because the transperineal and endovaginal approaches use different transducers and depict J Ultrasound Med 20: ,

6 Transperineal Versus Endovaginal Imaging of the Cervix the cervix in different orientations, it cannot simply be assumed that cervical measurements obtained in these views will be commensurate: indeed, measurements of the gravid cervix have been shown to be shorter on average at endovaginal ultrasonography than at transabdominal ultrasonography. 25 It is therefore important to know if there is a systematic difference in cervical length between the 2 views. Studies that have addressed this issue previously reported conflicting conclusions Prior results have included poor correlation between transperineal and endovaginal ultrasonography, 31 similar cervical length measurements by the 2 approaches, 29 and shorter lengths at transperineal ultrasonography. 30,32 Our study supports the previous observations of a tendency for cervical lengths to be shorter at transperineal ultrasonography. 30,32 In addition, we stratified cervical length data by gestational age and found that the shorter lengths at transperineal ultrasonography are mainly attributable to the 15- to 20-week subgroup. The mean cervical length difference in this subgroup was 5.5 mm, a discrepancy we consider large enough to be potentially clinically important. We postulate that this measurement discrepancy may stem from the greater likelihood that the caudal portion of the cervix is not depicted in the transperineal view early in the second trimester. The external os of the cervix was often obscured by shadowing from bowel gas from the transperineal perspective and, therefore, was likely to be erroneously located. The length measurement could then exclude some portion of the caudal cervix. Even though mean cervical length was shorter at transperineal ultrasonography than at endovaginal ultrasonography, our study did not detect a systematic relationship that could potentially be applied to cervical lengths measured at transperineal ultrasonography to convert them to equivalent endovaginal lengths. When pairs of measurements on each patient were assessed, cervical lengths at transperineal ultrasonography could be shorter than, longer than, or equal to those at endovaginal ultrasonography. This inconsistency in measurements by the 2 techniques is an observation that has also been made by other investigators. 31,32 Not surprisingly, greater variation in pairs of measurements between transperineal and endovaginal ultrasonography was observed when reviewers rated the transperineal images as inferior in diagnostic quality to the endovaginal images than when the transperineal and endovaginal images were judged to be of similar quality. Presumably when there was suboptimal depiction of the cervix at transperineal ultrasonography, the boundaries of the cervix were misidentified and the cervix length was mismeasured. In conclusion, this study shows that both transperineal and endovaginal ultrasonography can provide adequate images of the cervix in some patients, but endovaginal ultrasonography is more likely to produce superior images than transperineal ultrasonography in the remaining patients. In most situations, endovaginal ultrasonography should be considered the imaging approach of choice for evaluating the cervix. If the transabdominal and transperineal approaches are used instead and the cervix is not optimally depicted from these perspectives, then endovaginal ultrasonography is indicated. Transperineal ultrasonography remains an important adjunctive tool in settings where endovaginal ultrasonography may be undesirable, such as premature rupture of membranes, active bleeding of unknown etiology, or when a patient refuses or cannot tolerate endovaginal ultrasonography. Finally, it should be recognized that measurements of cervical length tend to be shorter from the transperineal perspective than from the endovaginal one, particularly before 20 weeks. Short cervical lengths documented at transperineal ultrasonography before 20 weeks should be confirmed by endovaginal ultrasonography. References 1. Wong G, Levine D. Sonographic assessment of the cervix in pregnancy. Semin Ultrasound CT MR 1998; 19: Watson WJ, Stevens D, Welter S, Day D. Observations on the sonographic measurement of cervical length and the risk of premature birth. J Matern Fetal Med 1999; 8: Sullivan CA. Sonographic evaluation of the uterine cervix. Obstet Gynecol Clin North Am 1998; 25: Gomez R, Galasso M, Romero R, et al. Ultrasonographic examination of the uterine cervix is better than cervical digital examination as a predictor of the 1076 J Ultrasound Med 20: , 2001

