Predicting defibrillator benefit in patients with cardiac resynchronization therapy: A competing risk study

Cardiac resynchronization therapy (CRT) reduces morbidity and mortality in selected heart failure patients, but decision-making regarding selection of CRT-defibrillator or CRT-pacemaker is an ongoing debate.

The purpose of this study was to construct predictive models and scoring systems for implantable cardioverter-defibrillator (ICD) therapy and death without ICD therapy (prior death).

We pooled 2 prospective cohorts of CRT-D patients with primary prevention indication and used Fine and Gray models to develop independent prognostic models for time to first ICD therapy (event of interest) or death without prior ICD therapy (competing event). We defined CRT-D benefit as a high probability of ICD therapy combined with moderate/low probability of prior death.

Seven hundred twenty patients were included. Median follow-up was 7.2 years, and 247 patients (34%) died. Cumulative incidence of ICD therapy/prior death at 5 years was 24%/17%. In multivariable models, higher New York Heart Association classes, diuretic use, and ischemic cardiomyopathy were predictors of ICD therapy (hazard ratio 1.89 [1.30-2.75], 1.91 [1.12-3.24], and 1.40[1.02-1.92], respectively) but not of prior death. Males with comorbidities (cancer, renal failure, peripheral artery disease, body mass index >30) or systolic blood pressure ≤100 were at higher risk for prior death. Higher age was associated with lower risk of ICD therapy but higher risk of prior death. One-quarter of patients had low predicted benefit from CRT-D implantation using a scoring system for the dual prediction of appropriate ICD therapy and death without appropriate ICD-therapy.

Different factors predict ICD therapy or prior death in CRT-D patients using competing risk models. Scoring allows identifying patients with predicted low benefit of CRT-D (low chance of ICD therapy, high chance of prior death).