Repeat measurements of nuchal translucency at 11–14 weeks of gestation: when do we need them?

Authors

  • L. J. Salomon,

    Corresponding author
    1. Obstetrics and Maternal-Fetal Medicine, GHU Necker-Enfants Malades, Assistance Publique-Hôpitaux de Paris, Université Paris Descartes, Paris, France
    2. Société Française pour l'Amélioration des Pratiques Echographiques (SFAPE), Paris, France
    • Correspondence to: Prof. L. J. Salomon, Obstetrics and Maternal-Fetal Medicine, GHU Necker-Enfants Malades, AP-HP, University of Paris Descartes, Paris, France (e-mail: laurentsalomon@gmail.com)

    Search for more papers by this author
  • R. Porcher,

    1. Départment de Biostatistique et Informatique Médicale, Hôpital Saint-Louis, Assistance Publique-Hôpitaux de Paris, Paris, France
    2. Institut National de la Santé et de la Recherche Médicale, Paris, France
    Search for more papers by this author
  • D. Socolov,

    1. University of Medicine and Pharmacy Gr. T. Popa, Iasi, Romania
    Search for more papers by this author
  • H. Lamrani,

    1. Hôpital Sud-Essone, Dourdan-Etampes, Dourdan, France
    Search for more papers by this author
  • Y. Ville

    1. Obstetrics and Maternal-Fetal Medicine, GHU Necker-Enfants Malades, Assistance Publique-Hôpitaux de Paris, Université Paris Descartes, Paris, France
    2. Société Française pour l'Amélioration des Pratiques Echographiques (SFAPE), Paris, France
    Search for more papers by this author

ABSTRACT

Objective

Previous studies have recommended that several repeat measurements of nuchal translucency (NT) be obtained to optimize sensitivity of the screening process. However, truncation is applied within the risk calculation for small NT measurements; therefore, repeating NT measurements in the lower range may be unnecessary. The aim of this study was to determine the optimal number of NT measurements and whether this is dependent on the initial value obtained.

Methods

We simulated the expected distribution of sets of five repeat NT measurements and their corresponding likelihood ratios (LR) for a variety of crown–rump length (CRL) values at 11–14 weeks' gestation, based on the published mixture model for first-trimester NT measurements and on published data regarding NT measurement variability. The ratio between LR obtained from the highest and the initial measurement in each set were computed, with a ratio of 1 indicating that repeat measurements would have no effect on risk estimation. We calculated NT cut-offs below which a change in estimated LR would occur in fewer than 10% of cases if repeat measurements were obtained, and we tested this approach on a real first-trimester screening dataset.

Results

The simulations performed indicated that repeating NT measurements when the first NT obtained ranged between ≤ 0.9 and ≤ 1.6 mm at CRLs of 45 and 84 mm, respectively, is not useful. When applied to a real dataset, our approach allowed avoiding repetition of measurements in 47/165 cases (28%).

Conclusions

Although there is some variability in NT measurements, repeat assessment is not useful to optimize screening performance for the smallest NT measurements due to lower truncation limits that are applied in risk calculation. Our study provides NT cut-off values, in relation to CRL, below which it appears that there is no need to repeat measurements once a good quality image has been obtained. Copyright © 2013 ISUOG. Published by John Wiley & Sons Ltd.

Ancillary