Vice President of Health Policy
Heart Rhythm Society
January 20, 2012
Centers for Medicare and Medicaid Services
Department of Health and Human Services
Room 445–G, Hubert H. Humphrey
200 Independence Avenue, SW
Washington, DC 20201
Re: Coverage with Evidence Development Public Solicitation
Dear Ms. Tavenner:
The Heart Rhythm Society (HRS) welcomes the opportunity to provide input to support CMS refinement of the Coverage with Evidence Development (CED) process. The Society is committed to accelerating access to safe, effective, reasonable, necessary and innovative technology, and minimizing the administrative burden associated with the collection of data to address critical clinical questions.
HRS is the international leader in science, education and advocacy for cardiac arrhythmia professionals and patients, and the primary information resource on heart rhythm disorders. Founded in 1979, HRS is the preeminent professional group representing more than 5,400 specialists in cardiac pacing and electrophysiology, consisting of physicians, scientists and their support personnel. HRS’s members perform electrophysiology studies and curative catheter ablations to diagnose, treat and prevent cardiac arrhythmias. Electrophysiologists also implant pacemakers, cardioverter defibrillators and cardiac resynchronization therapy devices in patients who are indicated for these life-saving devices.
Randomized clinical trials (RCTs) have contributed to the provision of evidence-based care for patients with heart rhythm disorders. However, clinicians and other healthcare stakeholders recognize that RCTs are most often designed to answer questions in a clearly delineated population. To maximize their utility, conclusions drawn from clinical trials should be applicable to real-world populations and unstudied patient subgroups.
Clinicians have the challenge of determining how closely the characteristics of an individual patient match those of the clinical trial population and estimating the likelihood of benefit and harm from a procedure, medication, or process of care. These challenges are particularly difficult when a patient differs in important ways from the study population.
The January 2005 Medicare national coverage determination that extended coverage for implantable cardioverter-defibrillators (ICDs) for primary prevention of sudden cardiac death to high-risk individuals used the CED mechanism to address the apparent evidence gaps following completion of RCTs. The ICD CED requires, as a condition of Medicare coverage, that a Medicare beneficiary receiving an ICD for primary prevention of sudden cardiac death be enrolled in a clinical trial or a qualified data collection system including an approved registry.
As outlined in the January 2005 coverage memorandum, the CED mechanism was designed to consider the following hypotheses:
1. The clinical characteristics of the patients receiving ICDs are similar to those of patients involved in the primary prevention randomized clinical trials.
2. The indications for ICD implantation in patients are similar to those in the primary prevention randomized clinical trials.
3. The in-hospital procedure related complications for patients are similar to those in the primary prevention randomized clinical trials.
4. Certified providers competent in ICD implantation are implanting ICD devices in patients.
5. Patients who receive an ICD represent patients for which current clinical guidelines and the evidence base recommend implantation.
6. The clinical characteristics and indications for ICD implantation do not differ significantly among facilities.
7. The clinical characteristics and indications for ICD implantation do not differ significantly among providers.
8. The in-hospital procedure related complications for ICD implantation do not differ significantly among facilities.
9. The in-hospital procedure related complications for ICD implantation do not differ significantly among providers.
10. The in-hospital procedure related complications for ICD implantation do not differ significantly among device manufacturer, types, and/or programming.
To test these hypotheses and meet the CED data collection requirements, the ICD RegistryTM was designed as a partnership of the Heart Rhythm Society and the American College of Cardiology Foundation.
In May 2007, CMS approved the design of a study to answer the questions posed in the CED. Unfortunately, funding could not be obtained until late 2009 for the Longitudinal ICD Study. As of November 2011, nearly 1500 of the planned 2500 recipients had been enrolled in the study. The investigators plan to assess the rates and correlates of mortality, hospitalization, device-related complications, and ICD therapies.
As of December 2011, 1653 hospitals are submitting data to the ICD registry. The ICD Registry holds over 780,000 records of which 54% are primary prevention device implants under the CED. The data collected in the ICD Registry have provided answers to ICD-related questions about differences in decisions to implant and outcomes related to gender, race and ethnicity, geographic region, appropriateness, complications, and device type including the number of leads. Differences in outcomes have been identified related to the professional qualifications of the implanting physician and the volume of implants performed at a facility.
The Heart Rhythm Society makes the following recommendations and provides lessons learned about the CED process:
1) The CED mechanism initiates a complex, multi-year enterprise, and should be considered when there are unanswered questions and when answers to those questions will lead to one of two actions: a) Medicare will cover the service or item and remove the CED requirement, or b) Medicare will withdraw coverage if the device, drug, or process of care is shown to not meet the statutory requirement that it is reasonable and necessary.
2) Careful analysis of existing datasets should be performed prior to a CED to ensure that the mechanism is only utilized when there are no alternative sources from which to obtain the answers. If the environmental scan for alternative data sources and/or the analysis of that data cannot be conducted prior to the publication of a coverage determination, it should be initiated as soon after the determination as possible. For example, in the case of the ICD NCD, in 2009, Drs. Gillian Sanders and Donald Berry both presented independent analyses using existing ICD clinical trial datasets to the Medicare Coverage Advisory Committee. The experts reported that many of the questions posed in the ICD coverage with evidence development determination could have been answered at the time of the CED’s implementation. http://www.cms.gov/medicare-coverage-database/details/medcac-meeting-details.aspx?MEDCACId=49&fromdb=true
3) CEDs are designed to answer specific questions. Caution should be exercised to ensure that conclusions are limited to the questions identified in the CED and that the CED is not utilized for additional research questions for which the data collection methods were not designed.
4) The funds necessary to perform required studies must be made available to ensure that the questions posed in the CED are answered in a timely manner.
5) The approval of the study design to answer the CED questions should occur concurrently with the CED’s implementation.
6) The design of patient registries required by the coverage with evidence development mechanism should closely follow the principles outlined in the Agency for Healthcare Research and Quality Effective Health Care Program’s Registries for Evaluating Patient Outcomes: A User’s Guide. This guide includes recommendations about data elements and sources, study design, data quality assurance, analysis and reporting, ethics, privacy and governance. The report also recommends that registry managers establish criteria to determine when a registry is no longer needed because of the fulfillment of the stated objectives.
7) The quality of the data can be suspect when there is limited funding available for the collection and audit of auditing and for the collection of the data. The number of data collection fields should be strictly limited to those that answer the CED questions and contribute to the accuracy of the data. These data have value but are distinct from the value drawn from randomized clinical trials. When the CED questions are answered, the requirement for reporting should sunset.
8) Collaboration among CMS, FDA, NIH, other federal agencies that fund research and non-government entities presents an opportunity to coordinate efforts to ensure funding and timely completion of well-designed studies to answer CED questions.
Thank you for the opportunity to provide input and share lessons learned. If additional questions arise, or if the Heart Rhythm Society may be of further assistance to you or your staff, please contact Laura Blum, Vice President, Health Policy at 202-464-3489 or firstname.lastname@example.org.
Bruce L. Wilkoff, MD, FHRS, CCDS
President, Heart Rhythm Society