14781-koppen

233 Treatment adherence 11 rank correlation ( P < .20) in the current study and to adherence according to previous studies. 11–13,26,37,38 The second block contained the variables only correlating to adherence according to previous studies. The linearity assumption was tested using scatterplots with regression lines. To test for homoscedasticity, a scatterplot of the regression standard residual against the regression standardized predicted value was made, and to test for multicollinearity an R-matrix was created. The minimal required sample-size for this regression model was calculated using the following equation: “n=10*k”. Based on previous studies and our own estimates, we expected to incorporate 10 predictors, requiring a minimal sample-size of 100 children. To test the reliability of different questionnaire subscales, Cronbach’s alpha was calculated. Values ≥0.7 and <0.8 indicate acceptable reliability , values ≥.08 and <0.9 indicate good reliability and values ≥0.9 indicate excellent reliability. All statistical analyses were performed with SPSS (IBM Corp. Released 2016. IBM SPSS Statistics for Windows, Version 24.0. Armonk, NY: IBM Corp), statistical significance was defined as a P -value < .05. RESULTS Patient characteristics Out of 150 invited families, 126 returned the questionnaire (84% response rate) and 115 patients were included. The patient inclusion process is shown in Supplemental file 1. Patient characteristics are shown in Table 1. No statistically significant differences in demographic variables were found between adherent and nonadherent children. The median age was 7.8 years (IQR 4.9-11.3 years, range 0.6-16.8 years), the median duration of symptoms was 42 months (IQR 24-72 months, range 2-192 months). Fifty-six children (49%) were ≥8 years of age, 48 of whom (86%) completed the children’s questionnaire. MARS-5 In 3/115 included children (3%), the parental MARS-5 questionnaire contained insufficient data to calculate a MARS-5 score. Overall, the median MARS-5 score was 22 (IQR 20-24, range 14-25) and 43 children (37%) had a score ≥23 and were considered adherent. Sixty- three children (55%) had a score ≥22 and 71 children (62%) had a score ≥21. There were no statistically significant differences in self- and proxy-reported MARS-5 scores when children and their parents completed the questionnaire (Supplemental file 2).