Obstetrics & Gynecology:
Second-Trimester Cervical Sonography: Features Other Than Cervical Length to Predict Spontaneous Preterm Birth
Yost, Nicole P. MD; Owen, John MD; Berghella, Vincenzo MD; MacPherson, Cora PhD; Swain, Melissa RN; Dildy, Gary A. III MD; Miodovnik, Menachem MD; Langer, Oded MD; Sibai, Baha MD; for the National Institute of Child Health and Human Development, Maternal–Fetal Medicine Units Network
*From the Department of Obstetrics and Gynecology, University of Texas Southwestern Medical Center, Dallas, Texas; Department of Obstetrics and Gynecology, University of Alabama at Birmingham, Birmingham, Alabama; Department of Obstetrics and Gynecology, Thomas Jefferson University, Philadelphia, Pennsylvania; The Biostatistics Center, George Washington University, Bethesda, Maryland; Department of Obstetrics and Gynecology, Wake Forest University School of Medicine, Winston-Salem, North Carolina; Department of Obstetrics and Gynecology, University of Utah, Salt Lake City, Utah; Department of Obstetrics and Gynecology, University of Cincinnati, Cincinnati, Ohio; Department of Obstetrics and Gynecology, University of Texas at San Antonio, San Antonio, Texas; and Department of Obstetrics and Gynecology, University of Tennessee, Memphis, Tennessee.
Received September 19, 2003. Received in revised form October 30, 2003. Accepted November 13, 2003.
G.A.D. is now with the Louisiana State University, Baton Rouge, Louisiana; M.M. and O.L. are now with Columbia University, New York City, New York; and B.S. is now with the University of Cincinnati, Cincinnati, Ohio.
Address reprint requests to: Nicole P. Yost, MD, University of Texas Southwestern Medical Center, Department of Obstetrics and Gynecology, 5323 Harry Hines Boulevard, Dallas, TX 75390–9032; e-mail: Nicole.Yost@utsouthwestern.edu.
This study was supported by the following grants from the National Institute of Child Health and Human Development: HD27869, HD21414, HD27860, HD27905, HD19897, HD36801, HD34116, HD34210, HD34208, and HD34136.
OBJECTIVE: To estimate whether cervical and lower uterine segment characteristics other than cervical length and funneling predict recurrent preterm birth.
METHODS: We conducted a secondary analysis of a multicenter, blinded observational study of 181 women with singletons and prior spontaneous preterm births. Endovaginal ultrasonic examinations were performed at 2-week intervals between 16 0/7 weeks and 23 6/7 weeks of gestation. Cervical canal contour (straight/curved), cervical position (horizontal/vertical), posterior cervical width, lower uterine segment thickness, vascularity, endocervical canal dilation, with or without associated membrane prolapse and chorioamnion visible at the internal os, were systematically assessed.
RESULTS: At the initial sonogram (16 0/7–18 6/7 weeks), membranes visible overlying the internal os (relative risk 1.9, confidence interval [CI] 1.2, 3.1) and canal dilation of 2–4 mm (relative risk 2.6, CI 1.4, 4.7) were significant predictors of spontaneous preterm birth of less than 35 weeks in univariate analyses. Only canal dilation remained statistically significant after controlling for cervical length (odds ratio 5.5, CI 1.1, 28.6).
CONCLUSION: Endocervical canal dilation of 2–4 mm during second-trimester endovaginal sonography was associated with an increased risk of recurrent preterm delivery independent of cervical length.
