It is believed that the proportion of patients who develop pulmonary complications after undergoing cardiac surgery is 8%. How large should the sample size for two groups (Intervention vs. Control) if...

It is believed that the proportion of patients who develop pulmonary complications after undergoing cardiac surgery is 8%. How large should the sample size for two groups (Intervention vs. Control) if an investigator wishes to detect a relative reduction of 25% in the pulmonary complication rate after implementing the preoperative intervention, assumed a power of 90% and a Type I error of 0.05?