Using Standards to Make Clinical Research More Effective

Session chaired by Phil Kirsch

Description:

There was a time when everything from clothing to farm equipment was hand-crafted, one item at a time. Then, standardization made manufacturing possible. In fact, if it weren’t for standards, most of the products consumers take for granted today would either not exist or be available only to the wealthy. But, we’ve only begun to reap the benefit of standards in Clinical Research. Despite CDISC, HL7, and similar efforts, each study tends to be a one-off, hand-crafted effort.

It doesn’t have to be that way. If clinical protocols were developed following an XML-tagged approach, computerized systems could draft study databases without human assistance and validate them against dummy datasets built from typical values. As study data is collected, it could be compared to the dummy data, and significant differences could be flagged for human investigation. Some computerized systems are already trained to observe enrollment trends and recognize adverse events that human caregivers might miss. And it won’t be long until someone starts using AI to analyze study results, produce tables, listings, and figures, and even draft clinical trial reports.

A few organizations have begun to explore these and similar possibilities, which will be presented and discussed. At least one prototype system has been designed, and example databases are under construction. Please expand your vision by joining us to consider ways to create better standards and use them more effectively to speed up and improve clinical research.