LEVEL OF EVIDENCE: II-2
Sonographic evaluation of the cervix has been widely investigated for the prediction of preterm birth.1 Studies of ultrasonic cervical length measurements in the general obstetric population have shown that the incidence of spontaneous preterm birth increased as cervical length decreased.2 In a cohort study of women with a prior early spontaneous preterm birth, the investigators observed that a second-trimester cervical length less than 25 mm was associated with a 75% recurrence of spontaneous preterm birth at less than 35 weeks of gestation, and that after controlling for cervical length, funneling at the internal os was not a significant predictor.3 The American College of Obstetricians and Gynecologists Committee on Obstetric Practice has concluded that cervical length may be useful in identifying women at high risk for preterm labor.4
Several other characteristics in addition to cervical length can be assessed with second-trimester endovaginal sonography. Our objective was to investigate whether cervical and lower uterine segment characteristics other than cervical length and funneling are useful predictors of recurrent preterm delivery in women with prior preterm births.
MATERIALS AND METHODS
This is a secondary analysis of a previously reported observational study3 conducted between March 1997 and November 1999 at 9 university-affiliated centers, all members of the National Institute of Child Health and Human Development, Maternal–Fetal Medicine Units Network. Inclusion criteria were a singleton pregnancy and at least 1 prior spontaneous preterm birth between 16 and 32 weeks of gestation. Women who received a cervical cerclage for a clinical diagnosis of cervical incompetence were not eligible. Endovaginal sonography was initially performed at 16 weeks to 18 weeks and 6 days of gestation, and subsequent scans were scheduled at 2-week intervals until 23 weeks 6 days. There was a maximum of 4 sonograms per woman. The study protocol was approved by the Institutional Review Board at each center, and informed written consent was obtained from all the women.
The ultrasound examinations were standardized, and images were obtained with the woman recumbent and immediately after voiding. Briefly, the appropriate view for the cervical length measurement in the sagittal plane was obtained by locating the internal os, external os, and endocervical canal. We avoided undue pressure on the cervix by first obtaining a satisfactory image, then withdrawing the probe until the image blurred, and finally reapplying only enough pressure to restore the image. The cervical length was measured from the external os to the functional internal os along a closed endocervical canal (Figure 1). The endocervical canal was measured in 2 linear segments if any subjective curvature was noted. The ultrasound cursor was used to connect the internal and external os with a third vector. The maximum excursion between the intersection of the 2 contiguous segments and the middle of the third vector was measured. If this excursion was 5 mm or greater, then the cervix was defined as curved, and the cervical length measurement was the sum of the 2 contiguous segments (Figure 2). Three cervical length images were acquired, and the shortest measurement associated with the best image was recorded.
In addition to cervical length measurement and assessment of cervical canal contour (curved or straight), the position of the cervix was defined as primarily horizontal or vertical. A cervix was classified as vertical if the endocervical canal appeared to have an angle greater than 45° in relationship to the horizontal plane (axis of the maternal abdomen). In addition, the posterior cervical width, defined as the distance between the midpoint of the endocervical canal and the posterior cervicovaginal interface perpendicular to the endocervical canal (Figure 1), was measured. The anterior lower uterine segment thickness was recorded as the distance between the vesicouterine interface and the amniotic cavity, and the vascularity of the lower uterine segment was subjectively assessed as either present or absent. If the chorioamnionic membrane was distinctly visible as a separate structure overlying the internal os (Figure 3), the membranes were defined as visible. Canal dilation was recorded if the endocervical canal was found to be dilated between 2 to 4 mm along its entire length (Figure 4); canal dilation of 5 mm or more defined a U-shaped funnel and was considered in the primary analysis.3 In the patients with canal dilation, it was noted whether the chorioamnionic membranes prolapsed into the canal.
After baseline assessments were performed, fundal pressure was applied for 15 seconds. Each examination lasted a minimum of 5 minutes, allowing for observation of changes in the cervix that might have occurred spontaneously or as a result of fundal pressure, which prompted repeat measurements. Considering that serial evaluations were performed, it was the shortest ever observed cervical length that was used to assess the relationship between cervical length and preterm delivery. The lower uterine segment thickness and posterior cervical width values used for the serial analysis were the smallest ever observed during the study period. The ultrasound scans were performed by protocol-certified sonographers and were videotaped for quality control. According to the study protocol, the results of each scan were not made available to the patients’ managing physicians except in cases of complete placenta previa or fetal death.
