Comparing transabdominal and transvaginal cervical length measurements at mid-trimester fetal anomaly scan: The impact of bladder fullness and lower uterine contractions

Abstract

Purpose

To assess the effects of bladder fullness and lower uterine contractions ultrasound on transabdominal and transvaginal cervical length measurements at the mid-trimester fetal anomaly scan (FAS).

Methods

Transabdominal and transvaginal cervical length measurements from 925 mid-trimester FAS examinations were retrospectively analysed. Images were assessed for lower uterine contraction and bladder fullness using a novel qualitative assessment. Bland–Altman plots and single-score interclass correlation (ICC) were used to determine correlation between transabdominal and transvaginal measurements. Sensitivity and specificity of transabdominal cut-offs were calculated.

Results

Transabdominal and transvaginal measurements of the cervix correlated poorly (ICC 0.306). An overfilled bladder and lower uterine contractions on average increased the length of transabdominal cervical length measurements. Removing these variables did not significantly improve correlation between transabdominal and transvaginal measurements of the cervix but resulted in an improved sensitivity of transabdominal assessment to detect a clinically relevant short cervix.

Discussion

Resolving the confounding factors of an overfilled bladder and lower uterine contractions can help improve the our ability to detect a short cervix on transabdominal ultrasound. Our data set supported a two-stage approach to cervical length screening which would allow 100% sensitivity when a cut-off of ≤35 mm is used on transabdominal ultrasound and would limit the need for transvaginal scanning to approximately 39% of patients. This cut-off is in line with the findings of other studies. The low prevalence of short cervix in our study did however make it difficult to extrapolate reliable calculations.

Conclusion

Although transabdominal measurements correlate poorly with transvaginal measurements of the cervix, we demonstrated an improved sensitivity for detecting a short cervix using a transabdominal approach when no contractions or overfilled bladder is present. This potential could be explored in a future study with a larger sample size.