Using JMP® as a Common Language to Interpret Lab-to-Lab Data: A First-Time JMP User’s Perspective
Aug 13, 2015 6:23 AM
A Quick Reference.jsl
JMP Summit Journal.jrn
Andrew Kampfschulte (kampfsca), Biostatistics Student Associate, Perrigo Company Rob Lievense(statpharmer), Senior Statistician, Perrigo Company Donna Seibert, Senior Principal Scientist, Perrigo Company
The United States Pharmacopeia (USP) and the International Conference on Harmonisation (ICH) are in the process of revising standards for the elemental impurities in pharmaceutical drug products. The new, lower limits for trace-level elemental impurities create the need for increased precision and sophistication of laboratory methods (ICP-MS) over the outdated heavy-metals wet chemistry testing methods. An interlaboratory study with 12 European and North American labs, eight elements, and over a dozen covariates was conducted to assess lab-to-lab consistency and to generate a data-driven way to communicate the complexity of ICP-MS methodology. JMP 11.2 was utilized as a common language to visualize and analyze ICP-MS data presented to the consortium, as it was easy to learn and apply for a first-time user. Extensive use of the Distribution tools provided dynamic exploratory analysis. The Measurement Systems Analysis platform was later used to investigate variance components, which led to the conclusion that lab-to-lab consistency was significantly improved when labs were given flexibility in their methods to suit their individual instrumental configurations. JSL scripting was used to create a common data set from multiple formats and automate analyses for efficient and thorough feedback.