Maternal depression, assessment methods, and physical symptoms affect estimates of depressive symptomatology among children with cancer.

Investigated the incidence of depressive symptoms and their covariates in a sample of 99 children undergoing treatment for cancer and their mothers. Although the prevalence of depressive symptoms falling within the clinical range was low (7 to 8%), classification of these children was highly dependent upon the informant and instrument used. Interrater reliabilities did not differ from chance levels. Separate multiple regression analyses of the mother's and nurse's ratings of the child's level of depression, the child's self-report on the Child Depression Inventory, and the mother's responses to the Child Behavior Checklist depression scales revealed different statistical models for each method of assessment. However, increased severity of the mother's self-report of depressive symptoms on the Beck Depression Inventory, which was predicted by low perceived social support and hospitalization of her child, was associated with higher levels of child depression on all child- and parent-report measures. Parental adjustment, sociodemographic, and medical factors as well as methods of assessment must be addressed by models explaining the etiology of depressive symptoms among pediatric oncology patients.