|First Author||Year||Ref. #||Non-Arrhythmic Mortality Risk Factors||Comparator Group∗||ICD Benefit Cut-Point†||CRT Included?|
|Parkash||2006||(9)||Age >80 yrs, AF, Cr >1.8, NYHA ≥III, Co-morbidity‡||No||N/A||No|
|Barsheshet||2012||(12)||BUN >26, NYHA ≥III, AF, age >70 yrs, QRS >120||Yes||3 RFs||No|
|Bilchick||2012||(13)||CKD, age >75 yrs, COPD, DM, NYHA ≥III, AF, EF <20||No||N/A||No|
|Steinberg||2014||(14)||Smoking, ischemic HD, CKD, DM, COPD, PVD||Yes||∼3–4 RFs||No|
|Providencia||2015||(1)||GFR <60, NYHA ≥III, AF, age >70 yrs, QRS >120||No||N/A||Yes|
AF = atrial fibrillation; BBL = beta-blocker; BUN = blood urea nitrogen; CKD = chronic kidney disease; Cr = creatinine; CRT = cardiac resynchronization therapy; DM = diabetes mellitus; EF = ejection fraction; GFR = glomerular filtration rate; ICD = implantable cardioverter-defibrillator; NYHA = New York Heart Association functional class; RF = risk factor; SCD-HeFT = Sudden Cardiac Death in Heart Failure Trial.
↵∗ Comparator Group: Inclusion and comparison of ICD and no-ICD patients with or without various risk factors = Yes; identification of factors associated with mortality in ICD-only cohort = No.
↵† ICD Benefit Cut-Point: The concept of “ICD benefit cut-point” addresses the co-morbidity burden above which an ICD might not improve survival.
↵‡ Co-morbidity factors included dementia, cancer, pneumonia, or infection.
↵§ Koller et al. (10): Total mortality predicted by age (per 10-yr increase), LVEF (per 10% decrease), absence of beta-blocker use, diuretic use, more recent implant era. However, only diuretic use predicted death before appropriate ICD therapy.
↵‖ SHFM (Seattle Heart Failure Model): includes age, sex, ischemic etiology, NYHA, EF, SBP, K-sparing diuretic, statin use, allopurinol use, hemoglobin, percent lymphocyte count, uric acid, sodium, cholesterol, and diuretic dose/kg. Patients in the SCD-HeFT placebo arm having the highest mortality rates by quintile did not have better survival with an ICD than with placebo.