2
Highlights NFCS and MBPS coding systems were examined using Confirmatory Factor Analysis. A revised 3-item NFCS maintained good psychometric properties of 7-item version. Redefinition of MBPS with cry as a sole indicator was suggested. The revised scales increase efficiency of coding based on improved psychometrics.
AbstractDiverse behavioural cues have been proposed to be useful cues in infant pain assessment, but there is a paucity of evidence based on formal psychometric evaluation to establish their validity for this purpose. We aimed to examine two widely-used coding systems, the Neonatal Facial Coding System (NFCS) and the Modified Behaviour Pain Scale (MBPS), by examining their factor structures with confirmatory factor analysis using a large archival dataset. The results indicated that an item-reduced NFCS scale with three items produced a one-factor pain model that maintained the good psychometric properties of the 7-item scale. In addition, it was found that MBPS also has challenging internal consistency, with items that are both weakly correlated and highly redundant. One-item of the MBPS may be able to capture the construct of pain equally well or potentially improve its psychometric properties. Redefinition of MBPS with cry as a sole indicator was suggested. This analysis provides two new iterations of the NFCS and MBPS that improve construct validity and internal consistency. These shorter versions also improve the feasibility of both measures and increase their potential for clinical use as less time is required for their administration.Perspective: This article presents new iterations of the NFCS and MBPS scales. These revised measures improve the internal consistency of the measures, feasibility of use of the tools in research settings and the efficiency of the coding process. The revised tools could also improve the feasibility of coding within clinical settings.