Introduction:

Medical biometrics has made it possible to identify predictive variables of birth weight.

Objective:

To determine the local discriminatory power and performance of fetal biometric variables at 22 weeks on the trophic condition of the newborn.

Methods:

An observational, analytical and retrospective study was carried out in three health areas of the Santa Clara municipality, in the period between January 2013 and December 2019. From a population of 6,035 births, 2,454 were selected by simple random sampling. Data were obtained from records of genetic consultations. In the analysis, areas under the Receiver Operating Characteristic curve were constructed and performance indicators for diagnostic tests were calculated.

Results:

The areas under the curve of the biometric variables discriminate those born small and large for gestational age. In the small ones they exceed 0.840 except for the length of the femur; in the large ones, the estimated fetal weight reaches a curve of 0.715, the rest are lower. Local cut-off points are estimated. The performance indicators of the biometrics maintain a regular behavior; those that are estimated by transforming the values ​​from the reference tables are more specific with values ​​above 80%; while those calculated after transforming the variables by the estimated cut-off points raise the sensitivity above 60%.

Conclusions:

All biometric variables have discriminatory capacity for deviations of the trophic condition at birth, preferably for small births for gestational age. The optimal cut-off points identified differ from those established in the reference tables. The performance indicators of the fetal biometric variables showed superiority according to the estimated cut-off points with respect to those of the reference tables.

        · abstract in Spanish     · text in Spanish     · Spanish ( pdf )