Interpretation
Recently, a multicenter Swedish cohort study involving 11,465
asymptomatic singleton pregnant women found that TVU ability to predict
sPTB at <37 weeks was poor: AUC of 0.63 (0.59–0.67) for
measurement at 21–23 (+6) weeks with best cutoff point 35mm; and the
number needed to screen (NNS) to detect one true positive test result
for sPTB at <34 weeks considering CL ≤25mm was 524. TVU
demonstrated good performance (AUC >0.75) for predicting
sPTB at <31 weeks’ gestation 21. Despite the
considerable differences between our population and theirs, including
the fact that our patients used progesterone if CL ≤30mm and the
difference between sPTB rates (7.1% our study versus 3.6% Swedish
study), both studies illustrate that 25 mm does not seem to be the best
cutoff point to identify women at PTB risk; moreover, TVU has moderate
or good accuracy when different gestational ages are considered in both
analyses. In addition, our NNS to identify one true positive
sPTB<34 weeks when patients with CL ≤25mm is considerably
lower than previous studies that considered populations with lower PTB
rate 21,22, what is an alert to correctly define the
applicability and cost-utility of TVU-CL measurement as a screening test
for PTB in different countries.