Interrater Reliability and Concurrent Validity of the Neonatal Eating Outcome Assessment.

Abstract

Few neonatal feeding assessments are currently available, and the Neonatal Eating Outcome Assessment is the only one that identifies feeding impairment while considering the developmental changes that occur from preterm birth to term-equivalent age. To determine the interrater reliability and concurrent validity of the Neonatal Eating Outcome Assessment. Prospective, observational study. Level 4 neonatal intensive care unit. A convenience sample of 7 neonatal therapists participated in reliability testing. For concurrent validity, a prospective cohort of 52 preterm infants born ≤32 wk gestation had feeding assessed at term-equivalent age. Intraclass correlations (ICCs) and Fleiss's κ statistics were used to define reliability across therapists, who independently scored five videos of preterm infants orally feeding using the Neonatal Eating Outcome Assessment. Concurrent validity was determined by evaluating relationships between the Neonatal Oral Motor Assessment Scale (NOMAS) and the Neonatal Eating Outcome Assessment using an independent-samples t test and χ² analysis. The ICC for the Neonatal Eating Outcome Assessment total score was 0.90 (confidence interval [CI] [0.70, 0.99]). Fleiss's κ scores for the 19 scorable items on the Neonatal Eating Outcome Assessment had predominately moderate, fair, and slight agreement, with 3 items having poor agreement. Dysfunctional NOMAS scores were related to lower Neonatal Eating Outcome Assessment scores (t[49.4] = 3.72, mean difference = 12.2, 95% CI [5.60, 18.75], p = .001). The Neonatal Eating Outcome Assessment has excellent reliability. Concurrent validity was established. This article reports that the final version of the Neonatal Eating Outcome Assessment (Version 5.7) has excellent interrater and concurrent validity and is an important tool to assess the occupation of infant feeding in clinical practice.