Ietje Perfors

115 arm were evaluated for contamination. All GP consultations within two weeks were registered in both groups. Data collection Patient characteristics were collected online directly after inclusion (baseline). Data extraction at baseline, including the number of GP contacts (year prior to inclusion), was performed in the free text and coded routine care data from the EMR of each GP practice. GP characteristics at T0 and rurality were collected from public Dutch online databases for GP experience. 25, 26 Comorbidities, date of diagnosis, cancer stage and treatment decision were extracted from the EMR in hospitals. The moment of treatment decision was defined as the moment the patient agreed with or chose the treatment. Sample size The sample size was based on the primary outcomes of the GRIP study, i.e., satisfactionwith care at threemonths after the end of therapy (excl. hormone therapy), with a maximum of one year. We assumed a medium effect size (0.5) to be a relevant difference between the two study groups. Using a power of 0.8 and an alpha less than 0.05, at least 64 patients per study group were required. Accounting for an estimated dropout of 15%, 75 participants in each group were needed. 19 Statistical analysis The study population was described descriptively. Intervention effects compared to usual care were analysed following the intention-to-treat principle. Additionally, outcomes were described stratified for patients with a TOC before treatment decision (conform protocol), a TOC after treatment decision, and no TOC. Paired sample T-test was used to calculatemean changes and 95%confidence intervals of self-efficacy from baseline to T1 within groups. ANOVA was used to calculate between-group differences (i.e., intervention versus control 5