Impact of SGA, AGA, and LGA birthweight subclassification on NICU mortality risk identification: a cohort study

Objective To assess whether subdividing small (SGA, <10th), appropriate (AGA, 10th–90th), and large (LGA, >90th percentile) for gestational age (GA) birthweight classifications improves mortality risk estimation among preterm infants in neonatal intensive care units (NICUs). Study Design Data from 25,779 singleton inborn NICU infants (24–30 weeks GA) from the Pediatrix Clinical Data Warehouse (2013–2018) were analyzed. Infant birthweights were subclassified into 12 subgroups (extremely SGA [eSGA], <3rd; moderately SGA [mSGA], 3rd-<10th, 8 AGA categories [3-AGA to 9-AGA]; mLGA, >90th-97th; eLGA, >97th percentile). Mortality probability by birthweight category was modeled using logistic regression. Result Mortality probability was higher for infants classified as eSGA (28.9%, 95% CI [26.4–31.4%]) than mSGA (9.9% [8.7–11.3%]). Smaller AGA infants had higher mortality than larger AGA infants (e.g., 3-AGA: 8.5% [7.5–9.7%] vs. 9-AGA: 5.0% [4.2–5.9%]). Subdividing LGA provided no additional information. Conclusion Subdivision of SGA and AGA birthweight categories may improve mortality risk prediction among premature infants.