Problems in the evaluation of new devices for coronary intervention: what have we learned since 1989?

The objectives of this study are to review the problems associated with the evaluation of new devices, the progress made in that evaluation process since 1989, and the role played by the New Approaches to Coronary Intervention (NACI) registry. In 1988-1989, the first wave of new coronary devices (stents, atherectomy, laser catheters) were entering clinical investigation. It seemed unlikely that the small manufacturer-run registries used to gain approval for earlier balloon catheters would be adequate to evaluate the host of complex new devices, each of which might be used for a restricted set of anatomic indications. Moreover, the wide range of arbitrary definitions then in use for fundamental outcomes (such as success, complication, and restenosis), effectively precluded meaningful device-to-device comparisons. Against this backdrop, the NACI registry was formed with National Heart, Lung, and Blood Institute funding to provide an independent and standardized evaluation of the first 8 new devices under evaluation in the United States, across the broad range of their application. The registry employed a unique modular form set to track the sequence of events during complex cases in which serial new devices and balloon angioplasty might be used, either in a planned way, or an unplanned way (to treat complications or suboptimal results). Outcomes were subjected to standardized criteria for (1) the reason for device use (planned, unplanned); (2) success (device, lesion, and procedural success); (3) complications (a) major (death, Q-wave myocardial infarction, and emergency coronary artery bypass grafting); or (b) other (groin complications, non-wave myocardial infarction, etc.); and (4) clinical restenosis (any subsequent revascularization, target lesion revascularization). Separate funding for an angiographic core laboratory was obtained in 1992, which analyzed 3,936 (88.9%) of the 4,429 films obtained on patients enrolled between November 1990 and March 1994. The NACI registry has addressed a broad range of problems inherent in the evaluation of new devices for coronary intervention. Whereas the approval process has moved progressively towards randomized clinical trials (and away from registries), the NACI registry offers a unique view of current practice, outside the narrow scope of the limited number of randomized trials that have been performed to date. This article shows, however, that we have learned about more than the devices themselves since 1989-we have also learned about the importance of knowing the reason for device use, using precise definitions of endpoint variables, understanding the financial and reimbursement ramifications of new device trials, and upholding strict investigator ethics during the conduct of such evaluations.