Patterns of family relationships in pediatric oncology: implications for children's adjustment upon treatment completion.

OBJECTIVE Family characteristics influence children's adaptation following a cancer diagnosis. Little is known about the importance of family relationships on children's social and emotional health immediately after treatment completion. In this study, we identify subtypes of family relationships and evaluate how these profiles relate to child adjustment upon treatment completion. METHODS Youth (N = 77; ages 7-14) who completed cancer treatment within the past 7 months provided their perceptions of their relationships with their caregivers and siblings. Caregivers and teachers rated participants' internalizing and externalizing symptoms, while participants and caregivers reported on child social and overall functioning. Latent profile analyses identified patterns of family relationships, and we examined associations between these patterns and youth outcomes. RESULTS A four-profile solution best characterized families as (1) low closeness/high sibling-only discord (12.4%); (2) low closeness/high discord (6.6%); (3) high closeness/low discord (47.6%); and (4) moderate closeness/moderate discord (33.4%). Family relationship profiles were differentially associated with self, caregiver, and teacher ratings of adaptation. Participants from families characterized by low closeness/high discord reported more difficulties with peer interactions than participants with high closeness/low discord in their families. Caregivers and teachers reported significantly higher externalizing symptoms for participants from low closeness/high discord and low closeness/high sibling discord families in comparison to participants with high closeness/low discord in their family. CONCLUSIONS Findings underscore the importance of assessing children's perceptions of family relationships in pediatric oncology and providing support to the family during and after treatment. Future directions include assessing how these profiles predict outcomes over time.