Advertisement

Transvaginal ultrasound is superior to transabdominal ultrasound in the identification of a short cervix

      Objective

      Many centers have implemented universal transabdominal cervical length (TACL) screening, despite evidence of poor reproducibility.
      • Khalifeh A.
      • Quist-Nelson J.
      • Berghella V.
      Universal cervical length screening for preterm birth prevention in the United States.
      Our objective was to evaluate the technical limitations that may reduce the effectiveness of 1 such program.

      Study Design

      We performed a cross-sectional study of singleton pregnancies between 160/7 and 236/7 weeks gestation. We defined eligible patients as those with singleton gestations without traditional risk factors for spontaneous preterm birth. In the preexposure period (June–December 2016), TACL was performed without a protocol for reflex transvaginal scanning. In the postexposure period (June–December 2017), routine TACL was performed according to the methods described by Saul et al,
      • Saul L.L.
      • Kurtzman J.T.
      • Hagemann C.
      • Ghamsary M.
      • Wing D.A.
      Is transabdominal sonography of the cervix after voiding a reliable method of cervical length assessment?.
      and reflex transabdominal cervical length (TVCL) ultrasound scans were performed if the TACL measured <30 mm or transabdominal views were limited. The primary outcome was the proportion of reflex TVCL ultrasound scans, which were defined as TVCL ultrasound scans that were performed in response to suspected shortened or suboptimal TACL. Secondary outcomes included the number of TACL <30 mm and various obstetric outcomes. Six sonographers participated in a reproducibility study in which we calculated intraclass correlation coefficients to evaluate inter- and intrarater agreement of TACL measurements. The study was approved by the Institutional Review Board.

      Results

      Complete data were available for 616 patients in the preexposure group and 669 patients in the postexposure group. Although there was a small difference in body mass index (median, 28.6 kg/m2 [interquartile range, 25.0–33.5 kg/m2] vs median, 25.8 kg/m2 [interquartile range, 22.3–30.2 kg/m2]; P<.001), there were no other differences in demographics (Table). In the postexposure group, 34 patients (5.1%) had transvaginal ultrasound scans. Almost one-half of transvaginal scans (16/34; 47.1%) were performed reflexively; 9 scans were performed because of TACL <30 mm, and 7 scans were due to poor visualization of TACL landmarks. The other transvaginal scans were performed for routine obstetric indications (eg, placental assessment). Only 2 patients (0.3%) in the postexposure group were found to have TVCL <25 mm. There were no differences in secondary outcomes. The intraclass correlation coefficients for TACL measurements was 0.15 (95% confidence interval, 0.04–0.45); the within-subject variance was 0.11 (95% confidence interval, 0.10–0.13).
      TablePatient demographic characteristics and select outcomes
      VariablePreexposure period (June–December 2016; n=616)Postexposure period (June–December 2017; n=669)P value
      Determined by Chi square or Fisher’s exact tests (categoric data) and Wilcoxon rank-sum test (for nonparametric continuous data)
      Demographic characteristics
      Data are given as median (25th–75th percentile).
       Age, y30 (26–34)30 (27–34).78
       Gravidity2 (1–4)2 (1–3).42
       Parity, preterm births0 (0–0)0 (0–0).74
       Gestational age at scan, wk20 (20–20)20 (20–20).99
       Body mass index, kg/m228.6 (25–33.5)25.8 (22.3–30.2)<.001
      Primary outcome: reflex transvaginal cervical length, n (%)5 (0.8)16 (2.4).03
      Secondary outcomes
       Transabdominal cervical length, n (%)
      N534 (86.7)624 (93.3)<.001
      <35 mm236 (38.3)250 (37.4).12
      <30 mm22 (3.6)9 (1.3).004
       Transvaginal cervical length, n (%)
      N25 (4.1)34 (5.1).38
      <25 mm2 (0.3)2 (0.3).57
       Vaginal progesterone, n (%)3 (0.5)2 (0.3).46
       Gestational age at delivery, wk
      Data are given as median (25th–75th percentile).
      39.1 (38.3–40.0)39.1 (38.3–40.0).72
       Preterm birth <37 weeks gestation, n (%)56 (9.1)64 (9.6).77
       Spontaneous preterm birth, n (%)
      <37 wk25 (4.1)30 (4.5).71
      <32 wk6 (0.97)6 (0.9).89
      Jayakumaran. Transvaginal ultrasound scan to identify short cervix. Am J Obstet Gynecol 2019.
      a Determined by Chi square or Fisher’s exact tests (categoric data) and Wilcoxon rank-sum test (for nonparametric continuous data)
      b Data are given as median (25th–75th percentile).

      Conclusion

      Cervical length screening is important to implement treatment with vaginal progesterone or other interventions to reduce preterm births.
      • Romero R.
      • Conde-Agudelo A.
      • Da Fonseca E.
      • et al.
      Vaginal progesterone for preventing preterm birth and adverse perinatal outcomes in singleton gestations with a short cervix: a meta-analysis of individual patient data.
      Compared with universal TVCL screening, the TACL screening program substantially reduced the frequency of transvaginal scanning. However, TACL screening identified fewer cases of short cervices than expected, and TACL measurements were poorly reproducible. This study underscores that many short cervices were missed
      • Hernandez-Andrade E.
      • Romero R.
      • Ahn H.
      • et al.
      Transabdominal evaluation of uterine cervical length during pregnancy fails to identify a substantial number of women with a short cervix.
      and that the poor intraclass correlation coefficients for TACL measurements were a likely explanation.
      Most studies that investigate TACL screening have performed transvaginal ultrasound scans concurrently and have highlighted the challenges that are associated with transabdominal scanning. These studies focused on test performance in ideal conditions and not in real-world scenarios.
      • Stamilio D.
      • Carlson L.M.
      Transabdominal ultrasound is appropriate.
      Accordingly, our study provides information about the impact of universal TACL screening using a 30-mm cut-off point in clinical practice, and the results raise important questions about the technical limitations that reduce the effectiveness of TACL screening. Until there are more convincing data to support TACL screening, we urge providers to consider the possibility that the primary benefit of universal TACL screening is the avoidance of transvaginal scanning rather than the identification of high-risk patients.

      References

        • Khalifeh A.
        • Quist-Nelson J.
        • Berghella V.
        Universal cervical length screening for preterm birth prevention in the United States.
        J Matern Fetal Neonatal Med. 2017; 30: 1500-1503
        • Saul L.L.
        • Kurtzman J.T.
        • Hagemann C.
        • Ghamsary M.
        • Wing D.A.
        Is transabdominal sonography of the cervix after voiding a reliable method of cervical length assessment?.
        J Ultrasound Med. 2008; 27: 1305-1311
        • Romero R.
        • Conde-Agudelo A.
        • Da Fonseca E.
        • et al.
        Vaginal progesterone for preventing preterm birth and adverse perinatal outcomes in singleton gestations with a short cervix: a meta-analysis of individual patient data.
        Am J Obstet Gynecol. 2018; 218: 161-180
        • Hernandez-Andrade E.
        • Romero R.
        • Ahn H.
        • et al.
        Transabdominal evaluation of uterine cervical length during pregnancy fails to identify a substantial number of women with a short cervix.
        J Matern Fetal Neonatal Med. 2012; 25: 1682-1689
        • Stamilio D.
        • Carlson L.M.
        Transabdominal ultrasound is appropriate.
        Am J Obstet Gynecol. 2016; 215: 739-743.e1