7 Hertzberg et al likelihood of premature delivery in patients with preterm labor and intact membranes. Am J Obstet Gynecol 1994; 171: Hasegawa I, Tanaka K, Takahashi K, et al. Transvaginal ultrasonographic cervical assessment for the prediction of preterm delivery. J Matern Fetal Med 1996; 5: Iams JD, Parasko J, Landon MB, Teteris JN, Johnson FF. Cervical sonography in preterm labor. Obstet Gynecol 1994; 84: Iams JD, Goldenberg RL, Meis PJ, et al. The length of the cervix and the risk of spontaneous premature delivery. N Engl J Med 1996; 334: Hertzberg BS, Kliewer MA, Farrell TA, DeLong DM. Spontaneously changing gravid cervix: clinical implications and prognostic features. Radiology 1995; 196: Berghella V, Kuhlman K, Weiner S, Wapner RJ. Cervical funneling: sonographic criteria predictive of preterm delivery. Ultrasound Obstet Gynecol 1997; 10: Murakawa H, Utumi T, Hasegawa I, Tanaka K, Zuzimori R. Evaluation of threatened preterm delivery by transvaginal ultrasonographic measurement of cervical length. Obstet Gynecol 1993; 82: Riley L, Frigoletto FD Jr, Benacerraf BR. The implications of sonographically identified cervical changes in patients not necessarily at risk for preterm birth. J Ultrasound Med 1992; 11: Rana J, Davis SE, Harrigan JT. Improving the outcome of cervical cerclage by sonographic follow-up. J Ultrasound Med 1990; 9: Fried AM. Bulging amnion in premature labor: spectrum of sonographic findings. AJR Am J Roentgenol 1981; 136: Michaels WH, Schreiber FR, Padgett RJ, Ager J, Pieper D. Ultrasound surveillance of the cervix in twin gestations: management of cervical incompetency. Obstet Gynecol 1991; 78: Michaels WH, Montgomery C, Karo J, Temple J, Ager J, Olson J. Ultrasound differentiation of the competent from the incompetent cervix: prevention of preterm delivery. Am J Obstet Gynecol 1986; 154: Feingold M, Brook I, Zakut H. Detection of cervical incompetence by ultrasound. Acta Obstet Gynecol Scand 1984; 63: Sonek J, Shellhaas C. Cervical sonography: a review. Ultrasound Obstet Gynecol 1998; 11: Dawson WB, Dumas MD, Romano WM, Gagnon R, Gratton RJ, Mowbray RD. Translabial ultrasonography and placenta previa: does measurement of the os-placenta distance predict outcome? J Ultrasound Med 1996; 15: Hertzberg BS, Bowie JD, Carroll BA, Kliewer MA, Weber TM. Diagnosis of placenta previa during the third trimester: role of transperineal sonography. AJR Am J Roentgenol 1992; 159: Zilianti M, Azuaga A, Calderon F, Redondo C. Transperineal sonography in second trimester to term pregnancy and early labor. J Ultrasound Med 1991; 10: Mahony BS, Nyberg DA, Luthy DA, Hirsch JH, Hickok DE, Petty CN. Translabial ultrasound of the third-trimester uterine cervix: correlation with digital examination. J Ultrasound Med 1990; 9: Hertzberg BS, Bowie JD, Weber TM, Carroll BA, Kliewer MA, Jordan SG. Sonography of the cervix during the third trimester of pregnancy: value of the transperineal approach. AJR Am J Roentgenol 1991; 157: Mahony BS. Ultrasound of the cervix during pregnancy. Abdom Imaging 1997; 22: Iams JD. Cervical ultrasonography. Ultrasound Obstet Gynecol 1997; 10: Andersen HF. Transvaginal and transabdominal ultrasonography of the uterine cervix during pregnancy. J Clin Ultrasound 1991; 19: Berghella V, Tolosa JE, Kuhlman K, Weiner S, Bolognese RJ, Wapner RJ. Cervical ultrasonography compared with manual examination as a predictor of preterm delivery. Am J Obstet Gynecol 1997; 177: Richey SD, Ramin KD, Roberts SW, Ramin SM, Cox SM, Twickler DM. The correlation between transperineal sonography and digital examination in the evaluation of the third-trimester cervix. Obstet Gynecol 1995; 85: J Ultrasound Med 20: ,

8 Transperineal Versus Endovaginal Imaging of the Cervix 28. Smith RS, Lauria MR, Comstock CH, et al. Transvaginal ultrasonography for all placentas that appear to be low-lying or over the internal cervical os. Ultrasound Obstet Gynecol 1997; 9: Kurtzman JT, Goldsmith LJ, Gall SA, Spinnato JA. Transvaginal versus transperineal ultrasonography: a blinded comparison in the assessment of cervical length at midgestation. Am J Obstet Gynecol 198; 179: Droste S, Trampe B, Martin CB. Comparison of transvaginal and transperineal measurements of cervical length. Paper presented at: International Society of Ultrasound in Obstetrics and Gynecology, 7th World Congress; 1997; Washington, DC. Abstract A 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: Carr DB, Smith K, Parsons L, Chansky K, Shields LE. Ultrasonography for cervical length measurement: agreement between transvaginal and translabial techniques. Obstet Gynecol 2000; 96: Hertzberg BS, Kliewer MA, Baumeister LA, McNally PB, Fazekas CK. Optimizing transperineal sonographic imaging of the cervix: the hip elevation technique. J Ultrasound Med 1994; 13: Guzman ER, Vintzileos AM, McLean DA, Martins ME, Benito CW, Hanley ML. The natural history of a positive response to transfundal pressure in women at risk for cervical incompetence. Am J Obstet Gynecol 1997; 176: Guzman ER, Rosenberg JC, Houlihan C, Ivan J, Waldron R, Knuppel R. A new method using vaginal ultrasound and transfundal pressure to evaluate the asymptomatic incompetent cervix. Obstet Gynecol 1994; 83: Yost NP, Bloom SL, Twickler DM, Leveno KJ. Pitfalls in ultrasonic cervical length measurement for predicting preterm birth. Obstet Gynecol 1999; 93: J Ultrasound Med 20: , 2001