Questionnaire-based reports of Quality of Survival and direct assessments of cognitive performance in children treated for medulloblastoma in the PNET 4 randomized controlled trial

摘要

Background: The relationship between direct assessments of cognitive performance and questionnaires assessing quality of survival (QoS) is reported to be weak-to-nonexistent. Conversely, the associations between questionnaires evaluating distinct domains of QoS tend to be strong. This pattern remains understudied.Methods: In the HIT-SIOP PNET4 randomized controlled trial, cognitive assessments, including Full Scale, Verbal and Performance IQ, Working Memory, and Processing Speed, were undertaken in 137 survivors of standard-risk medulloblastoma from 4 European countries. QoS questionnaires, including self-reports and/or parent reports of the Behavior Rating Inventory of Executive Function (BRIEF), the Health Utilities Index, the Strengths and Difficulties Questionnaire, and the Pediatric Quality of Life Inventory, were completed for 151 survivors. Correlations among direct cognitive assessments, QoS questionnaires, and clinical data were examined in participants with both assessments available (n = 86).Results: Correlations between direct measures of cognitive performance and QoS questionnaires were weak, except for moderate correlations between the BRIEF Metacognition Index (parent report) and working memory (r = .32) and between health status (self-report) and cognitive outcomes (r = .35–.44). Correlations among QoS questionnaires were moderate to strong both for parent and self-report (r = .39–.76). Principal Component Analysis demonstrated that questionnaires and cognitive assessments loaded on 2 separate factors.Conclusions: We hypothesize that the strong correlations among QoS questionnaires is partially attributable to the positive/negative polarity of all questions on the questionnaires, coupled with the relative absence of disease-specific questions. These factors may be influenced by respondents’ personality and emotional characteristics, unlike direct assessments of cognitive functioning, and should be taken into account in clinical trials.