ICDs in the Elderly With VT/VF: Be a Doctor Not an Implanter

John Mandrola, MD


February 13, 2017

There are times when neither a randomized trial nor a guideline statement assist a medical decision.

Use of an implantable cardioverter defibrillator (ICD) in an elderly patient who presents with sustained ventricular arrhythmia is one of those times.

The core problem in this scenario is that sudden death due to ventricular arrhythmia is only one of many possible adverse outcomes. The word morbidity hardly does justice to many of the bad things outside of dying that can happen to an older person: repeated hospital admissions, loss of independence, admission to a nursing home, and a slow death from diseases such as dementia, frailty, and cancer.

This is why I was drawn to a recent study on outcomes among older patients receiving an ICD for secondary prevention.[1]

Researchers used data from the National Cardiovascular Data Registry (NCDR) to study 12,420 Medicare patients undergoing first-time ICD implants for secondary prevention. They assessed mortality, rates of hospitalization, and admission to a skilled nursing facility over 2-year follow-up for multiple age strata.

Three key findings emerged from this study:

  • First was that a lot of very old patients get ICDs for secondary prevention: one in four patients in this analysis were over the age of 80.

  • Second, survival was pretty good. Nearly 80% of patients were alive at the end of the 2-year follow-up. As expected, mortality rates increased with age at implant.

  • Third, this impressive survival came at a cost. Almost two-thirds of these patients were hospitalized during the follow-up period and admission to a skilled nursing facility occurred in one in 10 patients overall and one in three patients over age 80.

These data are from 2006–2009. People are living longer with heart disease, so look for the number of 80- and 90-year-olds with ventricular tachycardia (VT) and ventricular fibrillation (VF) to increase.


Let's first address the main weakness of this paper: it's not a randomized trial. In an email, Prof Rachel Lampert (Yale University), who was not involved in the study, wrote, "Without a no-ICD group, nothing whatsoever can be said about [device] benefit. . . . An 83-year-old in your office doesn't get to choose whether they want to be 83 or 60. What they need to know is . . . will they live longer with an ICD?"

Her comment gets to the point that age is only a number. We have all seen vigorous 80-year-olds with VT who have better survival than a sedentary 50-year-old smoker who presents with VF from severe ischemic cardiomyopathy.

That said, this study and others argue for a cautious approach to the use of ICDs in the elderly—even for secondary prevention.

Consider first the small magnitude of ICD benefit for secondary prevention. A patient-level meta-analysis of the three ICD-vs-antiarrhythmic-drug trials in survivors of VT and VF (AVID)[2], CASH[3] and CIDS[4]) reported a 28% relative reduction in total mortality.[5] This statistically significant reduction translated to an average prolongation of life of only 4.4 months over 6 years. And these modest gains were noted in much younger patients—the mean patient age was 63 in the pooled cohort.

The second reason for pause is that it's possible ICDs provide no benefit in older adults. Another meta-analysis of the same trials by Healey et al focused on older people.[6] In patients ≥75 years, ICD implantation was not associated with a reduction in all-cause mortality (HR 1.06, 95% CI 0.69–1.64) or arrhythmic death (HR 0.90, 95% CI 0.42–1.95). It's important to note that only 252 patients (15%) in the trials were 75 years or older and that the interaction between older age and ICD use did not meet statistical significance. Those limitations prevent drawing any firm conclusions, but the findings surely give reason for caution.

Another factor that can affect the ICD decision is ventricular function. About 40% of the NCDR patients had a left ventricular ejection fraction (LVEF) >35%. There's good reason to believe this group gains little from an ICD. The original patient-level meta-analysis of the three trials found that the survival benefit for ICDs was largely in patients with LVEF at or below 35%.[5] A subgroup analysis of AVID reported no difference in survival between drug- and ICD-treated patients when the LVEF was 35% or higher.[7] These trends are intuitive; better ventricular function plausibly confers a lower risk of death from cardiac causes, and anything that lowers cardiac mortality relative to all-cause mortality lessens the likelihood of ICD benefit.

The issue of arrhythmic death vs overall death is especially important in the elderly. I recently cared for an older patient who, 6 months after her husband of 50 years died, was "saved" by an ICD shock. I put saved in quotes because her last year of life included a cycle of multiple admissions for broken bones, infections, and heart failure. She died alone in a nursing home. Indeed, the ICD extended this person's life, but, to me, the gain looked nothing like a win.

That's what this registry study helps us see. Older people rarely get ventricular arrhythmia as a fluke. They get ventricular arrhythmias because of organ decline—which is the normal order of life. And the heart isn't the only organ in decline. It's no surprise, therefore, that admissions to the hospital and nursing homes after ICD implants in the elderly are common.

Finally, the NCDR registry study highlights the vital role of the patient's viewpoint. Some elders may want to eke out every day; extra time is their goal. Others value their independence and want to avoid admission to a skilled nursing facility. Death for some is not the worst outcome. Algorithms and checklists declaring "ICD eligibility" offer the older person with ventricular arrhythmias nothing useful.

Our older patients deserve an honest framing of the expectations of life with or without an ICD. To do this, we have to see past the ECG and echocardiogram to the person. Such vision requires a clear-eyed view of the evidence—and the limits of human life-spans.

The older our patients are, the more they need a doctor—not an implanter.


Comments on Medscape are moderated and should be professional in tone and on topic. You must declare any conflicts of interest related to your comments and responses. Please see our Commenting Guide for further information. We reserve the right to remove posts at our sole discretion.