Determinants of health-related quality of life proxy rating disagreement between caregivers of children with cancer

Purpose: Proxy reports of health-related quality of life (HRQoL) are commonly used in pediatric oncology. However, it is not known if caregivers' reports differ. This study therefore aims to compare paternal and maternal proxy reports, and explore determinants of couple disagreement (sociodemographic and medical characteristics, and parental QoL and distress).

Methods: Both parents completed the PedsQL generic (child's HRQoL), Short Form-12 (own QoL) and Distress Thermometer for Parents. To assess agreement in child HRQoL, intra-class correlation coefficients (ICCs) were calculated. Differences between fathers/mothers were assessed with paired t tests. Systematic disagreement patterns were visualized with Bland-Altman plots. Characteristics of parental couples with a mean proxy difference in the highest quartile (highest proxy score minus lowest proxy score) were explored with multiple logistic regression analysis.

Results: Parents of 120 children with cancer (87% post-treatment, mean age 11.0 ± 5.7 years) participated. No significant differences were found between paternal and maternal proxy scores, and agreement was good on all scales (ICCs 0.65-0.83). Bland-Altman plots revealed no systematic disagreement patterns, but there was a wide range in magnitude of the differences, and differences went in both directions. Couples with a mean proxy difference (irrespective of which direction) in the highest quartile (± 20 points) were more likely to have a child in active treatment, with retinoblastoma or relapsed disease, and to diverge in their own QoL.

Conclusions: If proxy reports of only one parent are available, clinicians may reasonably assume that paternal and maternal reports are interchangeable. However, if in doubt, respondent's sex is not of major importance, but clinicians should be aware of patient's and family's characteristics.