Govert Veldhuijzen

121 CBE before colonoscopy is reducing operational costs, with lower patient and societal expense Obviously, the costs of these elements of the model varied between the trial sites in the original multicenter RCT. We therefore made robust assumptions for these different costs (regarding the pay rate of auxiliary staff) based on the experiences of the research team of the original trial. Finally, we used data with regard to the proportions of patients in each trial site from the earlier performed multicenter randomized controlled trial to enrich this model. Secondary outcomes Next to the costs made by the endoscopy units, we were interested in effects for patients and the society. Patients For patient costs, we calculated the cost of nurse counselling visit. We hypothesized that the main cost drivers in this category would be travel costs. To calculate this, we used iMTA costing tool to establish the cost per kilometre and average hospital parking fees. 23 Travel distances were not recorded on patient level as it was outside the remit of the original trial. Therefore, we used travel data provided by the Dutch national colorectal cancer screening program monitor. Their 2017 report included average travel distance from patients homes to their endoscopy unit (calculation based on ZIP code). 24 The total amount that all patients in the original trial have spent will be calculated from these cost components. In the nurse counselling group, this includes all patients (100%), in the CBE group we included patients that required additional nurse counselling visit (3%) after CBE. Society To assess the burden of patient education prior to colonoscopy on society, we assumed that the main societal costs were due to lost productivity of our patients. The average age of patients undergoing colonoscopies in the earlier trials was between 56 and 59 years. 20,25 Therefore still most patients were considered to take active part in the workforce. To establish the costs of productivity loss we used the friction cost method derived from the completed iMTA Productivity questionnaire. 26 To calculate the costs the following inputs were used: work status (employed, unemployed), the number of hours absent from work, absent from unpaid work or hours someone else had to absent from work (for instance to babysit) to allow the patient to get educated prior to endoscopy. For these three categories, the average hourly wages for male, females and unpaid work were provided by the iMTA costing tool. 23 6