Objective. To propose a standard definition of "out of specification" for cochlear implants and a paradigm for inclusion of category C of the ISO standard 5841-2:2000 for reporting in cumulative survival statistics. Hypothesis: A standard definition of "out of specification" and consistent reporting by manufacturers of cochlear implants will create a fair and consistent representation of cumulative survival. This will allow discernment of differences between manufacturers for reliability and for detection of trends in reliability between model types from the same manufacturer. Design: Three separate meetings with representatives of the three manufacturers of cochlear implants marketed in the United States were staged over a 13-mo period. Standard questions, created by the authors, were addressed by each representative to determine the current state of device reliability reporting. Results were presented to clinicians at the William House Cochlear Implant study Group and the Implantable devices sub-committee of the American Academy of Otolaryngology (2004, 2005) and at the 8th International Cochlear Implant Conference (2004) for feedback. After assimilation of feedback by all parties, the standard was written and reviewed by representatives from each manufacturer for accuracy of data. Results: A complaint-driven standard was developed. A "cochlear implant" as an internal device placed and skin closed in surgery. An internal device is "out of specification" when one or more technical characteristics is outside the limits of normal function and results in explantation or nonuse by the patient." Children will be reported separately from adults, each model of device will be reported on annually, a minimum of 200 devices must be in each model group for Cumulative Survival Reporting (CSR). Confidence limits are set at 95%. Explants will be determined to be "biological" or "technical." Technical explants are included in CSR reports. Devices failing to meet specifications set by the manufacturer, not in use but still in situ due to patient choice not to be re-implanted are considered category C and included in CSR reports. Implants that cannot be classified at explant are placed in an "under investigation" category while evaluation is completed. If no classification is made by 6 months, these devices will be included in the CSR report. Notification to the implant center regarding "in" or "out of specification" will be made within 60 d of the explant arriving at the manufacturer with final root cause of failure reported to centers when complete. Information will be passed on to patients by members of the implant team. A standardized form will be created to provide the manufacturers with necessary patient information to guide reliability analysis, including performance after re-implant. Conclusions: The standard for reliability reporting described in this paper improves patient care by presenting data which are understandable to clinicians delivering cochlear implant services. It fosters fair and accurate reporting without discriminating or granting perceived advantage to any manufacturer. This standard provides a basis for reporting research related to or including device reliability in the medical literature.
