Published in the February 2, 2005 edition of JAMA, the new tool utilizes the combination of three simple measures obtained through laboratory blood tests and by measuring vital signs. Heart failure is a condition that affects five million Americans and is the leading cause of hospitalization for those over age 65.
"The new tool is a first for the treatment of acute heart failure, and offers a simple quick way for clinicians to assess mortality risk upon hospital admission and quickly decide on a treatment strategy," said Dr. Gregg C. Fonarow, lead study author, The Eliot Corday Chair in Cardiovascular Medicine and Science, professor of cardiology, David Geffen School of Medicine at UCLA and director, Ahmanson-UCLA Cardiomyopathy Center.
Using data from a national registry of over 100,000 heart failure patients called the Acute Decompensated Heart Failure National Registry (ADHERE), researchers developed a risk model after analyzing 33,046 hospitalizations. The model was developed using a relatively new statistical technique know as Classification and Regression Tree Analysis (CART). The validity of the model was then tested using data from an additional 32,229 hospitalizations.
Researchers evaluated 39 possible factors as survival indicators upon hospital admission and found that the best single predictor for mortality was a high blood urea nitrogen level, (above 43 mg/dL), followed by a low systolic blood pressure (above 115 mm Hg) and a high serum creatinine (higher than 2.75 mg/dL).