The primary outcome was spontaneous preterm birth before 35 completed weeks of gestation and was coded on the data forms as secondary to labor or premature membrane rupture. In addition to cervical length and funneling, cervical canal contour (straight/curved), cervical position (horizontal/vertical), posterior cervical width, lower uterine segment thickness and vascularity, endocervical canal dilation, with or without associated membrane prolapse and chorioamnion visible at the internal os, were systematically assessed to preterm delivery. Data were analyzed using SAS 8.0 (SAS Institute Inc, Cary, NC). Categorical variables were compared by using χ2 or Fisher exact test, and continuous data were compared with the Wilcoxon rank-sum test. Logistic regression was used for multivariable analysis of spontaneous preterm birth and cervical ultrasound assessments. P < .05 was considered statistically significant.
One hundred eighty-three consenting women were enrolled in the study and underwent a total of 590 sonographic evaluations. Four-hundred sixty-six (79%) of the taped examinations were reviewed for quality control by 2 of the investigators, who were blinded to pregnancy outcome. Examinations reviewed were chosen at random and included all cases of observed canal dilation, funneling, cervical lengths less than 20 mm, spontaneous or fundal-pressure–induced dynamic changes, and cases in which the physician was notified. Two patients were excluded from this analysis because their initial sonogram was performed after 18.6 weeks of gestation. The 181 women in the study had a mean maternal age of 26 years (standard deviation [SD] 5 years); 65% were African American, 14% were white, and 21% were Hispanic. The earliest prior delivery occurred at a mean gestational age of 24 weeks (SD 4.8 weeks); 67% were associated with preterm labor and 33% were associated with preterm membrane rupture. One hundred thirty-four women had a single prior spontaneous preterm birth, 36 had 2 prior spontaneous preterm births, and 10 had more than 2 prior spontaneous preterm births. On review, we determined that 1 patient had not experienced a prior spontaneous preterm birth before 32 weeks but, rather, had experienced an induced preterm birth.
The mean gestational age at delivery was 35.3 weeks (SD 6.2 weeks). A total of 47 (26%) women experienced a spontaneous preterm birth before 35 weeks; 33 (18%) before 32 weeks; 29 (16%) before 28 weeks; and 20 (11%) before 24 weeks. An additional 5 women underwent an indicated preterm delivery at 31–34 weeks of gestation for obstetric indications. Of the 47 spontaneous births before 35 weeks, 33 (70%) were associated with preterm labor, and 14 (30%) were associated with preterm membrane rupture.
Shown in Table 1 are ultrasonic ancillary markers assessed between 16.0 weeks and 18 weeks and 6 days of gestation and their relation to spontaneous preterm birth. Neither lower uterine segment thickness (median 6.5 mm, range 3–17 mm versus 7.0 mm, range 2–27 mm, P = .14) nor posterior cervical width (median 18 mm, range 9–28 mm versus 18 mm, range 9–30 mm, P = .61) measurements were significantly different between women who did and did not experience spontaneous preterm birth at less than 35 weeks. Endocervical canal dilation and membranes visible at the internal os were found to be significant predictors of recurrent preterm delivery in univariate analyses. With multivariable analysis to control for cervical length, these 2 markers were included in a logistic regression model for spontaneous preterm birth at less than 35 weeks of gestation. The odds ratio of preterm delivery for women with canal dilation was 5.5 (95% CI 1.1, 28.6), whereas the odds ratio for membranes visible was 2.0 (95% CI 0.95, 4.0). None of the patients with canal dilation were noted to have chorioamnionic membranes prolapsing into the canal.
