Safety end points included adverse laboratory and events abnormalities. Statistical Analysis We estimated that 285 events would be needed to detect a hazard ratio for the primary end point with macitentan , as compared with placebo, of 0.55 over around maximum study duration of 4.1 years, assuming an anticipated hazard rate of 0.43 in the placebo group, an expected 5 percent annual attrition price, and an annual enrollment of 200 individuals. The type I mistake was set at 0.005 for the comparison of placebo with each dose of macitentan, with the use of Bonferroni correction to make sure an overall alpha degree of 0.01, and power was set at 90 percent. A well planned blinded reestimation of the sample size was performed three months prior to the end of the anticipated recruitment phase because the overall hazard rate was lower than anticipated, resulting in an increase in recruitment from 525 to 699 patients.