Agreement and disagreement between child-caregiver symptom and function reports in pediatric oncology.

Abstract

The use of patient-reported outcomes (PRO) is increasing in pediatric clinical practice and research, including in clinical trials. Treatment teams are now able to discern differences in symptom and function reports between children with cancer and their family caregivers. Few studies, however, have quantified such differences. Children and caregivers were matched to create a dyad sample, and child/caregiver agreement of responses to the Patient-Reported Outcome Measurement Information System (PROMIS) Pediatric symptom and function measures were examined using the established minimally important difference (MID) value (three points) derived from these measures. Latent class analysis (LCA) was then used to analyze the status/patterns of response agreement. A total of 430 dyadic cases (dyads) were used for LCA modeling. Three classes of child-caregiver concordance resulted: (a) Agreement-caregiver and child within one MID (200 dyads); (b) Disagreement-caregiver underestimating symptoms (47 dyads with caregiver scores more than 1 MID lower than the child scores); and (c) Disagreement-caregiver overestimating symptoms (183 dyads with caregiver scores more than 1 MID higher than child scores). As PRO use increases in pediatric oncology, clinician identification of child/caregiver differences on symptom and function reports will help to distinguish family perceptions of cancer treatment impact. The treatment goal, in addition to mitigating symptom and function treatment-related burdens for the child and caregiver, may now also be to discern the direction and clinical meaning of the difference within the child-parent dyad.