Do co-morbidities, frailty and functional status identify patients who may not benefit from ICD therapy?

Abstract Funding Acknowledgements Type of funding sources: Private grant(s) and/or Sponsorship. Main funding source(s): Boston Scientific ISR lite. Background In carefully selected patients, Implantable Cardioverter Defibrillators reduce all-cause mortality by reducing the burden of sudden cardiac death. Most patients who have an ICD implant do not receive life-prolonging therapy from it. Little research has been undertaken to determine which patients benefit the least from ICD therapy. As patients age and accumulate comorbidities, the risk of death increases and the benefit of ICDs diminishes. Purpose We sought to evaluate the impact of comorbidity, frailty, functional status on death prior to ICD therapy as this group has the least to gain from ICD therapy. Methods A prospective, multicentre, observational study involving 15 English hospitals was undertaken. Patients were eligible for inclusion for the study if they were scheduled to have a de novo, replacement or an upgrade to a high-energy device (transvenous or subcutaneous ICD or CRT-D). Baseline characteristics were collected including age, gender, device type, LV function, medication, baseline bloods if available and were asked to complete a frailty assessment, Fried score; a functional status questionnaire (EQ5D5L), Charlson Co-morbidity Score and EQ5D5L data were calculated. Results In total, 662 patients were enrolled, mean age 65.6 (IQR 65-75) years, 77.3% male, 55.9% de novo implants, 33.4% replacements and 10.5% upgrade. Overall, 32% were for a secondary prevention indication, 49.% had a single or dual chamber transvenous ICD, 45.3% had a CRT-D and 5% had an S-ICD) implanted. A total of 63 patients (9.5%) died across this period, of these, 54 died without receiving appropriate therapy. Frailty was present in 83/662 (12.5%) and severe co-morbidity was present in 69/662 (10.4%). The overall survival probability estimated to be 90.2% at 2.5 years, and was 68% in frail patients. Figure 1. Multivariate predictors of death with no appropriate therapy were identified and a risk score comprising frailty, severe co-morbidity, age >70 years, eGFR <60 ml/min/1.73 m2 EQ5D5L <0.6 was developed. Table 1. The model performance was: sensitivity= 80.7%; specificity = 81.6%; precision 27.3%; accuracy 81.6% and AUC 0.81. Conclusion Frailty and severe co-morbidity are powerful, independent predictors of death with no appropriate in ICD/CRT-D recipients.