March 13, 2006
Seattle Heart Failure Model is able to accurately predict survival rates for patients with heart failure
A new model developed at the University of Washington provides an accurate estimate of one-, two-, and three-year survival rates and average years of survival for patients with heart failure. The model incorporates medications and devices that are used to treat heart failure and how altering these affects survival.
The Seattle Heart Failure Model was created by Dr. Wayne C. Levy, associate professor of medicine in the Division of Cardiology at the UW, in collaboration with 13 co-authors. It is now available online at http://circ.ahajournals.org/ and will be published March 21 in the journal Circulation.
Heart failure has a mortality rate that can range from 5 percent to 75 percent per year. Patients and clinicians have not had an easy way to estimate survival. The Seattle Heart Failure Model was developed using very simple clinical and laboratory variables that are available to any health care provider. Some of these include age, gender, blood pressure, weight, heart failure medications/devices, and simple laboratory variables like hemoglobin, cholesterol, uric acid, and serum sodium. The model was derived by examining 1,125 heart failure patients, and validated in five additional groups, totaling 9,942 patients. The accuracy of the model was excellent.
“What is unique about this model is that one can estimate the change in an individual patient’s survival by adding medications or devices used to treat heart failure,” Levy said. “For example, a heart failure patient treated with only digoxin and diuretic therapy with a 20 percent annual mortality rate will live about four years on average. But according to the Seattle Heart Failure Model, if you add an ACE inhibitor the patient will live five years, and if you add an ACE inhibitor and a beta blocker the patient will live six and a half years.
“If you use an ACE inhibitor, beta blocker and an aldosterone blocker, the patient makes it to eight years, or double the original life span,” Levy said. “And if you add an implantable cardiovertor defibrillator (ICD) you would make it to nine and a half years.”
Heart failure medications are proven to be effective and are relatively inexpensive, as many are available in a generic formulation. However, in ADHERE, a 65,000-patient registry of heart failure patients admitted to the hospital, only 41 percent were taking an ACE inhibitor, and only 45 percent were on beta blockers.
“The question we are asking is: Why aren’t they on these proven life-saving heart failure medicines?” Levy said. “This model will actually illustrate why patients need to take them. The same applies to cardiac devices, such as biventricular pacemakers, implantable cardiovertor defibrillators, or left-ventricular assist devices. We are trying to encourage patients and physicians to use the medications and devices that we know will work in our heart failure patients,” he said.
The model may also be valuable in determining who should receive a heart transplant. “Unlike other organs, there is no ‘score’ for either listing patients for a heart transplant or to allocate who receives an organ,” Levy said. “This model could help determine who should be listed for a heart transplant and to allocate the heart to the highest-risk patient on the waiting list.”
Dr. David T. Linker, an associate professor of medicine in the Division of Cardiology at the UW and co-author of the paper, has developed a Web-based application to make this interactive model available to health care providers at www.seattleheartfailuremodel.org.