Measurements of heart and vessel size need to be adjusted on the children’s body size. In children with a weight problem, there may be an error in this adjustment, leading to under- or over-diagnosis. This project aims to improve adjustment for obese children to obtain more accurate diagnosis.
The objectives are to evaluate if excess weight causes underestimation of echocardiography Z scores in a large sample of pediatric echocardiography measurements and to develop and validate normalization methods that minimize the bias introduced by excess weight in pediatric quantitative echocardiography.
This project involved gathering cardiovascular 2D measurements from more than 20,000 normal echocardiograms (screen-negative) from several centers in Canada.
In a first study, we found that that BSA-based models did not completely adjust for body size. This resulted in an underestimation of z scores in subjects with higher body mass index (BMI) and an overestimation of z scores in subjects with lower BMI. We also showed that prediction models based on height and weight yielded more accurate AoV diameter predictions, including in subjects with extreme BMI. (Plante V et al. “Alternative to Body Surface Area as a Solution to Correct Systematic Bias in Pediatric Echocardiography z Scores.” Can J Cardiol 37(11): 1790-1797.)
In a second study, we confirmed that that Z-scores calculated with only BSA showed errors on the order of one full Z-score in some groups, notably newborns, lean school-aged children and overweight adolescents. Using a Generalized Additive Model for Location, Scale and Shape (GAMLSS) enabled the proposition of a new set of Z-scores that corrected these errors by considering height, weight, body mass index and age. The proposed GAMLSS Z scores offers a more accurate description of the expected cardiac diameters according to body size, which therefore enable a more accurate assessment of the impact of diseases on cardiac dimensions. (Lauzon-Schnittka J, Circ Cardiovasc Imaging: e017944. DOI: 10.1161/CIRCIMAGING.124.017944, PMID: 40631677)


Leave a Reply