(P10-033-24) Sensitivity and Specificity of Revised Malnutrition Indicators in the Neonatal Intensive Care Unit

Objectives: To investigate the sensitivity and specificity of neonatal malnutrition indicators compared to the historical standard, and to identify characteristics associated with false positive and negatives, among infants discharged from the neonatal intensive care unit (NICU) from 2016-2021.

Methods: Data from 1483 infants discharged from the NICU between 2016-2021 were used in this analysis. Infants with major congenital anomalies were excluded. Infants were classified as meeting the reference standard for growth failure if their weight-for-age z-score (WAZ) at discharge was less than -1.28 (less than the 10th percentile). Infants were classified as meeting revised neonatal malnutrition criteria, published by the Academy of Nutrition and Dietetics and American Society for Parenteral and Enteral Nutrition, of having at least mild, moderate, and severe neonatal malnutrition if their birth to discharge WAZ declined by ≥0.8 SD; >1.2 SD; and >2 SD, respectively. Sensitivity and specificity were calculated. Logistic regression was used to identify characteristics associated with having either false positive or negative result.

Results: Compared to the reference, the sensitivity of mild, moderate, and severe neonatal malnutrition was 60.0%, 24.1%, and 3.5%, respectively. The specificity  of mild, moderate, and severe neonatal malnutrition was 46.7%, 81%, and 96.5%. Gestational age was associated with a reduced odds of a false positive; WAZ was associated with increased odds of false positive result.

Conclusions: The specificity of severe neonatal malnutrition is high (96.5%), but no threshold for neonatal malnutrition resulted in a high sensitivity when compared to the reference. Infants born at earlier gestational age and higher WAZ have increased odds of testing positive using the mild neonatal malnutrition criteria while also testing negative using the reference. For these analyses, the historical standard of WAZ < -1.28 was chosen as the reference because its use in clinical settings continues, despite the publication of revised indicators. A limitation of this analysis is that it does not identify which indicator is more closely associated with health outcomes or clinical judgment; future research should address this knowledge gap. Clinical judgment should be used when evaluating infants’ nutritional status.

Funding Sources: No funding to disclose.