Statistical Issues in the Product Life Cycle of Implantable Medical Devices*Duane Steffey, Exponent, Inc.
Keywords: corrosion, fatigue, accelerated test, Bayes, post-market surveillance
Implantable medical devices, such as cardiovascular stents and orthopedic joint replacements, hold the potential to increase both the longevity and quality of people’s lives. During product development, statisticians may be involved in designing and analyzing engineering tests conducted under accelerated conditions to assess device susceptibility to corrosion or fatigue under in vivo conditions. Such tests complement clinical studies of device efficacy and safety, in which Bayesian ideas find growing acceptance as viable alternatives to traditional approaches. Statistical methods for controlling the production process and end-of-line quality, as well as data mining tools for troubleshooting problems, find application in the large-scale manufacture of devices. The frequency and severity of adverse events involving FDA-approved devices have stimulated ongoing research to develop more useful data sources and more effective analytic methods for timely and accurate post-market safety surveillance. This talk will address the role of the applied statistician at various stages of the product life cycle, touching on novel methodological aspects and pertinent regulatory considerations.