Medicare ICD Guidelines Exegesis

In the process of preparing a module for my EP Mobile program on the Centers for Medicare and Medicaid Services (CMS) guidelines for using implantable cardiac defibrillators (ICD), I have had to delve deeply into the source document for said guidelines, the Medicare National Coverage Determination (NCD) Manual, more specifically, Chapter 1, Part 1,  Section 20.4, Rev. 129, found here.  The purpose of the module is to allow the user to input a few important clinical facts about a patient, e.g. ejection fraction, heart failure class, history of myocardial infarction, etc., and come up with an answer as to whether CMS will cheerfully pay for the ICD implant and leave you in peace or will, after a few years, come back at you with a Department of Justice investigation, accuse you of criminal fraud, and haul you away to the slammer.  Actually, there are indications that the DOJ attorneys are not approaching this investigation in a ham-handed way.  Apparently they are coming to understand some of the nuances of actual day-to-day medical decision making and realizing that the all-hallowed, set-in-stone, “guidelines” may not indeed cover all cases in which ICD implantation is indicated.  In fact, strict adherence to the guidelines could actually increase costs and risks to patients, such as the now oft-cited situation of a patient within 40 days of an myocardial infarction (MI) who would qualify for an ICD after the 40 days, but needs a device now because of bradycardia.  It would make no sense to implant a pacemaker and then a few weeks later remove it and replace it with an ICD.  It seems that even our governing overlords are not dense enough to sanction something like that.

There are 9 guidelines in all.  Going back from the 9th to the first guideline is like going back in time, from 2005 (when the last guideline was added) to 1991.  As you travel back in time, the guidelines get simpler, more succinct, and less specific.  The multiple exclusions applied to the later guidelines disappear.  Nostalgic memories of a simpler, less complicated time surface when we read the original ICD guideline:

1. Documented episode of cardiac arrest due to ventricular fibrillation (VF), not due to a transient or reversible cause (effective July 1, 1991).

Nothing could be simpler or more straight-forward than that!  Things already start to get a little more complicated with the second and third guidelines, introduced in 1999:

2. Documented sustained ventricular tachyarrhythmia (VT), either spontaneous or induced by an electrophysiology (EP) study, not associated with an acute myocardial
infarction (MI) and not due to a transient or reversible cause (effective July 1, 1999).
3. Documented familial or inherited conditions with a high risk of life-threatening VT, such as long QT syndrome or hypertrophic cardiomyopathy (effective July 1, 1999).


Guideline #2 is particularly thorny.  What is the definition of sustained?  A minute? 30 sec? 15 sec? Requiring cardioversion?  Then there is the use of the abbreviation “VT.”  Compare with the use of “VF” for ventricular fibrillation in guideline #1.  VT is the common abbreviation for ventricular tachycardia, yet the holy text uses it as an abbreviation for “ventricular tachyarrhythmia.”  I would take this to mean either ventricular tachycardia or ventricular fibrillation, but, like Biblical apologists, I have to be careful not to be guilty of eisegesis (reading too much into the text) as opposed to exegesis (an objective literal interpretation).  More importantly, in the case of induced “VT,” why was the EP study done in the first place?  Can someone with nonsustained VT occurring after an MI (but not within the acute phase, whatever that is defined as — ? 48 hours) have an EP study to see if they have inducible “sustained VT” and then get an ICD.  Note that this Guideline #2 seems not to fall under the 40 day post MI No-ICD-Zone that the later guidelines have.  The answer to these questions is: Who knows? And the 3rd guideline also seems rather vague.  I actually like the vagueness.  It seems to put the physician back in the driver’s seat.  Maybe we as physicians can decide which patients with long QT syndrome and hypertrophic cardiomyopathy can get ICDs, based on our judgment.  The guideline seems to leave open the possibility of other conditions being covered, such as Brugada Syndrome and Arrhythmogenic Right Ventricular Cardiomyopathy/Dysplasia.  Overall, the first three guidelines, dating from the 1990s, are vague and broad, with some wiggle room.  Good for physicians and patients, but bad for CMS presumably.  Subsequent guidelines were to be much more specific and rigid.

The remaining guidelines are based on the major primary sudden death prevention trials: MADIT-II and SCD-Heft.  These trials, like all clinical trials, have inclusion and exclusion criteria that were primarily designed to increase the power of the trials: i.e. to make sure the primary endpoints are met with the fewest patients and lowest costs.  The best way to do this is enroll patients at  highest risk who are clinically in a “steady-state.”  This means excluding recent MI, recent stenting or bypass surgery, patients with class IV (very severe) heart failure, and so forth.  CMS decided to take a very literal interpretation of these studies, assuming the results only applied to the exact populations studied, and basically cut and pasted the exact same inclusion and exclusion criteria used in the studies into their guidelines.  Thus the 40 day post-MI and 90 day post-stent or CABG exclusions.  Very concrete thinking on the part of the CMS folks, and, as is always the case, an approach that leads to less spending.  I don’t disagree with some waiting period after interventions or infarction, but one must understand that these waiting periods perhaps by intent save the government money either because the patient’s ejection fraction (EF) improves and he or she no longer needs an ICD (a good thing!) or the patient dies and no longer needs an ICD (oops!).

