Implantable Defibrillators

I am writing to comment on the National Coverage Decision Memorandum on Implantable Defibrillators posted by the Centers for Medicare and Medicaid Services (CMS) on December 28, 2004.

Over the past 6 months Cambridge Heart has strongly supported the clinical review of expanded ICD coverage and the evidence based concept of a data registry. We appreciate you and your staff taking the time to meet with me and members of the medical community to discuss the merits of utilizing Microvolt T-Wave Altemans (MTWA) as a risk stratifier in the MADIT II and SCD-HeFT populations.

In the time remaining before publication of the SCD-HeFT peer-reviewed article, I wanted you to know we were pleased the Decision Memorandum referenced the numerous comment letters submitted in support of including MTWA as a registry data element. However, in addition to the Decision Memorandum, we reviewed the "ICD Registry Data Elements" and the "ICD Registry Hypotheses". The “Data Elements” document lists several different criteria under History and Clinical Characteristics but it does not include MTWA.

The recent DINAMIT study in the New England Journal of Medicine highlights the need for effective risk stratification in guiding ICD therapy. That is why numerous acknowledged cardiovascular experts as well as community practitioners emphasized so strongly to CMS how MTWA can further delineate individuals who are unlikely to benefit from ICD implantation. Based on the comments summarized in the Decision Memorandum, it would make clinical sense to simply include MTWA in the registry data elements following the bullet point for "QRS Interval."

This would also be entirely consistent with the draft decision memorandum where CMS specifically identified several risk factors, including MTWA. As you recall, the draft decision stated:

Of particular interest would be hypotheses that identify appropriate risk factors, such as age, NYHA Class, QRS (or other ECG) interval and L VEF and screening tests, such as T-wave alternans, that predict which patients will receive defibrillator therapy. Finally, appropriate questions might identify those provider and facility attributes that predict good patient outcomes.

The data registry/data collection activity may present the best possible alternative to assist the medical community in refining the guidelines, as undertaking a classic randomized study in this population may no longer be practical or ethical. But without a registry data element reflecting MTWA, CMS will be less able either refine the population who benefits from ICD implantation or identify which patients are at very low risk and therefore do not require defibrillator therapy.

Finally, you are aware that some commenters raised concerns about the registry's cost, complexity or its potential to delay coverage and limit patient access. In contrast, I believe that the proposed registry data elements could be inexpensively and easily incorporated into existing ICD manufacturer product/warranty registration forms. Currently, most ICD manufacturers collect a substantial amount of information regarding each ICD placed. The forms are electronic, but can be in paper form, are frequently filled out by manufacturer employees, and are electronically distributed to the manufacturer, physician, patient, hospital and other entities.

Using a standardized format, each manufacturer could continue to receive their own warranty validation. The relevant database fields could be transmitted to a reputable, third party data analysis firm which could compile the necessary registry information while protecting proprietary information, such as the number and type of each manufacturer's devices.

Cambridge Heart is strongly supportive of CMS' approach and welcomes the opportunity to work with the agency and other stakeholders on the completion of the ICD National Coverage Decision and the rapid implementation of the National ICD Registry.

Less

I am would like to provide what I hope will be helpful comments on the recently proposed ICD Registry Hypotheses and Data Elements. I am a clinical cardiac electrophysiologist and Professor of Medicine and Biomedical Engineering at Case Western Reserve University, Cleveland. I have been directing clinical and basic science research in the field of sudden cardiac death for more than 12 years. My research has been supported largely by the NIH, but I have also collaborated closely with industry. Specifically, my work has focused on arrhythmia mechanisms and developing novel means for identifying patients at risk for sudden cardiac death (SCD). I am currently the principal investigator for the Alternans Before Cardioverter (ABCD) Trial which is sponsored by St. Jude Medical. This trial is designed to test the hypothesis that noninvasive T wave alternans (TWA) testing can be used to guide implantation of ICDs in patients without previous life-threatening cardiac arrhythmias.

I would like to applaud CMS’s efforts to establish a patient registry to obtain much needed information regarding optimal selection of ICD patients. The original CMS statement emphasized, quite appropriately, the importance of improving methods of identifying those patients most likely to benefit from therapy. Despite the important and significant findings of the Madit2 and ScdHeft Trials, these trials also emphasize important limitations associated with using ejection fraction alone to identify “high risk” patients. The absolute benefit of ICDs to mortality was small (about 6%), there was no benefit at all in patients under 65 years old (Madit2), and 3 year survival following appropriate ICD therapy was only 30% (Madit 2). Clearly, delineating which patients are most likely to develop arrhythmias rather than progressive mechanical heart failure is critical to appropriate utilization of ICDs. Therefore, it is not surprising that ejection fraction, a measure of mechanical cardiac dysfunction, fails to optimally identify patients likely to benefit from ICDs (e.g. in Madit2 17 ICDs had to be implanted to save 1 life). Finally, the benefit of ICDs must also be weighed against the morbidity and cost of therapy. Consequently, it is important to consider that in SCDHeFT it is estimated that 80% of ICD shocks were inappropriate (i.e. not for life-threatening arrhythmias), and Madit2 showed an unexpected rate of heart failure admissions in patients treated with ICDs.

Our work and the work of many others have established that TWA is mechanistically linked to cardiac arrhythmogenesis, and that it is the most sensitive noninvasive marker for sudden death risk currently available. This is highly relevant to the difficult decisions CMS is presently confronting because the compelling question raised by recent primary prevention trials is not “who needs an ICD”, but rather “who does not”? Consequently, a noninvasive test like TWA, with negative predictive value in the order of 95%, can be potentially very helpful in screening for patients who will not benefit from ICDs. I believe it is most compelling that 3 recent trials in patients with clinical characteristics similar to those in Madit2 and ScdHeft demonstrated that event rate in TWA negative patients was lower than the event rate in similar patients in Madit2 treated with ICDs.

I was surprised to learn that the proposed ICD registry hypothesis and data registry elements do not include parameters of risk stratification. I would strongly suggest that you consider collecting data on risk stratifiers, such as T wave alternans. The registry will provide a unique opportunity to collect critically important data on the future usefulness of risk stratifiers in this population. Such information may provide critical insights to CMS, practitioners, and the scientific community for developing new strategies for identifying those patients most likely to benefit from ICDs, and avoiding unnecessary morbidity and waist of precious health care resources associated with inappropriate use of ICD technology.

Less

Please consider approval of oupatient (in addition to inpatient) ICD implantation for primary prevention indications. Many doctors routinely implant ICDs on an outpatient basis on non-medicare patients and discharge the same day. Even Dick Cheney had his EP study and ICD and discharge on the same day. Unfortunately, hospitals adminstrators tell me that in order to get Medicare payment for defibrillator implantation that overnight inpatient hospitalization is needed since outpatient implantation is not covered. This practice obviously increases the cost of care.

Less