Item-Level Psychometrics of the Neonatal Eating Outcome Assessment in Orally Feeding Infants

Brooke Mulrenin; Roberta Pineda; Cynthia Dodds; Craig A Velozo

doi:10.1177/15394492231212399

Item-Level Psychometrics of the Neonatal Eating Outcome Assessment in Orally Feeding Infants

OTJR (Thorofare N J). 2023 Nov 19:15394492231212399. doi: 10.1177/15394492231212399. Online ahead of print.

Authors

Brooke Mulrenin¹, Roberta Pineda^{2

3

4}, Cynthia Dodds¹, Craig A Velozo¹

Affiliations

¹ Medical University of South Carolina (MUSC), Charleston, SC, USA.
² University of Southern California (USC), Los Angeles, CA, USA.
³ Keck School of Medicine of USC, Los Angeles, CA, USA.
⁴ Washington University School of Medicine, St. Louis, MO, USA.

PMID: 37981785
DOI: 10.1177/15394492231212399

Abstract

Background: The Neonatal Eating Outcome Assessment determines feeding performance based on the infant's postmenstrual age (PMA).

Objective: To examine item-level measurement properties of this assessment's rating scale.

Methodology: In this retrospective study, Rasch analysis was completed on clinical data from the Neonatal Eating Outcome Assessment for 100 infants (52 preterm and 48 full-term) using Winsteps version 3.93.1. Instead of PMA-based scores, ordered letters converted to numerical scores were analyzed.

Results: Analysis demonstrated that Section I (Pre-Feeding Skills) represents a separate construct from Sections II and III (Oral Feeding and End of Feeding, respectively). Sections II and III were adequately unidimensional to complete Rasch analysis. These sections fit the Rasch model overall, but rating scale category underuse was common, which may be attributed to sample characteristics.

Implications: This analysis supports using validated ordered letter scoring of Sections II and III to measure oral feeding performance in preterm and full-term newborns.

Keywords: Rasch analysis; item response theory; occupational therapy; pediatrics; psychometrics.

Plain language summary

Examining Measurement Properties of Items in the Neonatal Eating Outcome AssessmentThe Neonatal Eating Outcome Assessment is used to evaluate infant feeding skills, considering expected performance based on the infant’s age. We completed Rasch analysis of this assessment using data from 100 infants (52 preterm and 48 full-term). Analysis used ordered letter scores (not based on infant age) which allowed comparison of the skill level of each infant and the difficulty level of each assessment item. This also provided information about how well each item measures the feeding behavior it intends to assess. Study limitations included underutilization of some rating scale categories and inability to analyze Section I. However, overall, results showed that items in Sections II and III of The Neonatal Eating Outcome Assessment provide good measurement of infants’ oral feeding performance. This information will help clinicians and researchers understand this assessment’s strengths and opportunities for improvement.