Pain-specific Reactions or Indicators of a General Stress Response?: Investigating the Discriminant Validity of 5 Well-established Neonatal Pain Assessment Tools

Abstract

Objectives: There are many neonatal pain assessment tools available. However, systematic psychometric comparisons between tools are lacking, particularly those comparing tools regarding their ability to differentiate between pain and stressful procedures. The aim of the present study was to compare 5 widely used neonatal pain assessment tools: Neonatal Facial Coding System-Revised, Premature Infant Pain Profile-Revised, Neonatal Pain, Agitation and Sedation Scale, Neonatal Infant Pain Scale, and Bernese Pain Scale Neonates.

Materials and methods: Two coders applied all pain assessment tools to videos of 42 neonates who were videotaped during a pain and a stressful procedure.

Results: Interrater reliability and relative convergent validity were high and internal consistency good to excellent for all 5 assessment tools. All tools discriminated between painful and stressful events. Tools differed regarding their overall effect sizes as well as their items' effect sizes. Behavioral items tended to have larger effect sizes than physiological ones. Yet, effect sizes of items from one behavioral category differed greatly, which may be due to different operationalization of coding schemes.

Discussion: Given that the tools investigated in the present study appear to be fairly comparable psychometrically. Aspects of their clinical utility are discussed and ways of improvement identified.