Training Decay Selection for Usability Validation

Abstract

When conducting usability validation testing, representative users must use the device in the expected conditions of use in the field. There is usually a period of time—days or weeks—between the point in time a user is trained, and the moment they use the device for the first time. For this reason, the FDA acknowledges the need for “training decay” as part of usability validation testing, but manufacturers face challenges simulating real-time decays. In response to challenges associated with lags of days or weeks between training and usability validation testing, medical device manufacturers typically simulate shortened training decay periods. This paper discusses the theory behind the shapes of various training decay curves and the variables that drive differences between training decay curves. The author proposes to use a task-based approach for defining training decay curves in usability validation studies and sets out generalized training decay curves at a high level. Future research could reveal detailed and generalizable training decay curves. Identifying generalizable training decay curves could standardize the usability testing required for medical devices, and ultimately improve use error identification while avoiding an undue toll on manufacturer resources.