Abstract
Objective: To assess the predictive value of selected growth phenotypes for neonatal morbidity and mortality of very low birth weight (VLBW) infants and compare with INTERGROWTH-21st (IG21).
Methods: Retrospective analysis of data from the Brazilian Neonatal Research Network (BNRN) database on VLBW infants (≥ 22 to < 30 weeks gestational age) admitted to neonatal intensive care units at 20 public, tertiary-care, university hospitals. The database was split into training (70% cases) and validation (30% cases) datasets. Weight to length ratio (W/L ratio) was calculated as kg/m, small for gestational age (SGA) was defined as being below the 3rd or 10th centiles of the weight for GA and sex. Large for gestational age (LGA) was defined as being above the 97th centile of the weight for GA and sex. Stunting was defined as being below the 3rd centile of the length for GA and sex and Wasting was defined as being below the 3rd centile for the BMI for GA and sex. The composite neonatal morbidity and mortality (CNMM) consisted of in-hospital death, oxygen use at 36 weeks corrected postnatal age, intraventricular hemorrhage grade 3 or 4, or necrotizing enterocolitis Bell’s grade 2 or 3. Single and multiple log-binomial regression models were fitted to estimate the adjusted relative risks (aRR) with 95% confidence intervals of the CNMM for growth phenotypes, comparing to the IG21.
Results: A total of 4,072 infants were included (2,900 in the training set, and 1,172 in the validation set). The occurrence of the CNMM was 58.6% for the whole cohort. The BNRN reference yielded higher aRR than IG21 for stunting and W/L ratio. BNRN showed a lower aRR for SGA compared with IG21. No differences were observed for LGA, and the agreement between the BNRN and IG21 was variable. The growth phenotypes had an excellent specificity (> 95%) and good positive predictive value (70- 90%), with a weak sensitivity and negative predictive values for the CNMM.
Conclusion: The BNRN phenotypes at birth differed markedly from the IG21 standards, and showed a weak accuracy in predicting adverse neonatal outcomes. More studies of growth phenotypes in VLBW infants associated with the CNMM are still needed before introduction and implementation.
Los contenidos publicados en esta revista están protegidos bajo una Licencia Creative Commons Atribución 4.0 Internacional (CC BY 4.0). Esto significa que cualquier persona es libre de compartir, usar y construir a partir de este artículo, incluso con fines comerciales, siempre que se otorgue el crédito apropiado al autor original, se proporcione un enlace a la licencia, se indique el nombre y edición de la Revista.
Esta licencia no impone restricciones adicionales, lo que garantiza la libre circulación y reutilización del conocimiento con respeto y transparencia hacia los derechos de los autores. (Véase El efecto del acceso abierto).