Analysis Plan
Analyses were conducted using SPSS (v.26). Descriptive statistics were conducted to examine the frequencies, mean and standard deviation of sociodemographic factors, cancer-specific stress, general stress, parent-adolescent communication, and HRQOL. Similar to previous work, the impact of sociodemographic factors was examined by creating a cumulative sociodemographic score [2]. Each sociodemographic variable of interest was dichotomized so that participants scored a 0 or 1 to indicate lesser or greater risk [2]. The scoring is as follows: multi-adult (0) vs. single-adult (1) household status, income > $50,000 (0) vs. income ≤ $50,000 (1), mother education > 12th grade (0) vs. mother education ≤ 12th grade (1), White race (0) vs. non-White race (1), and 0-2 children living in the household (0) vs. 3 or more children living in the household (1). Bivariate correlations were conducted to examine associations between the sociodemographic score, mother’s general stress, adolescent’s cancer-related stress, mother-adolescent communication, and survivor HRQOL (mother proxy-report and child self-report) at 5 years’ post-diagnosis (Table 1).
Two hierarchical regressions were used to examine predictors of: (a) mother’s proxy-report of survivor QOL and (b) survivor’s self-report of HRQOL. In Step 1, only mother’s age and the sociodemographic score were included. Step 2 included adolescent’s cancer specific stress and mother’s general stress. Lastly, Step 3 added mother-adolescent communication.