September 8th, 2011

When Is the Patient “Too Old” for an ICD?


An 86-year-old man presents to you for the first time with lightheadedness and increasing dyspnea on exertion, which he reports experiencing for the past 4 months. He has a history of coronary artery disease requiring 3-vessel bypass, dyslipidemia, hypertension, and chronic lymphocytic leukemia (CLL) that has been stable for many years.

His vital signs upon presentation are as follows: blood pressure, 110/80 mm Hg; heart rate, 35 bpm; afebrile; oxygen saturation of 98% on room air. His physical exam is significant for a 2/6 holosystolic murmur at the apex. He has no evidence of jugular venous pressure elevation, third-heart sound, rales, hepatosplenomegaly, or lower-extremity edema. ECG reveals sinus bradycardia at 35 bpm, a first-degree atrioventricular block, and a left bundle-branch block with a QRS duration of 160 ms. Lab tests show a creatinine level of 1.5 mg/dL and hematocrit of 30%.

A transthoracic echocardiogram reveals an LV ejection fraction of 30%. Records from the patient’s previous cardiologist document an LVEF of 30% more than 1 year ago. The patient has been on a stable regimen of an ACE inhibitor, furosemide, aspirin, and a statin. He was taking a beta-blocker for many years until 2 weeks back, when his primary care physician took him off the medication “due to slow heart rate.” The patient denies a history of syncope or cardiac arrest.

The patient lives alone, but his daughter visits him weekly and prepares food that lasts the entire week. His mobility is limited; at most he walks to a store down the street. He and his family want “everything done.” When questioned about the patient’s life expectancy with CLL, his oncologist says, “Your guess is as good as mine.”



  1. Would you consider pacemaker/ICD therapy for this patient? If so, would you suggest cardiac resynchronization therapy (or CRT with defibrillator)?
  2. What aspects of the patient’s clinical profile would influence your decision? How would you approach the conversation about an implanted device?
  3. How important should age be in choosing an expensive therapy such as an ICD? Should the decision be left to the patient and his family or to the caring physician?
  4. If an ICD is chosen, would you discuss the option of turning it off with the patient and his family?


James Fang, MD
September 29, 2011

This elderly man has symptomatic systolic heart failure from chronotropic incompetence and electromechanical dyssynchrony (e.g., wide left bundle-branch block) despite medical therapy. Pacing and cardiac resynchronization therapy (CRT) should be offered in this situation, given that there is reasonable evidence (although extrapolated from randomized trials involving patients younger than 70) to suggest biventricular pacing would enhance his quality of life by improving both the chronotropic incompetence and the dyssynchrony. Although the patient has chronic lymphocytic leukemia, it appears to be clinically stable (indeed, stable CLL can persist for years).

However, the evidence is modest that octogenarians, specifically, derive a clinically relevant improvement in overall survival with ICD therapy. The patient should be given the alternative of CRT without ICD, although that option is rarely offered in the U.S. The patient and his family need to know that concomitant ICD therapy will not improve his quality of life and could potentially make it worse, as the chance of receiving inappropriate shocks (e.g., from rapid atrial arrhythmias) is increased.

In ICD trials, the survival curves took at least one year to separate, and it is not known when (or if) this separation occurs in octogenarians. Notably, subsequent survival following an appropriate ICD shock is related to heart failure, which will not improve with ICD therapy alone. Thus, patients should be counseled that ICD therapy will simply change the mode of their ultimate demise from sudden death (which many people prefer) to an insidious decline in health, functional capacity, and loss of independence that accompany chronic heart failure.

Finally, patients and families should clearly understand that “turning off” the ICD is not tantamount to “turning the patient off.” Rather, it is an inactivation of an electrical discharge that would interfere with the peace and dignity of end of life when it inevitably comes.


