During clinical trials, clinicians and coordinators record descriptions of symptoms, procedures, and medication when describing adverse events, medical history, and concomitant medications. These descriptions are known as verbatims. Once these verbatim terms are recorded, the study Sponsor is required by the FDA and other regulatory authorities to code them to industry standard dictionaries.

During a study, medical coders have to code thousands of terms. The traditional ‘browse and code’ approach can take medical coders several minutes per term, adding up to tens or even hundreds of hours per clinical trial.

In our next-generation medical coding solution, Rave Coder+, Medidata has used over 60 million historic coding decisions to develop, train and test an algorithm that delivers highly accurate coding predictions for clinical verbatims. This white paper explains how the algorithm in Rave Coder+ was developed, trained, and tested to provide a solution that automatically codes terms in seconds rather than minutes, saving hours of coding time per study.