The above analyses were then broadened to include the presence of these ultrasonic ancillary markers at any time during the study period (16 weeks to 23 weeks 6 days). Ancillary markers and their relation to preterm birth in these women are shown in Table 2. Neither median lower uterine segment thickness (median 5 mm, range 3–17 mm versus 5 mm, range 2–16 mm, P = .71) nor posterior cervical width (16 mm, range 9–23 mm versus 16 mm, range 9–29 mm, P = .82) measurements were significantly different between women who did and did not experience a preterm delivery. Endocervical canal dilation and membranes visible at the internal os were again found to be significant predictors of recurrent preterm delivery in the univariate analyses. After adjusting for cervical length, the odds ratio for preterm delivery in women with canal dilation was 3.5 (95% CI 1.0, 12.2, P = .048), and the odds ratio for membranes visible was 1.4 (95% CI 0.5, 3.8). Only one of the patients with canal dilation was noted to have chorioamnionic membranes prolapsing into the canal. Of the 14 women who experienced canal dilation, 9 delivered preterm (4 were associated with preterm membrane rupture, and 5 were associated with preterm labor) at a mean gestational age of 29.2 weeks (SD 8.0 weeks).
Our objective was to determine whether any of 8 ancillary cervical and lower uterine segment characteristics, other than cervical length and funneling, assessed during second-trimester endovaginal sonography, were predictive of recurrent preterm delivery. Only one ancillary finding, ie, canal dilation of 2–4 mm, was associated with recurrent preterm delivery after adjusting for cervical length. These results suggest that only cervical length, as originally reported,3 and now reported canal dilation are potentially useful in the prediction of preterm delivery in women undergoing second-trimester endovaginal sonography.
Ample evidence exists of the association between a shortened cervix and shortened gestation.2,3 In a recently published analysis, Owen et al3 found that a cervical length less than 25 mm at 16.0 weeks to 18 weeks 6 days of gestation in women with a prior preterm delivery had a positive predictive value of 75% for delivery before 35 weeks. However, after controlling for cervical length, another ultrasonic finding, funneling, was not a significant independent predictor of preterm birth. The authors postulated that although some cervices likely shorten through a process of funneling, the remaining cervical length is more important than the precise method of cervical shortening. In our investigation, canal dilation, which is a separate entity from funneling, was defined as endocervical canal dilation of 2–4 mm throughout the entire length of the canal. In fact, in this analysis now reported, women with canal dilation at the initial sonographic examination were just as likely to deliver prematurely as women with shortened cervices (< 25 mm). Of note, the technique in this study was well controlled, and the applicability of the results is limited to those ultrasound units using similar established methods.
Canal dilation was observed less frequently than a shortened cervix (4% versus 10%, respectively). This low incidence of endocervical canal dilation may in part be due to the technique in which the images were obtained. Specifically, the evaluations in this study were performed in the sagittal plane, the traditional view for cervical sonography. In one investigation, the authors reported that 67% of the women with a cervix that appeared undilated in the sagittal ultrasonic image plane were actually clinically dilated at least 2 centimeters.5 These investigators found a better correlation between digital and ultrasonic findings when the vaginal probe was oriented in the transverse plane.
The remaining 7 ancillary cervical and lower uterine segment characteristics evaluated were not found to be significantly associated with recurrent preterm delivery after adjusting for cervical length. The rationale for analyzing these markers was the possible association they may have with preterm birth. For example, membranes visible as a separate entity over the internal os and membranes that protrude into the canal represent separation of the membranes from the underlying cervix. This could be a marker of inflammation or simple mechanical disruption of the chorioamnion-decidual interface, allowing for the possible ascent of organisms into the amniotic cavity, resulting in an increased risk for preterm delivery. This theory was not confirmed in our investigation. Posterior cervical width was thought to be a surrogate for cervical size, the hypothesis being that a smaller or more hypoplastic cervix would function poorly compared with a thicker cervix. This does not appear to be the case. In summary, of the 8 ancillary cervical and lower uterine segment characteristics evaluated, it appears that only cervical canal dilation discovered during second-trimester endovaginal sonography increases a woman's risk of a preterm delivery independent of cervical length.
1.Iams JD. Cervical ultrasonography. Ultrasound Obstet Gynecol 1997;10:156–60.
