Date of Award


Document Type

Open Access Thesis

Degree Name

Medical Doctor (MD)

First Advisor

Nina Kadan-Lottick


The purpose of the current study is to understand emotional and neurocognitive functioning in childhood cancer survivors. In this single-evaluation cross sectional study, 41 childhood cancer survivors (53.2% of those eligible) in the Health Education Research Outcomes in Survivors (HEROS) Clinic at Yale-New Haven Hospital between the ages of six and eighteen and their parents and teachers completed the Behavior Assessment System for Children, Second Edition (BASC-II), Behavior Rating Inventory of Executive Functioning (BRIEF) and Pediatric Quality of Life Inventory (PedsQL). Patients, parents and teachers scores were compared to those of normative populations for emotional, neurocognitive and quality of life measures. Frequencies of impairment and within population differences in the study sample were calculated. Responses by patients were compared to those of parents and teachers to determine inter-rater reliability. The patterns of co-existing neurocognitive and emotional difficulties in the sample were described. Lastly, elements of emotional functioning, neurocognitive impairment, and patient characteristics predictive of impaired quality of life were identified with unadjusted and multivariate analyses. Overall, 56.1% patients were female, the mean age at diagnosis was 3.6 years, and the mean age at study completion was 12.8 years. In examining inter-rater reliability, for most areas of emotional functioning there was poor agreement (kappa < 0.40). Parents and teachers showed moderate agreement in reporting problems with attention (kappa = 0.57), memory (kappa = .61), and metacognition (kappa = 0.52). Parents and children showed greater inter-rater reliability when reporting quality of life than symptoms, with agreement in every realm of the PedsQL (kappa > 0.40). Co-morbidities between emotional and neurocognitive impairments for the most part did not occur together. However, impairments in somatization and withdrawal tended to co-exist with impairments in memory, shift and metacognition (p < 0.05). In the multivariate analyses looking at social, emotional and school functioning, only neurocognitive functioning was a consistent predictor of poor quality of life (OR = 12.94, p = 0.008; OR = 11.48, p = 0.044; OR = 33.5, p = 0.003, respectively). Poor emotional functioning was also predicted by female gender (OR = 15.58, p = 0.025). Similar to previous studies, a significant proportion of childhood cancer survivors in our sample endorse difficulties with internalizing symptoms and executive functions, as well as lower physical, emotional and social functioning, than normative populations. The current study shows inter-observer variability, especially among indexes of emotional symptoms, indicating a need for multiple reporters to determine areas of deficit and true levels of functioning. Problems with memory, shift, initiation and coordination of problem-solving behaviors co-exist with symptoms of somatization and social withdrawal in our study. However, it is the neurocognitive rather than emotional symptoms that are shown to be key predictors of how child-age survivors perceive their quality of life after therapy. These results highlight neurocognitive impairments as a target for intervention during and after treatment for pediatric cancers.

Open Access

This Article is Open Access