Objectives: To evaluate whether the Enhanced Feedback for Effective Cardiac Treatment (EFFECT) scale for 30-day prediction of mortality is applicable to elderly adults with acute heart failure (AHF) in emergency departments (EDs) and whether discriminatory power is added with the inclusion of the Barthel Index (BI) to this scale (BI-EFFECT scale).
Design: BI-EFFECT is a multipurpose, nonintervention, multicenter cohort study.
Setting: Twenty EDs.
Participants: Individuals aged 65 and older with AHF.
Measurements: Information on baseline and episode characteristics and 30-day mortality was collected, and participants were categorized according to the EFFECT scale. Baseline degree of functional dependence was measured using the BI. Receiver operating characteristic (ROC) curves were made of the EFFECT and BI-EFFECT scales to predict mortality.
Results: One thousand sixty-eight participants were included. Thirty-day mortality was 5.1% and was directly and independently associated with high and very high risk categories of the EFFECT scale and with severe dependence. These two variables remained significant after adjustment of the model for both (OR = 4.5, 95% CI = 1.8-11.1 and OR = 2.9, 95% CI = 1.6-5.4, respectively). The EFFECT and the BI-EFFECT scales had significant ROC curves (area under the ROC curve (AUC) = 0.69, 95% CI = from 0.62 to 0.76; and AUC = 0.75, 95% CI = 0.69-0.81, respectively), and the difference in discriminatory power between the second and the first was also statistically significant (P = .02).
Conclusion: The EFFECT scale may be applied in the elderly population, and inclusion of functional status according to the BI in the new BI-EFFECT scale significantly improves the model for the prediction of 30-day mortality.
© 2012, Copyright the Authors Journal compilation © 2012, The American Geriatrics Society.