In Reply: Dr Kaneda is correct. In the original version of the SORT OUT II study protocol, we hypothesized major adverse cardiac event (MACE) rates of 20% over 5 years and selected a 25% relative reduction of MACE rates as the minimal relevant difference, which required randomization of 1800 patients (2-sided P < .05; study power, 0.80). The protocol also prespecified an interim analysis after 9 months. At this earlier time point, the event rate must be smaller than after 5 years. The event curves were expected to be curvilinear with an initial steep rise in events followed by a smaller slope, and we assumed a 9-month MACE rate of 10.0% and a relative difference of 25% (ie, 7.5% vs 10%), assumptions that would have required randomization of 4006 patients to achieve the same power.
Galløe AM, Bligaard N, Hansen PR. Study Power in the SORT OUT II Trial—Reply. JAMA. 2008;299(17):2021-2022. doi:10.1001/jama.299.17.2022-a