As a statistician working at a CDMO, I frequently conduct comparability studies on diverse data from various departments. These studies involve comparing methods, scales, sites, years, and more. Our general approach includes visual comparisons, variance tests, t-tests, and TOSTs. However, analyzing numerous variables and data sets can be time-consuming, particularly as TOSTs require manually entering practical equivalence values for each variable, which significantly adds to the workload.

To streamline this process, we've developed a comparability script/add-in that automates the analysis and adapts to different data sets. Users can input parameters to select study variables, define test and reference groups, and specify practical equivalence values for TOSTs, whether as a multiplier of standard deviation or a specific value. The script generates comparison visuals automatically, conducts variance tests to guide appropriate t-tests, and performs TOSTs. It produces a JMP report containing the selected data, a summary table of test results, and analytical output. This automation and standardization have significantly reduced the time required for comparability studies across various projects.