RESUMO
AIM: This paper is a report of a meta-analysis to investigate the association between self-report pain ratings for the dyads of child and parent, child and nurse and parent and nurse. BACKGROUND: Existing research has shown conflicting results with regard to agreements of self-report pain ratings between the three dyads. DATA SOURCES: The CINAHL, Medline, Ovid and PsycINFO databases were searched using keyword, such as 'children/adolescents', 'parents/nurses', 'pain assessment', 'pain ratings', 'association' and 'agreement'. Studies published in English in or after 1990 were included. METHODS: Meta-analysis methodology was applied to 12 pain assessment studies published between 1990 and 2007 which met the inclusion criteria. In the 12 studies a common effect size was estimated using the Pearson's correlation coefficient. Therefore, a fixed-effects model was chosen for this meta-analysis. RESULTS: We found moderate summary effect sizes between self-reported pain ratings for the dyad of child and parent (r = 0.64) and the child and nurse dyad (r = 0.58) and a weak summary effect size of r = 0.49 for the dyad of parent and nurse. The summarized effect sizes for each of the three dyads varied across the studies. A test of homogeneity (Q-statistic) indicated that all effect size estimates were not homogeneous. CONCLUSION: Parents' and nurses' perceptions of children's pain should only be considered as estimates rather than expressions of the pain experienced, and not the same as children's self-reports. There is a need for education on selection of appropriate pain assessment scales in relation to the age and development of the child.
