In 2005, the results of the Sudden Cardiac Death in Heart Failure Trial (SCD-HeFT) were released and published.1 The study followed approximately 2,500 ischemic and non-ischemic cardiomyopathy patients over at least five years; patients were randomized to receive standard medical therapy alone, standard medical therapy plus amiodarone, or standard medical therapy and an ICD. Study results showed a 23% decrease in the mortality rate among patients in the ICD arm. SCD-HeFT was the first large trial to prove the benefit of an ICD in the non-ischemic population. The Centers for Medicare and Medicaid Services (CMS), the agency responsible for funding Medicare and Medicaid, anticipated that the study results would lead to approximately 25,000 new implants in the year following the release of the study findings.2 In January of 2005, CMS released revised and expanded coverage guidelines for ICD implantation that included the use of ICDs for primary prevention in the ischemic and non-ischemic population, and also abolished the requirement of QRS duration >- 120 ms for non-CRT devices. However, the expanded coverage is not without condition. Data on all patients who receive an ICD for primary prevention must be entered into a national database. Data will be analyzed in an attempt to determine which patient populations may or may not receive the greatest benefit from ICD implantation, which in turn may lead to improved patient screening and avoid unnecessary procedures. Hospitals had until January 2007 to begin registering patients in this database, known as the ACC-NCDR ICD Registry. ICD Registry at LGH Lancaster General Hospital (LGH) is a 590-bed community hospital in Lancaster, Pennsylvania. The Department of Electrophysiology and Pacing consists of 4 dedicated electrophysiology labs, and employs 15 technologists and 7 registered nurses. Eight electrophysiologists and cardiologists perform approximately 3,600 procedures a year, including diagnostic and interventional EP studies, device implants and revisions, and pulmonary vein isolations. For fiscal year 2006, we implanted 350 new ICD systems, of which approximately 300 were for primary prevention. For fiscal year 2007, we have been averaging approximately 25 primary prevention implants per month. So how does the Registry work? As it turns out, it is not terribly difficult, but it can be a fairly cumbersome and time-consuming process, especially for institutions like ours with a high implant rate. The first step of the process, of course, is the implant procedure itself. After the patient is discharged, the Registry data collection form must be completed. The form, which is three pages long, tracks a large variety of data, including demographics, admission data, history and risk factors, diagnostic studies, implant data, adverse events, discharge data and discharge medications (Figure 1). After the data is collected, the form is forwarded to a database administrator (currently the administrator for the cardiac surgeons is entering our data) and is entered into the Registry database; data input takes about 15 or 20 minutes for each patient. Thus, at our institution, actual data collection and entry takes an average of 45 minutes per patient. In order to comply with the new requirements by the deadline, we had to gather and input data on a backlog of approximately 310 patients. This task required many hours of work for several employees, who squeezed it in among their regular duties. In fact, it has been so time-consuming that we are planning to hire a data abstractor and a database administrator, whose duties will include maintaining the Registry database. The data retrieval and entry process may be more or less cumbersome for an individual institution, depending on their system of chart storage. At our institution, patients' paper medical records are scanned and stored electronically, and are easily accessible for review at any time. In facilities without electronic records, the paper charts must be requested, retrieved, stored, and re-filed, a process that can be time-consuming and somewhat labor-intensive. At Lancaster General we have internally collected data on our ICD patients for over a year. We have used the data not only to help fulfill our registry requirements, but also to ensure that patients met implant and reimbursement guidelines. As a result, our Registry experience may have been eased somewhat, as our data collection apparatus was already in place when the Registry opened. Hospitals that are starting from scratch may well have found the process much more difficult. Data entry is not the end of the process, though. Each quarter, the institution receives a report listing the number of deficiencies in each data field (demographics, admission, etc.). Unfortunately, the report only lists the number of deficiencies in each field it does not include the name(s) of the patient(s) for whom data is missing. As a result, one must go though the data in that particular field case by case until the deficiency is found; then the information must be retrieved from the chart and entered into the database. This requires varying amounts of additional time and effort. Also, the database cannot be queried or customized for an individual institution, so additional data fields cannot be added and the information you have entered cannot be used for QA, benchmarking, etc. ACC-approved software is available that allows you to run queries and add fields, but that is at additional cost to the institution. The ICD Registry is a Federal requirement, and its long-term benefit remains to be seen. The process of data retrieval and entry can be cumbersome, time-consuming, and somewhat labor-intensive, but may become less so as we become more familiar with the process. It is unfortunate that the data we collect for the Registry cannot be used for QA or benchmarking at this time, but future software upgrades may provide us with that capability. Nevertheless, we are hopeful that the time and resources dedicated to the Registry will result in a database of information that provides new insight into our clinical practice, while helping to keep a lid on ever-rising healthcare costs as well.