Introduction: About 25% of patients who develop left ventricle (LV) systolic dysfunction will have improvement in LV ejection fraction (EF) overtime. This patient cohort is generally excluded from large sudden death trials and hence under studied.

Methods: Patients who had transient LV systolic dysfunction between 2010-2014 within Aurora Health System who had LVEF improved≥40%, irrespective of implantable cardioverter-defribillator (ICD) implant, were studied and predictors of mortality were identified using Cox proportional hazard model.

Patients were then divided into groups based on LVEF >50% or

Results: 1364 patients met inclusion criteria. 58.4% were male and mean BMI was 29+/-7. Mean age at improved LVEF was 66 years+/- 14, and with each added year the hazard rate increased by 5% [Hazard Ratio (HR) 1.05,p =0.0001]. Several clinical characteristics emerged as predictors of mortality, including smoking (HR 1.8, p=0.0002), chronic renal disease (HR 2.3,p = 0.0001), atrial fibrillation (HR 1.4, p=0.013) and no-ICD (HR 2.1, p=0.012). But with each percentage increase in LVEF hazard rate decreased by 2% (HR 0.97, p=0.007). However, presence of ICD did not significantly improve mortality in the group with LVEF >50% (figure).

Conclusions: Clinical predictors of mortality in patients with transient LV systolic dysfunction may help further risk stratify this cohort of patients. It appears that patients with LVEF 40-49% continue to derive benefit from ICD therapy.