Response to: 'baseline differences in SAVOR trial' – prespecified vs. post hoc analysis – a potential source of bias. Issue 12 (December 2015)