FIG. 28:  The primary endpoint of SERVE-HF was time to first event (death from any cause, unplanned hospitalizations for worsening heart failure, cardiac transplant, resuscitation of cardiac arrest, or appropriate ICD shock for ventricular arrhythmia).^[15]  Following randomization and initiation of ASV, follow-up visits were done annually; this means the trialists were only seeing the patients once a year, which is not very much (most heart failure trials would see the patients twice a year).  The Data and Safety Monitoring Board (DSMB) performed only two interim analyses over the 7-year period, when 50% and 75% of the predicted events occurred.