Update of the National ICD Registry: Interview with Stephen Hammill, MD

Interview by Jodie Elrod
Interview by Jodie Elrod
In November 2007, EP Lab Digest spoke with Dr. Hammill about the first-year results of the NCDR® ICD Registry™. In this feature interview, Dr. Hammill provides an update of the registry. For those who may be unfamiliar, remind us of why the national ICD Registry™ was initially created. The ICD Registry™ was required by CMS. In fall of 2004, Medtronic requested CMS to expand their ICD coverage based upon the SCD-HeFT trial. CMS agreed to expand coverage, but felt that there were some areas where the data was not sufficient. Therefore, they established a new program called Coverage with Evidence Development (CED). The CED process allowed them to expand coverage but required that additional data be collected on groups of patients in which CMS felt more data was required. To collect that data, CMS imposed the requirement of participation in a registry, so if a hospital wanted to get reimbursed for a Medicare patient seeking a primary prevention ICD, the hospital had to participate in the registry. There was also the National ICD Registry Working Group of which I chaired, which put together the concept and infrastructure of the registry. Eventually the registry developed by the Heart Rhythm Society and the NCDR was chosen by CMS as the registry to handle this data. What have been some of the benefits of having this registry? Approximately how many implants now have been reported? The most current totals are 330,000 implants and over 1,500 hospitals. From the hospital and physician perspective, I think the most important aspect of the registry is being able to use it as a tool for quality improvement. On a quarterly basis, hospitals receive a benchmarking report that compares their outcomes with those hospitals of similar size. Outcomes include things such as the type of device used, the complications associated with the procedure, and hospital length of stay. Every item that is collected is compared nationally, and this gives hospitals an idea of their quality performance and benchmarking. Accreditation organizations are also very interested in hospitals doing this type of comparison using this benchmarking approach; the Joint Commission in particular is quite supportive of this type of registry. In addition, there are some insurance companies that are now looking at requiring hospital participation in this registry for the patients that they insure. How has the registry changed over the years? Have pediatric ICD implants and surveillance of lead performance been added to the registry? The major change is that we keep enrolling more implants, and this helps to improve the quality of the data that can be used for research. Regarding pediatric ICD implants and the surveillance of lead performance, it has not at this point been added to the registry. The NCDR has a process where periodically they will look at an existing registry and then go over what was working and what was not working on the data analysis form, what was causing confusion, what needed to be added, and if there were any other things that we should start collecting based upon either national guidelines or something that may have happened nationally. That is what pushed it to develop to the next version of the registry, which is called version 2.0. This has been developed and testing has been finished, and it is now with the programmers at the NCDR changing the format and data entry of the forms. It is also with the vendors at the NCDR, because there are some software companies that are providing software to enter data into the registry. Version 2.0 cleans up the basic form and adds pediatric patient data as well as lead data. Some of the pediatric patients were being entered into the registry if the ICD was implanted in a participating hospital that had chosen to submit ICD implantation data on all patients that were being done, not just Medicare patients. However, most of the hospitals that place ICDs in pediatric patients are specialized hospitals or children’s hospitals. Therefore, we teamed up with the Pediatric & Congenital Electrophysiology Society (PACES), who was quite interested in participating in the registry, and will start to collect data from pediatric ICD patients beginning in January 2010. Then there is the issue of lead performance. When you look at problems that can occur with an ICD implant, occasionally there are problems that develop with the generator, causing a generator recall, but it is more likely that there will be a problem with one of the leads. In general, leads have a higher percentage of failure problems. We at the NCDR have been working with the Food & Drug Administration (FDA) and the Heart Rhythm Society to develop a new aspect of the data entry form that will follow ICD lead data. This will cover any lead that is implanted at the time of the ICD generator, including atrial leads, ventricular leads, left ventricular leads, defibrillator leads, and epicardial leads. We will collect data at the time of implant, and then whenever the generator or one of the leads is seen again. So if a generator has to be replaced for example because of normal end of battery, or if there is a problem with the generator or if it needs to be upgraded to a more sophisticated device, then new data will be collected and we will receive follow-up on the generator. The same will occur with leads, so if a lead has to be replaced, revised or extracted, if there’s an infection, if new lead setup is needed, or if there is a recall, then that new lead data will be entered into the registry. This will allow us a way of following lead performance. The FDA is quite interested in using this as one of the various tools available for lead surveillance once the lead is improved and starting to be implanted. How was the registry funded in the past and currently? What roadblocks are being run into with financing and support? There are two aspects to this. The Baseline ICD Registry, in which data is entered at the time of the implant, was initially funded (for the first two years) in part by industry at the request of CMS; hospitals also participated, but with a lower level of funding. After two years, hospitals have been the group that have been funding the registry. It costs about $3,000 for a hospital to participate in the registry, regardless of how many patients they enroll in the registry — this is similar to the other NCDR registries such as the CathPCI Registry®. So the funding for the registry now is done through hospitals, and hospitals look at it as a quality improvement initiative. Of course, hospitals have the additional cost of paying for the personnel that are needed to enter the data into the registry. There is also the Longitudinal ICD Registry Study, which is the research aspect that is being done to answer questions from Medicare to see if there could be full expansion of Medicare coverage to three specific groups: patients with an ejection fraction of 31-35%, patients with class IV heart failure receiving a CRT defibrillator, and patients with a dilated cardiomyopathy of less than nine months’ duration. CMS is interested in seeing the number of times the devices are used and if the devices are functioning appropriately, meaning if there are life-threatening rhythms that are properly detected and treated by the device with either pacing or a high-energy shock. However, the Longitudinal ICD Registry Study has had a great deal of difficulty getting funded. It is a Medicare mandate, but there was no funding that came with this. Therefore, we have pieced together funding from industry, AHRQ, NIH, and the insurance companies, so the Longitudinal ICD Registry Study will now be operated by the Cardiovascular Research Network, which is a group of large HMOs including the Kaiser organizations. Will the current administration affect or help this situation? We would hope so. Part of the stimulus package includes funding for comparative effectiveness research, which is basically to look at different strategies for treating patients, so there may be some research that is developed that will utilize the registry. Are smaller registries also needed to monitor the reliability of implantable devices? There are some registries that are operated by the device companies. They follow their own devices for a certain period of time in long-term studies after the devices are approved. The Medtronic study is called the System Longevity Study. These smaller registries are looking at long-term device performance. However, one of the great strengths of the national ICD registry is that it enrolls a large number of patients, so if there is a problem with a lead or a generator, it can be easy to quickly find how many people are affected. Another type of registry is being operated by Dr. Robert Hauser at the Minneapolis Heart Institute, and we at the Mayo Clinic have participated in this as well. So now if we see a problem with a lead or a device, we send that information to the FDA, the device company, and to Dr. Hauser’s registry. With his registry, Dr. Hauser gathers reports that show if a particular lead is starting to have more problems than would be expected; however, there is no way of showing how many leads are out there and how big a problem it is. Therefore, I can fully see in the future that a registry such as this will send an alert that there may be a problem, and then we would refer to the ICD registry to get an idea of how extensive that problem is.