123 Determinants for side effects 4 Statistical analysis Descriptive statistics with either means and standard deviations or medians with interquartile ranges (IQR) were calculated. The difference between SDQ and SDSC scores at T1 versus T2 was tested with a Wilcoxon signed-rank test. The intra-class correlation coefficient (ICC) was calculated to compare SDQ data of parents and children (two-way mixed effects model, single measures, absolute agreement). Univariable logistic regression models were estimated to explore associations between the potential determinants and either neurobehavioral or sleep problems. Odds ratios (OR) with 95% confidence intervals (95%-CI) were estimated. Significant determinants with a p-value <0.20 and four or more patients in each cell of the contingency table were included in the multivariable model. Furthermore, variables that were clinically relevant based on reported literature (patient age and sex15,33,34 and maintenance week35,36), as well as T1 SDQ score (for neurobehavioral problems) and T1 SDSC score (for sleep problems), were included in the multivariable logistic regression model. Multicollinearity was checked as described before.37 All analyses were performed using IBM SPSS Statistics version 26.0.
RkJQdWJsaXNoZXIy MTk4NDMw