A modified actuarial life-table approach to the analysis of implantable device performance

The actuarial life-table method is often used by pacemaker manufacturers and the pacing research community to describe pacemaker and lead performance. Most life-table methods allow for differing lengths of follow-up but assume that all devices were followed from implant. Occasionally, however, devices come under follow-up observation sometime after implant. This presentation describes an extension of the actuarial method to accommodate these kinds of data. The specific example to be considered involves follow-up data collected by CardioCare, a commercial cardiac monitoring service, on the performance of Medtronic polyurethane leads. Patients subscribe to this service, generally at some time after actual device implant. Results showed that of 12,112 patients with Models 4002, 6971, and 6972 leads who were followed by CardioCare, only 85 were followed from implant. If one were to exclude patients not followed since implant, more than 99% of the data would be lost. Using the modified approach with allowance for postimplant, entry resulted in an estimated three-year cumulative survival probability for these leads of 95.7%. Treating all patients as if they were followed since implant, the probability would be 96.9%, an optimistic and biased estimate.