Authors reply: Diabetes treatments and risk of amputation, blindness, severe kidney failure, hyperglycaemia, and hypoglycaemia: open cohort study in primary care
We thank Schrijnders et al for their detailed and thoughtful response to our recent paper on clinical outcomes in type 2 diabetes among patients prescribed different glucose lowering agents1. We agree that there are likely to be within class differences for different drugs and this is likely to include different sulphonylureas and risk of hypoglycaemia. However, as we stated in the paper, our main research question was concerning the risks associated with the newer agents such as glitazones and gliptins and this question informed the study design. We used an incident user design for new users of gliptins and glitazones as a result but prevalent users of metformin or sulphonylureas were included in the main study cohort. To do a robust analysis focusing on sulphonylureas as the main exposure, we would similarly need to do an incident user design focusing on patients who had a first prescription of sulphonylureas during our study period. To then undertake post-hoc analyses for each individual sulphonylurea would be a significant task given the number of different drugs within this class. We think this is an important research question which would be more appropriately addressed with a new study than a post-hoc analysis of a dataset assembled to answer a different question so that the relevant results can be presented and discussed in sufficient detail.
Competing interests: see statement in original paper