Tariq Ahmad, MD, MPH
October 10, 2011

After long discussions among the patient, his family, and the cardiologists, spanning several outpatient clinic visits, a decision was made to implant a CRT device with ICD capabilities. Although it was believed that the patient would be likely to benefit from pacing and cardiac resynchronization, the decision regarding an ICD was more difficult. After input from the oncologist, the patient’s life expectancy was predicted to be at least 1 year. The patient and his family were told about the risks versus the benefits of an ICD, and they elected in favor of it. They were also heavily influenced by input from a family friend whose ICD had apparently “saved his life many times.”

The patient was brought in as an outpatient, the device was implanted without complications, and he was discharged the next day. He was seen in the EP device clinic 3 months after discharge and, according to the clinic note, appeared to “have more energy.” He was also seen in the heart-failure clinic, where he reiterated his improved HF symptoms. However, objective assessments of his functionality, such as peak VO2 testing, have not been performed. Also, discussions about turning the ICD off under certain clinical scenarios have not been broached.

9 Responses to “When Is the Patient “Too Old” for an ICD?”

  1. Shaumik Adhya, MBBS BSc MRCP CCDS says:

    1. Yes – I’d suggest CRT-P. He needs pacing, and I wouldn’t feel comfortable giving him a dual chamber device even programmed to AAI-DDD mode.

    2. Difficult – if I thought he’d understand the issues, I’d explain that he fulfils guidelines for an ICD to prevent sudden cardiac death, but that sudden cardiac death is not a bad mode of death, hence suggest CRT-P. Whilst there isn’t any good direct evidence comparing CRT-P to CRT-D the incremental benefit is likely to be small. Also there’s the possibility that CRT-P would improve his LV function to the point where he wouldn’t qualify for an ICD.

    3. Should be taken individually.

    4. Yes

    We recently had a somewhat similar case with a 90+ year old who’d experienced symptomatic, but not syncopal VT. Over a number of days we discussed the issue with him and his family. He wanted the ICD. He was living independently at home and definitely knew what he wanted!

    Competing interests pertaining specifically to this post, comment, or both:

  2. I agree entirely with Shaumik and think a CRT-P would be a good choice. I think we all have difficult decisions over these cases. It would be good to see some registry data looking at outcome and age. I would not be surprised if there were not an age at which due to limited survival ICDs are not really cost effective any more (at least according to NICE criteria).

    One thing I do find is that many elderly patients are much more accepting of death than many of us are.

    I try to discuss the turning off of ICDs when I consent patients for such procedures.

    Competing interests pertaining specifically to this post, comment, or both:
    I have received educational sponsorship from Boston Scientific and Medtronic and speaker fees from Medtronic.

  3. David Powell , MD, FACC says:

    I also tend to agree with Shaumik. This patient likely met current revised CRT-D criteria one year ago, with EF 30, LBBB, and probable CHF class 2 presumably on standard good medical therapy (to which I would add spironolactone). Let us assume that a discussion took place at that time and a” conservative” course was chosen. Now more symptoms mandate relief of bradycardia. To be entirely theoretically consistent, a RV pacemaker may restore the patient’s prior symptomatic status and would be the least invasive means to do so. Only for persistent symptoms after a RV PM would an” upgrade” to CRT be considered under these strict “minimize invasive procedures only for current symptom relief” strategy. RV pacing should not be more deletirious compared with the previous LBBB ( echo studies suggest less dyssynchrony for RV pacing compared with most native LBBBs, particularly thoae with long QRS). RV pacing from high in the septum is a reasonable approach. But practicably speaking, how much more” invasiveness” is a biventricular pacemaker compared with an RV pacemaker? The complications are more, but how much more? I would urge the patient/family to make a slight departure from the original plan in order to improve 5 year prognosis. The ICD issue should also again be addressed, as this “must insert at least. one wire” situation may change the patient/family outlook.

