[An example of clinical controversy related to evidence-based medecine].

An example of clinical controversy related to EBM. The DCCT study for type 1 diabetes (1993) and UKPDS for type 2 (1998) had demonstrated the benefit of the hypoglycemic treatment with, for 1 point of HbA1c less, 30% less microangiopathy events. The cardiovascular benefit was less obvious, as these studies were performed in patients with primary prevention. The benefit became significant only during the follow-up, 10 years after the end of the study. After these results, it became unethical to consider randomized studies in which patients in the control group would have HbA1c greater than or equal to 8%. More recent studies in type 2 diabetics with a very high cardiovascular risk therefore compared "intensive" treatment to "good" treatment. The "super intensive" treatment of the ACCORD study, with an objective of HbA1c lower than 6%, resulted in increased mortality leading to premature discontinuation of the study. The excess cardiovascular mortality was correlated with the occurrence of very severe hypoglycaemia, three times more frequent in the "super intensive" group. In contrast to the concept "the lower is the better", patients with high cardiovascular risk should avoid the risk of severe hypoglycaemia secondary to treatment. In addition, the drug industry was asked to demonstrate the cardiovascular safety of the new antidiabetic molecules, regardless of their hypoglycemic effect, by conducting randomized double-blind studies with an identical target of HbA1c between the two groups. The adaptation of glycemic targets to cardiovascular risk in case of risk of hypoglycemia on the one hand, and the poor understanding of the methodology of cardiovascular safety studies on the other hand, have led to a radical questioning of the benefit of treating hyperglycemia with potentially serious consequences for patients.