Govert Veldhuijzen

154 Chapter 8 alpha and item-total correlations for verifying the reliability. A Cronbach’s Alpha between 0.7-0.95 was accepted for internal consistency. The item-total correlations were allowed to vary between 0.2-0.8, with <0.2 meaning there is not enough relation and >0.8 meaning there is too much relation, and therefore not providing additional information. A p-value < 0.05 was accepted as statistically significant. SPSS 25.0 and R Studio 3.4.4 were used to perform the statistical analyses. All answers were scored according to the Likert scale. One point for the most positive answer and 5 points for the least positive answer. The same applied for binary (1 or 5) and three-point Likert scales. The D-GESQ total score was calculated by the following formula: (Score – lowest possible score) / score range x 100. This results in a range from 1 (very unsatisfied) to 100 (very satisfied). This study followed the principles of the Declaration of Helsinki 21 and was in accordance with the Dutch Medical Research Involving Human Subjects Act (WMO). Ethics approval was obtained from the METC Arnhem-Nijmegen, number 2016-3020. The trial was registered in the Dutch Trial Register (NTR), number 26875. 22 RESULTS Think-aloud method The data saturation point was reached after completing the questionnaire by 17 patients. Based on data received by performing the think-aloud procedure, two questions were corrected. Question 3 had an inadequate number of answer options; 89% of the patients answered plenty and none of them answered the last option “no”. During the think-aloud procedure patients reported that they missed the option sufficient, so this option was added. Question 17 was rewritten. We replaced “your questions” into “any questions”, because patients suggested that if they had questions, they would be answered by the given explanation, but in most cases, they did not have any questions. Main study From January 2018 until July 2019, we sent the 2.0 version of the D-GESQ to 1065 patients who were informed via the CBE platform. In total, 227 of 1065 patients completed the final questionnaire and were included in the study, resulting in a response rate of 21.3%. Baseline characteristics of these patients are summarized in table 1.