Primary outcome
The primary outcome was the proportion of patients reaching the therapeutic objectives decided by the physician at inclusion: HbA1c for type 2 diabetes, LDL-c blood levels for hypercholesterolemia and systolic BP for hypertension. The objectives were set individually for each patient by the physician himself in accordance to French or international recommendations for hypertension[49], type 2 diabetes[51] and hypercholesterolemia[52]. If the patient had more than one pathology, the primary outcome was based upon the pathology specified as the most relevant in terms of cardiovascular prevention by the physician before inclusion.
We proceeded to a per protocol analysis on the primary outcomes, in which we excluded all the drop out patients from the analysis. This study was not a drug trial, and we then presumed that loss of follow-up was not due to the study per se nor to the intervention but to external reasons. Then, the drop out patients should not necessarily be considered as they did not reach the therapeutic objectives. A per protocol analysis seemed acceptable to conclude on our results.