Score Predicts Risk for Death in Heart FailureBy Todd Neale, Senior Staff Writer, MedPage Today
Published: June 04, 2012
Reviewed by Robert Jasmer, MD; Associate Clinical Professor of Medicine, University of California, San Francisco and Dorothy Caputo, MA, BSN, RN, Nurse Planner
Take PosttestAction PointsA risk score, called the Emergency Heart Failure Mortality Risk Grade, derived from variables that are routinely gathered in the emergency department may help guide the decision to admit or discharge patients presenting with acute heart failure.Note that there were 10 variables significantly associated with a greater likelihood of dying in the first week that were included in the risk score, including transportation by emergency medical services, elevated serum troponin, use of metolazone at home, and active cancer.
A risk score derived from variables that are routinely gathered in the emergency department may help guide the decision to admit or discharge patients presenting with acute heart failure, researchers found.
The 10-factor algorithm showed a high level of discrimination in predicting the likelihood of patients dying within 7 days of presentation in both a derivation and validation cohort, according to Douglas Lee, MD, PhD, of the Institute for Clinical Evaluative Sciences in Toronto, and colleagues.
The areas under the receiver-operating characteristic curves were 0.805 for the derivation cohort and 0.826 for the validation cohort, higher than those seen with previous risk tools based on data from hospitalized patients (0.528 to 0.707), the researchers reported in the June 5 issue of the Annals of Internal Medicine.
The new risk tool -- called the Emergency Heart Failure Mortality Risk Grade (EHMRG) -- may help physicians decide whether a patient with acute heart failure requires a costly hospitalization or can be sent home safely, a decision typically made using clinical judgment alone.
"The caveat that applies to all decision-support tools is pertinent, because the EHMRG is intended to assist the clinician, who should ultimately synthesize the risk score with clinical judgment, because the decision to hospitalize or discharge patients with heart failure weighs several potential factors," the authors wrote.
"Symptomatic improvement, ability of the patient to seek follow-up care, and social circumstances should also be considered, along with quantification of acute prognosis," they said.
To develop the algorithm, Lee and colleagues collected data from 12,591 patients presenting with acute heart failure to 86 hospitals in Ontario from April 2004 through March 2007. There were 7,433 patients in the derivation cohort and 5,158 in the validation cohort. Roughly two-thirds of the patients were admitted to the hospital.
Within 7 days of presentation, 2% of the patients died, and the researchers examined numerous vital signs, clinical and presentation features, as well as laboratory tests to identify factors associated with mortality.
After multivariate adjustment, there were 10 variables that were significantly associated with a greater likelihood of dying in the first week that were included in the risk score: Transportation by emergency medical services (OR 2.84, 95% CI 1.92 to 4.21) Elevated serum troponin (OR 2.75, 95% CI 1.86 to 4.07) Use of metolazone at home (OR 2.65, 95% CI 1.07 to 6.61) Active cancer (OR 2.11, 95% CI 1.33 to 3.33) Potassium level of 4.6 mmol/L or higher (OR 1.71, 95% CI 1.13 to 2.58) Lower triage systolic blood pressure (OR 1.52, 95% CI 1.31 to 1.77, for every 20 mm Hg) Older age (OR 1.40, 95% CI 1.16 to 1.69, for every 10 years) Higher creatinine (OR 1.35, 95% CI 1.14 to 1.60, for every 1 mg/dL) Reduced oxygen saturation (OR 1.16, 95% CI 1.01 to 1.33, for every 5%) Higher triage heart rate (OR 1.15, 95% CI 1.03 to 1.30, for every 10 beats per minute)
The researchers divided the patients into six groups according to their risk score. From the lowest to the highest scores, the 7-day mortality rates were 0.3%, 0.3%, 0.7%, 1.9%, 3.5%, and 8.2% in the derivation cohort, with similar results in the validation cohort.
Overall, 27% of patients had a predicted mortality risk greater than the observed mortality rate of 2%.
The authors acknowledged some limitations of the study, including the lack of information on left ventricular ejection fraction and brain natriuretic peptide in the risk score, and the determination of medication use based on prescriptions in the medical record.
In addition, the researchers noted, the algorithm is not for use in patients with chronic, symptomatically stable heart failure.
The Institute for Clinical Evaluative Sciences is supported in part by a grant from the Ontario Ministry of Health and Long-Term Care. The research was supported by an operating grant from the Canadian Institutes of Health Research, a Canadian Institutes of Health Research clinician-scientist award, a Career Investigator Award from the Heart and Stroke Foundation of Ontario, a Canadian Institutes of Health Research Applied Chair in Health Services and Policy Research, and a Canada Research Chair in Health Services Research.
Lee reported submitting a U.S. patent application based on the research described in the current study along with one of his co-authors, and the decision is still pending. That co-author reported that he received money through his institution from a Canadian Institutes of Health Research Team Grant in Cardiovascular Outcomes Research.
From the American Heart Association:
Acute Heart Failure Syndromes in the Emergency Department Decision Making in Advanced Heart Failure
Primary source: Annals of Internal Medicine
Source reference:
Lee D, et al "Prediction of heart failure mortality in emergent care: a cohort study" Ann Intern Med 2012; 156: 767-775.
