PURPOSE: Early feeding skills need to be examined across the entire feeding spectrum and in response to changing conditions to determine appropriate interventions. The purpose of this research was to introduce a method of continuous examination of early infant feeding skills, the Dynamic-Early Feeding Skills (D-EFS) coding system, and to examine its reliability and validity.
PARTICIPANTS: Twenty preterm infants at risk for lung disease and younger than 32 weeks gestational age participated in this study. A total of 75 videotaped feeding observations were conducted when infants were orally feeding half of their prescribed volume for a 24-hour period.
DESIGN: Descriptive, correlational design.
METHODS: A small microphone was placed lateral to the infant's trachea and connected to the video camera to amplify the sounds of sucking, swallowing, and breathing on the videotape. Videotapes were a close-up of the infant's face and upper body. Observations were subdivided into the bottle-in feeding periods for coding purposes. By using the D-EFS coding system, infant feeding behaviors that were audible and/or observable, along with caregiver feeding behaviors, were coded continuously across all feeding periods.
MAIN OUTCOME MEASURES: Code dimensions included infant state, engagement in feeding, behavioral organization, swallowing function, respiratory regulation, infant position, and caregiver feeding actions (eg, stimulating sucking, stopping milk flow). In addition, breathing waveforms, oxygen saturation (Sao2), and heart rate data were collected throughout the feeding.
PRINCIPAL RESULTS: Interobserver reliability of the coding system was demonstrated with [kappa] coefficients of 0.54 to 0.93 for each code (median 0.79) and 0.84 overall. As a group, infants were drowsy (62%) and demonstrated low engagement in feeding (58%). Behavioral disorganization, swallowing, and respiratory dysregulation occurred in 23%, 37%, and 29% of the feeding, respectively. The 4 markers of distress (low energy, behavioral disorganization, respiratory and swallowing dysregulation) were associated with child characteristics (increased days on ventilator and continuous positive airway pressure, increased time to full oral feeding, and oxygen requirement). Respiratory dysregulation was associated with lower minimum Sao2, higher percentage time below 85% Sao2, less breaths, and longer intervals between breaths during the feeding. Feeder behaviors associated with less distress included providing more rest periods, positioning side-lying, decreased number of stimulate suck, and increased number of stopping milk flow events.
CONCLUSIONS: The D-EFS coding system is a valid and reliable method of assessment of early feeding skills and caregiver feeding behaviors. It can be built upon in future studies to assess the effect of specific interventions on early infant feeding.