Designation and distribution of events in the Multicenter UnSustained Tachycardia Trial (MUSTT).

Patients with coronary artery disease, depressed left ventricular ejection fraction, and nonsustained ventricular tachycardia (VT) have a high mortality rate due to arrhythmic (arrhythmic death/cardiac arrest) and other cardiac causes. The Multicenter UnSustained Tachycardia Trial (MUSTT) investigated whether electrophysiologic study (EPS) was helpful in choosing drug or defibrillator therapy in patients induced into sustained VT. The events committee attempted to categorize follow-up events in patients in MUSTT and to present a detailed breakdown of events. A derivative of the Hinkle-Thaler classification was used, incorporating lessons from other multicenter studies. The committee was blinded to results of EPS and implantable cardioverter-defibrillator (ICD) or other antiarrhythmic therapy status of patients. The primary end point was cardiac arrest or death from arrhythmia. Secondary end points were death from all causes, cardiac causes, and spontaneous sustained VT. Classifications were death and cardiac arrest. Each was similarly divided as arrhythmic with 14 subcategories, e.g., unwitnessed or related to EPS and nonarrhythmic with 10 subcategories, e.g., ischemia. Terminal VF in progressive heart failure was considered nonarrhythmic. Events were reviewed by 2 members. Disagreements were resolved by the 2 members or, if needed, by the full committee. Of the 2,202 patients in MUSTT, there were 902 deaths. Sustained VT requiring cardioversion occurred in 182 patients. An additional 94 patients had resuscitated cardiac arrests. Events occurred in 1,027 patients, and all were reviewed. The 3 leading events were deaths that were classed as sudden/unwitnessed (23% of 902), due to progressive heart failure (22%), or due to noncardiovascular causes (18%). Arrhythmic deaths or cardiac arrests were highest in inducible patients randomized to no antiarrhythmic therapy; next were inducible patients receiving an ICD; and lowest were in patients who were noninducible. In conclusion, the classification system provided a detailed breakdown of events in consistent categories, showing utility for event analysis and interpretation and development of therapeutic strategies. The classifications assigned by the committee were used in all MUSTT outcomes reports, thus affecting all reported outcomes and overall interpretations of the MUSTT.

Duke Scholars

Author Kerry L. Lee Biostatistics & Bioinformatics, Division of Biostatistics