January 20, 2005
Shocking
Top officials at Medicare – the retirement program that unlike Social Security really is going broke – today announced a $3 billion to $5 billion expansion in annual benefits. The decision will double or triple the number of patients with congestive heart failure (but no previous history of heart attacks) eligible for implantable cardioverter-defibrillators (ICDs), which send a massive electric shock to the heart after a serious coronary incident. Each device costs $30,000 to $40,000. The eligible population now totals more than half a million.
Despite the fact today is inauguration day (and a holiday here in the nation’s capital), the Washington Post gave this story the front page coverage it deserved. The New York Times buried it inside and the Wall Street Journal covered it with wire reports, which was odd given the financial implications for Medtronic, Guidant and St. Jude Medical, the three publicly-traded medical device manufacturing companies with the most to gain from the decision.
The ICD story is a classic case of medical bracket creep. Once a medical intervention is proven effective in some patients, the companies that make the drug or device support clinical trials to expand it to groups where there will be diminishing, even if still positive, medical returns. While the latest trial of ICDs, reported in today’s New England Journal of Medicine, was primarily funded by the government’s National Heart Lung Blood Institute, the device industry’s fingerprints were all over the study. Lead author Gust H. Bardy of the Seattle Institute for Cardiac Research received research support from Medtronic and serves as a consultant for Guidant. Alan Kadish, the Northwestern University medical professor who wrote the accompanying editorial, received research support from all three manufacturers.
At first glance, the trial’s results seemed impressive. ICDs reduced the death rate from 29 percent to 22 percent over four years. In other words, instead of 7.2 in 100 patients with congestive heart failure dying each year, only 5.5 died with the implanted defibrillators.
But the trial raised as many questions as it answered. The trial’s architects deliberately included less sick individuals measured by either the ability of their left ventricle to pump blood or by their performance on standard measures like a stress test. Yet the sicker patients in this latter group showed no benefit!
Moreover, inappropriate firing is emerging as a major issue for the tens of thousands who now have these devices stuck in their chests. Only 31 percent of the patients in the trial had their ICDs go off during the study period, but one in three of those events (or about 10 percent of all firings) were inappropriate. Cardiologists tell me their patients compare it to being kicked in the chest by a mule. That’s something even a liberal like myself wouldn’t wish on Vice President Dick Cheney, the nation’s most famous ICD patient.
Expanded ICD usage is also raising ethical issues about end of life care. Admittedly, these are anecdotal reports. But in the wake of the last Medicare expansion of ICD reimbursement in 2003, aggressive marketing by the manufacturers and physicians specializing in implantation resulted in many older patients with deteriorating hearts – including some who were demented – getting ICDs. Some of them later experienced defibrillation storms that left them begging to be released from their agony, according to some cardiologists with whom I’ve talked.
Most of these issues were discussed at the last Medicare Payments Advisory Committee meeting that discussed this issue. A number of industry consultants sat at the table at that February 2003 meeting, which considered evidence from a previous industry-funded trial. The Center for Medicare and Medicaid Services (CMS) hired a consultant from Johns Hopkins to poke holes in the data, but the physicians with conflicts of interest dominated the discussion.
CMS’s response to this latest trial was to grant the expanded coverage but demand reports via mandatory registration so it knows how ICDs are used in the general population. It also wants to know the medical results. The Post article suggested this raises a tricky ethical question, comparing it to requiring people to participate in a clinical trial to gain access to approved medical care.
But that gets the question exactly backwards. Reporting medical outcomes is like reporting adverse events. It provides critical information so that future patients (and the taxpayers) may be spared paying for inappropriate, expensive care down the road. In this case, CMS is on the right track.