    Biological age with comorbidities and evaluation of patient’s preferences all go into the complex decision. Cost is not a direct factor for a recommendation, but is highly relevant indirectly . Reimbursement for example is not a current barrier in US for this case (whether age alone should dictate reimbursement is a societal issue which we should play a major role in).

  4. Leonard S. Sommer, B.S., Yale; M.D., Columbia says:

    I would be interested in the result of ambulatory ECG monitoring to assess the patient’s response to the mild level of exertion he performs, in deciding on intervention.

    Competing interests pertaining specifically to this post, comment, or both:
    None. /LSS

  5. Mani Prasad Gautam, MD says:

    Occasionally it seems difficult while counseling regarding the need of pacing in elderly people. This is especially true in out set up where there is no proper health insurance system. Recently I implanted a PPI in a 97 years female who had presented with history of dizziness for last 2 months and baseline ECG showed complete heart block with ventricular rate of 28/min. Her echo showed LV diastolic dysfunction, rest of the study was normal, she is doing very good.
    Regarding this case I would have preferred CRT with ICD combo device, Rest of the issues I agree with Saumik.

    Competing interests pertaining specifically to this post, comment, or both:

  6. VAZHA AGLADZE, MD., PhD says:

    I think, doctors must do what the patient’s family declares. They want to do everything for him, so let doctors do the same. According to every guideline, he needs CRT, only problem is his age. One contra, many pro.
    So, the device must be implanted.

    Competing interests pertaining specifically to this post, comment, or both:

    • Moshe Gunsburg, M.D. says:

      I agree that a patient’s and family’s wishes are extremely important in the decision making process but it is also important for a knowledgable and caring physician to be able to sort through the elements of the clinical situation at hand and then apply the scientific data (to date) regarding appropriate modalities of treatment so as to offer objective and truthful information and advice. The absolute question is: does the therapy help the patient, the family (psychologically) or both. The harder measure to discern is: if so, by how much? Then comes the risks vs benefits analyses. At the end of the day, if a therapy is not posing excessive risk and potential harm to a patient, I believe that no one would question the decision to proceed with it if the patient and/or family are the ones paying for it themselves. Most of the head-scratching torment comes from the reality that it is a third party (and in many instances “society”) who bears the financial burden of utilizing the therapy. At what point in the decision algorithm does society weigh in and have the right to say, OK, we understand that you “want this” but it’s something that you will have to pay for. I certainly do not have an answer but in my own experience I sometimes find myself biting my lower lip and performing a procedure on such a patient as the one presented here because it’s what the family wants (aka – no one wants to carry the mantle of guilt for allowing the medical commnity to withhold a “guideline supported therapy” for their loved one), even when my clinical judgment tells me that “less is more” in a particular scenario. Guidelines are developed and released to serve as a source of guidance, but not to mandate or to perfectly address every single clinical scenario. Hence, when a patient or family insist on getting a device and the guidelines support the indication, what choice does a physician have when living in a litiginous society?

      Competing interests pertaining specifically to this post, comment, or both:

  7. Leonard S. Sommer, B.S., Yale; M.D., Columbia says:

    With due respect to colleague Vazha Agladze, MD PhD, (second opinion above) both patient and family need guidance from the specialist-physician involved. Family’s motivation is based on emotion and fear. Intervention is not automatically advisable based on diagnostic and publication parameters. Each patient is an individual candidate for therapeutic guidance and decisions involving surgical intervention, however “low-risk”, need to be carefully individualized. “First, do no harm” is our historic guide.

    Competing interests pertaining specifically to this post, comment, or both:

  8. When my great grandmother was 90 she developed an acute gall bladder. Her doctors were hesitant to operate on someone of her years. Finally as the symptoms persisted, they operated and removed her gall bladder.

    She did much better for the next 22 years!

    It is hard to judge the quantity of a person’s life based upon the age of a birth certificate. I feel that to use age alone as a criteria for treatment is inappropriate.

    Competing interests pertaining specifically to this post, comment, or both:
    I hope to be old some day.