Author + information
Whether all patients with systolic heart failure (SHF) of unclear etiology should undergo coronary angiography is controversial. We sought to derive and validate a clinical prediction tool to “rule-out” coronary artery disease (CAD) as cause of SHF.
A derivation cohort was formed of consecutive patients with a primary diagnosis of SHF of unclear etiology with ejection fraction (EF) < 50% who underwent angiography over a five-year period. Patients with known CAD, prior myocardial infarction, or another etiology for SHF (eg, valve disease) were excluded. Using multivariate logistic regression analysis, independent predictors of severe CAD were identified and used to derive an algorithm that would insure 100% sensitivity in identifying patients with severe CAD (left main, 3-vessel, or 2-vessel involving the proximal left anterior descending). Subsequently, the defined algorithm was validated prospectively in a separate validation cohort o consecutive patients with identical inclusion criteria from 2 institutions.
Of 124 patients (58% male, age 55±11 yrs, EF 32±10%) in the derivation cohort, 27% had CAD (>50% stenosis), including 15% with severe CAD. Independent predictors of the absence of severe CAD included diabetes [OR 0.25 (CI 0.08-0.82); p=0.02] and ≥2 other risk factors [OR 0.26 (CI 0.08-0.84); p=0.02]; age and electrocardiographic Q waves or left bundle branch block (Q/LBBB) were borderline (p=0.06). Consequently, an algorithm of performing angiography only in patients with diabetes, ≥2 other risk factors, or Q/LBBB) would have identified 94% of patients with CAD, including 100% with severe CAD. In the validation cohort, of 143 patients [62% male; age 56±10 yrs, EF 22±9%], 37% had CAD and 19% had severe CAD. The algorithm identified 98% of patients with CAD and 100% of patients with severe CAD.
Severe CAD is not uncommon in patients with SHF of unclear etiology. We derived and validated a simple clinical algorithm that may accurately eliminate the need for angiography in as many as 30% of patients without CAD. Considering the prevalence of SHF and high cost of coronary angiography, the subsequent savings from applying our algorithm are substantial.
Poster Sessions, Expo North
Monday, March 11, 2013, 9:45 a.m.-10:30 a.m.
Session Title: Epidemiology, Risk Modeling and Prediction of Outcomes in Heart Failure
Abstract Category: 15. Heart Failure: Clinical
Presentation Number: 1308-298
- 2013 American College of Cardiology Foundation