Should methods of correction for multiple comparisons be applied in pharmacovigilance? Reasoning around an investigation on safety of oral antidiabetic drugs