Statistical Analyses
Descriptive statistics were used to characterize the sample at baseline and follow-up. Patients were divided by age at baseline into younger (age < 6 years) and older (age ≥ 6 years) age groups, with 6 years set as the cutoff to match the age-appropriateness criteria for the measures used (e.g., BASC and BRIEF Preschool versions <6 years), to assess the vulnerability of this young age group to the detrimental effects of radiation, and to be consistent with previously published results from our group.13 The two age groups were divided by radiation field (CSI or focal). For the present study, the BASC Behavioral Symptoms Index and Adaptive Skills and the BRIEF Global Executive Composite, Behavioral Regulation, and Metacognition indices were used as primary outcome measures. One-way analysis of variance (ANOVA) was conducted to determine the differential impact of demographic, diagnostic, and treatment-related characteristics on these primary outcome measures. Pearson correlation was used to investigate the relationship between the primary outcome measures and the time interval between baseline and follow-up. A one-way ANOVA was conducted to compare mean differences in the primary outcome measures at follow-up across the four age-by-radiation-field groups. When the one-way ANOVA resulted in a significant group difference, a least significant difference post hoc analysis was performed.
One-sample t-tests were used to compare mean scores with published normative means. Rates of impairment were calculated and defined as the frequency of scores > 1.5 S.D. above the mean of 50 (T-score > 65). Chi-square analysis was conducted to evaluate the differences in rates of impairment by age groups. Pearson correlation was used to examine the relationships between those scales where rates of impairment exceeded expectations in the general population.
Analyses were performed using SPSS version 24 (IBM 2016, Chicago, IL). Two-tailed analyses were used in all comparisons; statistical significance was defined as p < 0.05.