The analysis shows that more than half of the patients who would be good candidates to receive implanted cardioverter-defibrillators (ICDs) according to criteria from the MADIT II clinical trial would not qualify to be reimbursed for them under Medicare's current guidelines.
And some of them will die from sudden cardiac death that could probably have been prevented by an ICD, a pacemaker-like implanted device that monitors heart rhythm and shocks the heart back into a normal pace when an arrhythmia develops.
Today at the Annual Scientific Session of the American College of Cardiology, University of Michigan Cardiovascular Center researcher Richard Otten, M.D., will present the results of the study in an oral presentation and a press conference.
Otten and his colleagues analyzed data on patients who were treated at U-M for acute coronary syndrome in the three years before ICDs received FDA approval for low ejection fraction heart failure patients with a history of myocardial infarction, also called heart attack. That approval was based largely on the results of MADIT-II, a national study published in 2002 in the New England Journal of Medicine.
In June 2003, the Centers for Medicare and Medicaid Services decided to cover ICDs only for Medicare patients who had heart failure, myocardial infarction and a QRS interval greater than 120 milliseconds. The QRS interval is a parameter on electrocardiograms that represents the time it takes for the heart's ventricles to depolarize.
That cutoff was based on a sub-analysis of MADIT-II data, though MADIT-II was not designed to determine how the QRS interval affected patients' likelihood of sudden cardiac death. Private insurers have tended toward covering patient in accordance with MADIT-II.
Otten's presentation will show how, just at the U-M center alone, 140 patients over three years would theoretically have qualified for an ICD under MADIT-II criteria, regardless of their QRS score. That's 10 percent of all patients with acute coronary syndromes (heart attack or unstable angina) treated at U-M in the three-year period.
But had those patients all tried to get coverage for an ICD under Medicare, 80 of them would have been denied under Medicare's current criteria, because their QRS score was too low.
"Assuming all the patients were under Medicare, that would have saved Medicare, and therefore society, $989,000 annually under current reimbursement rates," says Otten, a research fellow in the U-M Medical School's Department of Cardiovascular Medicine. If all 140 U-M patients received ICDs, the total cost would be about $1.73 million annually.
"But the savings would also come at the cost of about half of the sudden cardiac deaths that could be expected in the entire group," Otten continues. "That's because we found no difference in death or rehospitalization between those patients with a QRS interval less than 120 milliseconds, and those with a QRS interval greater than 120 milliseconds, when we checked in with them six months after hospitalization."
If all 140 patients had received an ICD, MADIT-II data suggest that their risk of a rhythm-related death would be 61 percent lower than if they had not received the device. In the conventional therapy, non-ICD arm of the MADIT-II study, 53 percent of the deaths were due to sudden cardiac death.
The U-M analysis estimates that approximately two of the 140 patients a year would be saved from sudden cardiac death if all the patients received an ICD. Even if some of the 140 patients died from non-rhythm-related causes, the net result would be that one person a year from the original 140 would be saved if all the patients received ICDs.
The senior author on the study, U-M Cardiovascular Center clinical director Kim Eagle, M.D., FACC, notes that the another study coming out at the ACC meeting may add even more evidence to the debate over whether Medicare should alter its coverage criteria for ICDs. The National Institute of Heart, Lung, and Blood Institute Sudden Cardiac Death in Heart Failure Trial, also known as SCD-HeFT, showed that ICDs reduced the risk of sudden cardiac death by 23 percent among patients with moderate heart failure, compared with those who took anti-arrhythmia medication.
Hopefully, the addition of more data to the debate will help guide coverage decisions for Medicare and others, Otten notes.
"As clinicians, we're in the uncomfortable position of knowing that clinical trial data show that more people could be helped by this technology, but being unable to provide it to them unless they're able to pay out of pocket," he says. "This debate shows once again the balance that our society is having to strike, regarding how much money we're willing to pay to save a life."
In addition to Otten and Eagle, the study's authors include others at the Michigan Cardiovascular Outcomes Research and Reporting Program, including Eva Kline-Rogers, MS, RN, NP; Jianming Fang, M.D.; Darryl Elmouchi; Aman Chugh, M.D., Firas Marayati; and Frank Pelosi, Jr., M.D.