Here’s Guideline #4:

4. Coronary artery disease with a documented prior MI, a measured left ventricular ejection fraction (LVEF) ≤ 0.35, and inducible, sustained VT or VF at EP study. (The MI
must have occurred more than 40 days prior to defibrillator insertion. The EP test must be performed more than 4 weeks after the qualifying MI.)

This is the guideline based on the original MADIT study.  Somewhere along the line the need for showing that induced VT be non-suppressible by procainamide (part of the the MADIT study) seems to have disappeared — I could swear it was in there at one time.  It seems that hardly anyone does EP studies post-MI anymore, mostly because most of these patients are covered by the later guidelines based on SCD-Heft.  The only difference is that SCD-Heft requires Class II or III heart failure symptoms.  Criterion #4 could be applied to a SCD-Heft type patient with Class I heart failure and EF between 30 and 35%,, though at least in my experience it is not a common practice.  Note also that this guideline seems to conflict somewhat with guideline #2.

Guideline #5:

5. Documented prior MI and a measured LVEF ≤0.30 and a QRS duration of >120 milliseconds (the QRS restriction does not apply to services performed on or after
January 27, 2005).

This guideline still leaves in the controversial QRS > 120 msec criterion based on a overly-strict analysis of MADIT-II.  Patients with wider QRS complexes had more benefit from ICDs.  CMS later reversed this unpopular decision.  It is somewhat amusing that CMS does not make its rulings retroactive.  Presumably if you implanted an ICD in someone with a narrow QRS on January 26, 2005 you were committing Medicare fraud whereas if you had  just waited one more day everything would have been legal.

This is the first guideline encumbered with the Holy Exclusions, which can be summarized as:

Thou shalt not have Class IV heart failure

Thou shalt not have cardiogenic shock

Thou shalt not have had a CABG or PTCA within 3 months

Thou shalt not have had an MI within 40 days

Thou shalt not need coronary revascularization

Thou shalt not have a life expectancy of less than one year

Guidelines 6 and 7 are from SCD-Heft.

6. Patients with ischemic dilated cardiomyopathy (IDCM), documented prior MI, NYHA Class II and III heart failure, and measured LVEF ≤ 35%
7. Patients with non-ischemic dilated cardiomyopathy (NIDCM) > 9 months, NYHA Class II and III heart failure, and measured LVEF ≤ 35%

Of interest, the time limit of 9 months (later shortened to 3 months in Guideline #9) only applies to non-ischemic cardiomyopathy.  Although it is frequently stated (and certainly is good practice) that patients should be on optimal medical therapy before getting an ICD (i.e. beta blockers, ACE inhibitors), this is not actually stated in the guidelines.  Thus patients with ischemic cardiomyopathy do not have to wait to see if it gets better (as long as they don’t fall under the Holy Exclusions mentioned above) in order to get an ICD.  I find that this is a frequent misconception (I held it myself for some time).

Guideline 8 is the only one that addresses cardiac resynchronization therapy (CRT):

8. Patients who meet all current Centers for Medicare & Medicaid Services (CMS) coverage requirements for a cardiac resynchronization therapy (CRT) device and have
NYHA Class IV heart failure

The ironic thing about this guideline is that CMS actually does not have a national coverage document (NCD) for CRT, CRT guidelines are decided state by state.  After reading this many times, I’m still not sure what it means.  Surely it doesn’t mean that any one with Class IV heart failure with a wide QRS can get an ICD with CRT?  What about the various waiting periods (40 days post-MI, 3 months post-revascularization, 3 months after diagnosis of nonischemic cardiomyopathy)?  Do they apply?  Most commentators I have read feel this guideline essentially extends the other guidelines that exclude patients with Class IV heart failure to include those patients, as long as they meet CRT criteria (essentially have a widened QRS), but I have to say this one wins the prize for vagueness.  The Department of Justice could have a field day with this one.

Finally Guideline 9 is the same as 7, just shortening the waiting period for patients with non-ischemic cardiomyopathy from 9 months to 3 months.  This guideline also adds another Holy Exclusion: patients with irreversible brain damage (seems rather obvious).

The above guidelines have not been updated since 2005.  They are the Nine Commandments, written, as it were, in stone.  Ambiguous, vague, self-contradictory they may be, but, ignore them at your peril!

By mannd

I am a retired cardiac electrophysiologist who has worked both in private practice in Louisville, Kentucky and as a Professor of Medicine at the University of Colorado in Denver. I am interested not only in medicine, but also in computer programming, music, science fiction, fantasy, 30s pulp literature, and a whole lot more.

One reply on “Medicare ICD Guidelines Exegesis”

This is really excellent! Have you updated your thoughts, especially with regard to the value of a wearable external defibrillator as a bridge to permanent ICDs ?

Leave a Reply

This site uses Akismet to reduce spam. Learn how your comment data is processed.