Characterization of the consistency of healthcare components using LDA

Life data analysis (LDA) using Weibull modeling and the assumption of constant parameters can often be misleading. This paper will show that healthcare components may require a modification to the common analysis methods due to a manufacturing process variability over long production campaigns. The specific medical subsystem, on which this paper is focused, referred herein as Component B, has some interesting reliability behaviors. There exist several failure modes, any of which could render the component unusable. Each of these modes manifests themselves at different times and at various percentages depending on the production and the use environment. These present the reliability engineer with the undesirable situation of waiting a considerable amount of time for the process to “mature”. This maturation thus allows characterization and resulting predictions of Component B´s life. Weibull modeling by time period in conjunction with Nevada chart technique is used to better understand the lifecycle of Component B. Misinterpretations surrounding healthcare component warranty analysis are often caused by the lack of proper understanding regarding field failure data. Several examples presented in this paper regarding failure modes and data analysis by quarter of production proves that some healthcare components and their subsets require independent analysis.