Simple tool accurately predicts 30-day mortality for heart failure patients in the ED

The tool is based on 13 independent risk factors, including systolic blood pressure, age, N-terminal pro B-type natriuretic peptide level, potassium level, troponin level, respiratory rate, and oxygen saturation.


A simple tool using readily available data can accurately estimate the 30-day mortality risk for patients admitted to the ED with acute heart failure, a study found.

Researchers for the Spanish Ministry of Health used data from an established registry of 34 Spanish EDs with diverse patient and facility characteristics to create the prediction tool model. They identified 88 variables and tested them against outcomes in a cohort of 4,867 consecutive ED patients admitted with acute heart failure from 2009 to 2011. Within 30 days of ED admission, 500 patients (10.3%) died. The findings were published on Oct. 3 by Annals of Internal Medicine.

Thirteen independent risk factors were identified in the derivation cohort and were combined into an overall score: Barthel index score at admission, systolic blood pressure, age, N-terminal pro B-type natriuretic peptide level, potassium level, troponin level, New York Heart Association class at admission, respiratory rate, low-output symptoms, oxygen saturation, episode association with acute coronary syndrome, hypertrophy on electrocardiography, and creatinine level.

The variables were tested in an independent validation population of 3,229 patients gathered three years later from the same EDs. In this validation cohort, 299 (9.26%) patients died within 30 days of ED admission. The score had excellent discrimination and calibration for predicting 30-day mortality, the researchers found. Thirty-day mortality risk was dramatically different across groups defined by the score (<2% for patients in the two lowest risk quintiles and 45% in the highest risk decile). Multiple sensitivity analyses did not find substantial confounding or bias.

“Identification of both groups [high- and low-risk] has important management implications,” the authors wrote. “For a patient with very high risk, attention should be focused on ensuring that the patient and their relatives are aware of the severity of the situation. In addition, the patient should receive prompt and aggressive treatment if appropriate, with an emphasis on early admission to an intensive care unit. For a patient with low risk, attention should be focused on treatment that will lead to early discharge from the ED to home, which is consistent with a recent consensus opinion about patients with less than 2% all-cause mortality who undergo sufficient observation in the ED.”

An editorial noted that this is the fourth major study aiming to define prognosis of heart failure patients in the ED, all of which claimed excellent discrimination. “If any of these models are to gain acceptance, they will need to be prospectively tested in diverse populations. That is the easy part; the next steps are more challenging. If 40% of ED patients with [heart failure] are truly at very low risk, we must find commonalities among them. This information may guide development of an alternate infrastructure to successfully treat these patients out of the hospital,” the editorialist wrote.