Objective The frequency and predictors of improvement in left ventricular ejection fraction (LVEF) in ischaemic cardiomyopathy and its association with mortality is poorly understood. We sought to assess the predictors of LVEF improvement ≥10% and its effect on mortality.

Methods We compared characteristics of patients enrolled in The Surgical Treatment for Ischaemic Heart Failure (STICH) trial with and without improvement of LVEF ≥10% at 24 months. A logistic regression model was constructed to determine the independent predictors of LVEF improvement. A Cox proportional hazards model was created to assess the independent association of improvement in LVEF ≥10% with mortality.

Results Of the 1212 patients enrolled in STICH, 618 underwent echocardiographic assessment of LVEF at baseline and 24 months. Of the patients randomised to medical therapy plus coronary artery bypass graft surgery (CABG), 58 (19%) had an improvement in LVEF >10% compared with 51 (16%) patients assigned to medical therapy alone (p=0.30). Independent predictors of LVEF improvement >10% included prior myocardial infarction (OR 0.44, 95% CI: 0.28 to 0.71, p=0.001) and lower baseline LVEF (OR 0.94, 95% CI: 0.91 to 0.97, p<0.001). Improvement in LVEF >10% (HR 0.61, 95% CI: 0.44 to 0.84, p=0.004) and randomisation to CABG (HR 0.72, 95% CI: 0.57 to 0.90, p=0.004) were independently associated with a reduced hazard of mortality.

Conclusions Improvement of LVEF ≥10% at 24 months was uncommon in patients with ischaemic cardiomyopathy, did not differ between patients assigned to CABG and medical therapy or medical therapy alone and was independently associated with reduced mortality.

Trial registration number NCT00023595.