A large proportion of children needing palliative care have severe physical and learning disabilities limiting their ability to communicate pain.
Aim: To develop a behaviour rating scale suitable for assessing pain in this population.
Methods: Parent surveys (n=121) and interviews with parents (n=20) and clinicians (n=26) were used to identify potential cues for inclusion in the scale, resulting in an initial item set of 56. A second survey of parents (n=46) examined concurrent validity, internal consistency and reliability of those items. Elimination of some and combination of others reduced the scale to 20 items. Children (n=29) were filmed and videos scored by three raters using the 20-item scale to assess concurrent and extreme group validity, internal consistency and intra- and inter-rater reliability. During filming, saliva samples were collected from children and saliva cortisol concentration estimated as a potential criterion for pain and correlate with the PPP score. A slightly adjusted 20-item scale was then used by parents to assess children (n=130) "at their best" and when in pain and the responsiveness of the scale to the administration of analgesics to children (n=39) was assessed.
Results: The 20-item PPP was found to be internally consistent (0.75 to 0.86 Cronbach's alpha) and to have good inter-rater reliability (0.73 to 0.87 Intraclass correlation). Validity of the scale was supported by increase in the PPP score in line with global evaluations of the child's pain by observers and significant differences were present between score of children "at their best" and when in pain. Significant decrease in mean PPP score followed administration of analgesics.
Conclusion: The Paediatric Pain Profile appears a valid and reliable tool for assessing pain in children with severe to profound disability. The tool has potential for both clinical and research use.