Appropriate Use of Defibrillators

There have been numerous news reports over the last week regarding defibrillators.  A paper, published in last week’s Journal of the American Medical Association (JAMA), suggests that one-fifth of all patients who received a defibrillator in the last few years in the United States didn’t actually need one and may have been exposed to unnecessary harm and expense.   One of the many* reasons I started this blog was to have a venue to be able to elaborate on stories such as this one and I think this issue deserves some clarification.

First, let me set the stage with a little background.  The implantable cardioverter-defibrillator (ICD) is a device similar to a pacemaker that can deliver a heavy electrical shock to restore normal heart rhythm in patiens in the throes of cardiac arrest.  This technology has been around for a few years but only recently have we had a relatively clear understanding about who should get one.

As cardiologists the best way for us to determine who would most benefit from an ICD is to lay our hands on a crystal ball and have it forecast which of our patients will suffer cardiac arrest.  We simply implant an ICD in each of those and just wait for it to save our patient’s life.  Unfortunately, crystal balls have yet to receive an FDA indication and we are thus left to grope for different methods.  Numerous research studies have been done to help us better predict who might be at highest risk for sudden cardiac death and over the years we’ve managed to piece together a somewhat reliable and reproducible method.

It all comes down to the strength of the heart – what we call the ejection fraction (EF; please see a previous post for a long and boring discussion of this parameter).  In general, the lower the EF, the higher the risk of dropping over dead.  A normal EF is above 50%, and with anything below about 35% the risk begins to rise to the point where it’s probably worthwhile to implant an ICD.  I have many patients with EF values between 10 and 20%, a group that is clearly at highest risk and which has most to gain by receiving an ICD.

There are several caveats to the EF rule.  Some patients can expect to have their EF improve with appropriate therapy and should therefore not receive an ICD until they’ve been given a sufficient period of time.  One example of this is the patient who suffers a heart attack and undergoes some type of procedure to improve blood supply to the heart (stent placement or bypass surgery).  In such a patient we’d expect to see the muscle strength improve substantially over the course of weeks to months.

This observation—a heart weakened by a recent heart attack can improve in just a few weeks—has led to the recommendation that an ICD be deferred for at least a few months.  Our current guidelines mandate a particular period of time before reassessing the EF to see if the heart has strengthened with the idea that the patient may no longer be at risk for cardiac arrest.

A second caveat is that some types of heart problems can improve with medication alone and in these patients (with so-called nonischemic or dilated or idiopathic cardiomyopathy) a waiting period of 3 months is recommended prior to concluding that the EF is not going to get any better.

Several years ago, after our current guidelines were codified and put into effect, the Centers for Medicare and Medicaid Services (CMS) began collecting information on each patient who undergoes placement of an ICD.  It is this collection of data that the authors of the JAMA article mined to come up with the results that have made such a splash in the news.

The results were interesting.  It appears that 22.5% of ICDs implanted between 2006 and 2009 did not fulfill the criteria for appropriateness—suggesting that more than one-fifth of ICD surgeries were done outside the accepted guidelines.

Some commentators have wondered if this is just another case of doctors consciously performing unnecessary surgeries in order to enhance their income.  If true it wouldn’t be the first time.   Some readers might recall the scandal that arose in Redding, California where two heart specialists became immensely wealthy by performing angiograms and bypass surgeries on scores of patients with no significant coronary artery disease.  Or the recent reports about the Baltimore cardiologist Dr. Mark Midei who was investigated for implanting an astonishing number of stents into patients who probably didn’t need them.  His record was 30 stents in one day, an accomplishment that executives at the stent manufacturer Abbott Labs lauded in internal emails.  After his hospital dumped Dr. Midei, Abbott promptly hired him as a consultant.  I would love to have been in on that job interview: “So, Dr. Midei, you’re under investigation by the Justice Department and your own hospital has tossed you out the front door, but—wow!—thirty stents in one day!  Welcome aboard.”

I’ve looked closely at the JAMA report and I don’t honestly think this is another case of doctors taking advantage of the system.  Most of the cases considered inappropriate were patients who received ICDs prior to the conclusion of the mandatory waiting period.  By and large these patients were quite sick and clearly had diseased hearts (the average EF was 25%)—nothing like the fraudulent cases cited above.  Rather than a bunch of money-hungry cardiologists looking to buff their procedure numbers this is probably simply a case of doctors trying to be as aggressive as possible with a disease that is known to carry a high mortality rate.

As a side note, I was able to gain access to Alegent’s contribution to the same database cited in the JAMA article.  I was pleased to find that our 5 metropolitan hospitals ranked among the highest in the country (above the 95th percentile) for implanting ICDs with strict adherence to the CMS guidelines.

When I see stories like this one in the news I’m always a little concerned about the effect they have on patients who would really benefit from the therapy being called into question.  As an example, in this instance we’ve had a couple of patients call our office expressing doubts about going forward with a planned ICD surgery.  Both of these individuals fit clearly into the guidelines and are currently at high risk of cardiac arrest without a defibrillator.

The irony about the flap over the JAMA article is that the greater problem isn’t even really addressed: too many patients with a need for an ICD aren’t getting one.  This paradox—that we’re putting too many in the wrong people while putting too few into the right people—is not lost on many of the experts asked to comment on this issue, such as Dr. William Abraham, director of the Division of Cardiovascular Medicine at Ohio State University, as cited on ABC news:

"Registry studies are notoriously difficult to draw conclusions from," he said. "The big challenge is that we get a snapshot of patients who received defibrillators, but we don't get to know anything in this study about patients who might need one but didn't have one implanted.

"We know that over half of individuals who should get an ICD implanted based on guidelines actually don't receive one, for a variety of reasons. This is a far bigger problem."

While this new data may not be a smoking gun that points toward cardiologists running amok, it does represent a reminder to us who implant ICDs that we need to always assess the appropriateness of any procedure we do.  Our specialty has invested years into scientific research to get us to where we are now, research that helps us decide who needs an ICD and who doesn’t.  The patient is best served when we pay attention to the guidelines that we’ve worked so hard to create.

*The other reasons: keeps me off the streets and out of a life of crime;  gives me a good excuse to tell my wife I need an iPad; and disproves my high school English teacher who told me I’d never be able to string together a coherent sentence.