Wednesday, September 03, 2008

ICD Shocks - A Blessing or a Curse?

It's all across the media: ICD shocks predict impending death!

Thanks. My patients appreciated these headlines, I'm sure.

But if we're going to tell the 234,000 people who have implantable cardiac defirillators (ICDs) that they're going to die in less than a year after their first shock, shouldn't the media attempt to exercise even a modicum of discernment before going public with these headlines?

First, these findings come from analysis of the Sudden Death in Heart Failure trial (SCD-HeFT) - a well-designed prospective multicenter randomized trial in adults over age 18 which randomized over two thousand patients between three therapeutic treatments for heart failure: (1) conventional medical therapy, (2) converntional medical therapy plus Amiodarone, and (3) conventional medical therapy plus a single chamber, shock-only ICD therapy. The main finding of the SCD-HeFT Trial, in its original form, was that medical therapy (be it with o without Amiodarone), mortality in this group of patients was 7.2% annually. In other words, our best antiarrhythmic at the time, Amiodarone, failed to effect patient mortality. The addition of an ICD for primary prevention of sudden death is this sick population, actually reduced mortality (to 5.5% per year in patients followed for 5 years - a 23% reduction in mortality compared to medical therapies).

So now, is it any wonder that patients who have received an ICD to prevent death on a preventative bases, might just get a shock as their condition deteriorates?

When you can take fairly sick individuals and implant an ICD in them as primary prevention (i.e., they never had an arrhythmia before) and then look at those who develop arrhythmias suddenly, it seems intuitive to me that those patients would be sicker or have had a change in their clinical situation.

The data from the study support this theory. Therefore, evaluating for the development of worsened heart failure or additional ischemic burden might be prudent in these patients, since these causes seemed to be the largest culprits resulting in ICD shocks. Additionally, the study found that patients with either appropriate (shocks for ventricular arrhythmias) or "inappropriate" (shocks for rapid heart rates from non-ventricular causes, like atrial fibrillation onset) have a poorer prognosis compared to those who did not receive shocks, but this also might be a way to risk-stratify the sicker patients of the overall implant population. Does this mean the patients should not receive devices to live longer? Of course not. But there has to be balance to recommending ICDs in the sickest patients, since recurrent shocks can impact the patient's quality of life if they occur frequently. Infrequent shocks, interestingly, did not seem to convey a worse quality of life in this same group as reported (ironically) in the same issue of the New England Journal of Medicine.

Remember that all mortality curves of different therapies that divert favorably from one therapy, will again meet with the alternate therapy some time in the future, since none of us are immortal. ICD's reduce mortality in sick patients with heart failure, but an ICD shock might suggest that, of the group implanted, those patients with shocks have a better indication for the device, since their mortality is higher. It is important, however, to keep in mind who was NOT studied in this paper:

The patient population studied did not include children.

The patients studied did not have cardiac resynchronization devices.

The devices were not programmed for antitachycardia pacing and did not have dual chamber devices which are better capable of detecting atrial arrhythmias.

So before our patients go out to buy funeral plots, let's keep the issues of the risks and benefits of ICD therapy, and the benefit or curse of their shocks, in perspective.


Reference: Poole, JE et al, "Prognostic Importance of Defibrillator Shocks in Patients with Heart Failure," New England Journal of Medicine Sep 4, 2008 Vol 359(10):1009-1017.


Anonymous said...


What are the latest mortality/survival rates in patients with IHD and VT/VF with an EF in the 40s after dual chamber ICD placed (after aborted arrest)? Thanks.


DrWes said...

JC -

The data in the NEJM article to not apply to the patient you describe, since the NEJM article only pertains to heart failure patients implanted with an ICD for primary prevention.

The best summation of data close to the patient you describe implanted for "secondary prevention" (aborted sudden death with ischemic heart disease) would be found here. In this group, the prevention of arrhythmic death is especially strong, but the overall mortality data still are tough to estimate, since only a "relative risk" reduction in mortality of 28% is described.

In the AVID Trial which compared ICD therapy in patients with EF's <40% to Amiodarone drug therapy, a good review article states "the two-year mortality in the drug-therapy group was 25 percent. Therefore, assuming that the drugs produced neither benefit nor harm, the 27 percent reduction in relative risk corresponds to a 7 percent absolute reduction in risk among the total population of patients who would receive implantable cardioverter–defibrillators according to the AVID criteria. Although 7 percent seems small, as compared with the effects of other types of preventive therapy, it is actually a good rate of therapeutic efficiency. Of somewhat greater concern is the magnitude of the extension of life, an important measure of estimated benefit among patients with chronic and progressive diseases. In the AVID Trial, the mean additional survival provided by the defibrillator at 3 years of follow-up was only 3.2 months. However, since the outcome curves were not yet converging, continuing follow-up may (have)demonstrate(d) longer extension of life."

Hope that helps a bit...

Anonymous said...

Yes. thanks.

Anonymous said...

The finding that ICD shocks predicted patients who will die soon inspired me to conduct research into other so-called "protective" devices to evaluate whether they really do save lives, or conversely whether their use might actually be associated with INCREASED death. I was shocked (no pun intended) to find that a similar relationhip between protective devices and increased death is actually common! In fact, my research showed that deployment of airbags in personal vehicles is strongly associated with traumatic death, ofthen within minutes of being deployed. I wonder why the media hasn't picked up on this association to demand the immediate recall of airbags?!