Trial design has a major impact on events rates

A systematic review of adverse events in clinical trials for type 2 diabetes medications has highlighted the importance of quantifying baselines for cardiovascular disease (CVD) history and proteinuria. 

The authors examined all-cause death and CVD event rates in 29 large-scale randomised controlled trials that had at least 1000 participants, published 1998 to mid-2010. 

For all-cause death, the presence of CVD at baseline was associated with a threefold higher death rate compared with trials with no baseline CVD. CVD at baseline also led to a fivefold increased risk of CVD death. Patients with baseline proteinuria had six times the risk of all-cause death and 16 times the risk of CVD death. When there was no baseline of CVD history or proteinuria, events rates were low. 

Therefore, trials including diabetic participants without CVD or proteinuria generate few events and require substantial participant numbers to achieve adequate power