Symptom assessment in pediatric oncology: how should concordance between children's and parents' reports be evaluated?

BACKGROUND: Clinical evaluations in pediatric oncology are often triadic, involving children or adolescents, parents, and clinicians. However, few studies have evaluated the concordance between children's and parents' reports of symptom occurrence.
OBJECTIVES: The purposes of this study were to evaluate the concordance between children's and parents' symptom reports during the week of chemotherapy administration using 5 statistical approaches and determine which factors are associated with higher levels of dyadic concordance.
METHODS: Independent assessments of symptom occurrence were obtained from children and adolescents with cancer (n = 107) and their parents using the Memorial Symptom Assessment Scale 10-18. Concordance was assessed using (1) percentage of overall agreement, (2) Cohen κ coefficients, (3) McNemar tests, (4) positive percentage agreement, and (5) negative percentage agreement.
RESULTS: For each dyad, an average of 20 of the 31 symptom reports were concordant. Using children's reports as the "gold standard," parents rarely underestimated the children's symptoms. However, compared with children's reports, parents overestimated 7 symptoms. Advantages and disadvantages of each of the statistical approaches used to evaluate concordance are described in this article.
CONCLUSIONS: A variety of statistical approaches are needed to obtain a thorough evaluation of the concordance between symptom reports. Discordance was most common for symptoms that children refuted, particularly psychosocial symptoms. IMPLICATIONS FOR PRACTICE: Clinicians need to interview children and adolescents along with their parents about the occurrence of symptoms and evaluate discrepant reports. Effective approaches are needed to improve communication between children and parents to improve symptom assessment and management.