2.Iams JD, Goldenberg RL, Meis PJ, Mercer BM, Moawad A, Das A, et al, and the National Institute of Child Health and Human Development Maternal Fetal Medicine Unit Network. The length of the cervix and the risk of spontaneous premature delivery. N Engl J Med 1996;334:567–72.
3.Owen J, Yost N, Berghella V, Thom E, Swain M, Dildy GA, et al, for the National Institute of Child Health and Human Development, Maternal–Fetal Medicine Units Network. Mid-trimester endovaginal sonography in women at high risk for spontaneous preterm birth. JAMA 2001;286:1340–8.
4.Assessment of risk factors for preterm birth. ACOG Practice Bulletin No. 31. American College of Obstetricians and Gynecologists. Obstet Gynecol 2001;98:709–16.
5.Yost NP, Bloom SL, Twickler DM, Leveno KJ. Pitfalls in ultrasonic cervical length measurement for predicting preterm birth. Obstet Gynecol 1999;93:510–6.
Other members of the National Institute of Child Health and Human Development Maternal–Fetal Medicine Units Network are as follows: University of Alabama at Birmingham: C. Neely, A. Northern, J. Hauth, D. Thom; University of Chicago: A. Moawad; University of Cincinnati: N. Elder, T. Haskins, D. Schultz; George Washington University Biostatistics Center: E. Thom, S. Leindecker; Magee Women's Hospital: S. Caritis; University of Miami: M. O'Sullivan; National Institute of Child Health and Human Development: C. Catz, S. Yaffe, C. Spong; Ohio State University: J. Iams; University of Tennessee: R. Ramsey, M. Peterson, J. Fricke, J. Livingston; University of Texas at San Antonio: S. Barker, C. Leja; University of Texas Southwestern Medical Center: K. Leveno, J. McCampbell, R. Benezue; Thomas Jefferson University: M. DiVito, R. Wapner, G. Bega; University of Utah: M. Varner, E. Taggart, R. Zollinger; Wake Forest University: P. Meis, A. Henshaw; Wayne State University: M. Dombrowski.
This article has been cited 12 time(s).
Ultrasound in Obstetrics & GynecologyTransvaginal sonographic measurement of cervical length to predict preterm birth in asymptomatic women at increased risk: a systematic reviewUltrasound in Obstetrics & Gynecology
Obstetrics and Gynecology
Prediction and prevention of recurrent spontaneous preterm birth
Obstetrics and Gynecology, 110(2):
Obstetrics and Gynecology Clinics of North AmericaRecurrent risk of adverse pregnancy outcomeObstetrics and Gynecology Clinics of North America
Seminars in PerinatologyPrevention of Preterm Birth Based on a Short Cervix: CerclageSeminars in Perinatology
American Journal of Obstetrics and GynecologySonographic myometrial thickness predicts the latency interval of women with preterm premature rupture of the membranes and oligohydramniosAmerican Journal of Obstetrics and Gynecology
American Journal of PerinatologyFrom concept to practice: The recent history of preterm delivery prevention. Part I: Cervical competenceAmerican Journal of Perinatology
Journal of Maternal-Fetal & Neonatal MedicineMethods of sonographic cervical length measurement in pregnancy: A review of the literatureJournal of Maternal-Fetal & Neonatal Medicine
International Journal of Gynecology & ObstetricsPredictors of preterm birthInternational Journal of Gynecology & Obstetrics
Bjog-An International Journal of Obstetrics and GynaecologyDiagnosis of early preterm labourBjog-An International Journal of Obstetrics and Gynaecology
Clinical Obstetrics and GynecologyThe 20-week Scan:Beyond Biometry and AnatomyClinical Obstetrics and Gynecology
Current Opinion in Obstetrics and GynecologyThe short and funneling cervix: when to use cerclage?Current Opinion in Obstetrics and Gynecology
© 2004 The American College of Obstetricians and Gynecologists