By Michael H. Crawford, MD

Professor of Medicine, Chief of Clinical Cardiology, University of California, San Francisco

Dr. Crawford reports no financial relationships relevant to this field of study.

SYNOPSIS: This cardiac catheterization-based study of patients with newly diagnosed reduced left ventricular ejection fraction of unknown etiology showed that 15% had ischemic cardiomyopathy and they could be identified by clinical characteristics and an ECG-based risk score.

SOURCE: Smilowitz NR, Devanabanda AR, Zakhem G, et al. Comparison of clinical and electrocardiographic predictors of ischemic and nonischemic cardiomyopathy during the initial evaluation of patients with reduced (≤ 40%) left ventricular ejection fraction. Am J Cardiol 2017;119:1650-1655.

In newly discovered reduced left ventricular ejection fraction (LVEF), it is considered appropriate to perform invasive coronary angiography to identify ischemic cardiomyopathy (ICM), which is potentially treatable by revascularization. However, are angiograms always necessary? Investigators from New York University retrospectively identified patients referred for left heart catheterization with a new diagnosis of LVEF < 40% by echocardiography between 2010 and 2014 to see if they could develop a clinical risk score to predict the likelihood of obstructive (> 70% diameter reduction) coronary artery disease (CAD). Patients were excluded if the etiologic diagnosis was obvious, such as those admitted with acute coronary syndrome, previous revascularization or myocardial infarction, severe left heart valve disease, right ventricular pacing for conduction disease, or had an established non-ischemic cardiomyopathy diagnosis.

Inclusion criteria were met in 273 of the 5,030 patients referred for coronary angiography, and ischemic cardiomyopathy (ICM) was found in 41 patients (15%). Patients with ICM were older, more likely to have diabetes, peripheral arterial disease, and use tobacco. Also, ICM patients were more likely to have ECG evidence of Q wave infarction (34% vs. 13%; P < 0.001) and ischemic ST-T changes (22% vs. 9%; P = 0.02), but left bundle branch block was less likely (2% vs. 15%; P = 0.03). A risk model including all clinical and ECG data was highly predictive of ICM (C statistic = 0.81). A simplified model that only included age, hypertension, diabetes, tobacco, Q wave infarction, and ST-T changes on ECG also was highly predictive (C = 0.80). When the risk score (range -1 to 9 points) was dichotomized at 3.5, the negative predictive value was 95%. The authors concluded that specific clinical and ECG abnormalities could be used to estimate which patients with reduced LVEF were at low risk of having ICM.


The issue of performing an invasive evaluation of the coronary arteries in patients with newly diagnosed LV systolic dysfunction of unclear etiology is important because myocardial ischemia is potentially treatable. Many centers, such as the one where this study was performed, have had a policy that all such patients undergo a left heart catheterization. This study hypothesized that clinical and ECG features of the patient may be able to identify a low-risk group in which catheterization is not mandatory. They found that four clinical factors (age, hypertension, diabetes, and tobacco use) and two ECG abnormalities (Q waves and ischemic ST-T changes) could be used to calculate a point score that would identify a low risk of ICM group. The feature with the highest points assigned was age > 65 years (3 points). Age 55-64, diabetes, and Q waves all were assigned 2 points each. All the rest were 1 point, except for hypertension, which was scored as a -1. The maximum score was 9, but any score < 3.5 was considered a low risk for ICM. The up to 3.5-point cutoff for the low risk of ICM group had a negative predictive value of 95%. From a qualitative approach, if a patient is < 65 years of age and does not have diabetes or Q waves, their score will not be > 3.5, even if they have all the other factors in the score, so invasive angiography could be avoided in such patients. Since the false-negative rate is not zero perhaps in those with low scores, a non-invasive test for coronary disease should be used, such as stress testing or CT angiography. Other interesting findings were that 12% of non-ICM patients had ECG Q waves and only 2% of ICM patients had left bundle branch block. Both these findings have been found in other studies, and the paucity of left bundle in ICM is well-known.

There are several weaknesses to this study. It is a single-center retrospective study and suffers from the selection bias of catheterization-referred patients. However, at this center the policy was to perform invasive coronary angiography in this type of patient. CAD was defined as significant obstructive disease, but this is the disease most likely to be amenable to revascularization. Right ventricular dysfunction on echocardiography also is known to be associated with non-ICM, but wasn’t considered in this study. Also, the authors didn’t look at segmental left ventricular wall motion abnormalities either, perhaps because previous studies have not shown this to be a useful discriminator.

I believe many clinicians are using clinical judgment based on the characteristics of the patient, the ECG, and the echocardiogram to cull the low likelihood group and not performing invasive angiography on them. However, this paper provides a more systematic approach, which would promote more uniformity in decision-making.