1 Ardilouze et al present a different way to analyze and interpret our results derived from a recent randomized prospective crossover study investigating the tolerability of 2-day use versus 4-day use of insulin infusion sets during continuous subcutaneous insulin infusion (CSII) treatment.2 The authors are surprised by the absolute number of adverse events (AEs) recorded, and criticize our methodology of using poorly defined AE categories, nonvalidated AE questionnaires and having only a small number of participants. We appreciate the effort but would like to comment as follows:1. The study has been conducted in accordance with all existing quality standards (good clinical practice [GCP], local regulatory and legal requirements, regular monitoring by external clinical research associates (CRAs) including source data validation, etc). So the results are what they are. We apologize that the outcome is not in line with the expectations of health care professionals, who are in great favor of CSII as a treatment option. It is well known that studies specifically targeting adverse events as a primary endpoint result in higher AE frequencies than studies with clinical endpoints, where AEs are collected only in the process of the routine visits. In our opinion, our study does not even discredit CSII as the most optimal treatment option for insulin-treated patients with diabetes, but provides a real-world insight into the number of issues occurring in daily practice. Our way to categories the AEs and our standardized questionnaires are not validated, because there is no precedent research existing. They should preferably be understood a first suggestion for future validations. Unfortunately, we are not aware about any other study conducted yet with similar design, and we hope that our publication may trigger further research on this topic. 2. We respectfully disagree to the notion, that our sample size was too small for the purpose. In the absence of alternative studies, we used a pilot study, which we performed several years ago, where patients were using their catheters for steadily increasing periods of time (ie, 1 day, 2 days, etc) until skin problems started to occur. 3 In that prior study, the problems started to occur on day 3. According to the power analysis based on these pilot findings, the number of patients for this study to see a statistically significant difference at a power level of 80% with a first order error of 5 % was as small as 12.6 for the crossover design. By enrolling 24 subjects, we tried to accommodate the expectation of preferably larger cohorts to come to an at least acceptable number. However, a larger number was clearly not necessary to confirm the study hypothesis, because of the size of the negative effect of prolonged catheter use on treatment tolerability in the majority of the study participants. 3. We agree to the point that frequent catheter replacement every second day is more expensive. We would like to emphasize, however, that next to the higher frequency of AEs, a prolonged use ...