Clinical trials are collecting far too much data

The scientific analysis ecosystem has by no means been extra technologically succesful — or extra burdened by its personal ambition.

In December 2025, the Tufts Heart for the Examine of Drug Improvement and our group at TransCelerate printed analysis that implies practically 30% of the information collected in scientific trials don’t immediately inform key selections, but sufferers are nonetheless requested to offer it. Moreover, since 2005, the variety of procedures per protocol has elevated by practically 140%, endpoints by greater than 200%, and information factors collected by over 600%.

I work with heads of R&D from 18 of the world’s largest pharma firms. It’s abundantly clear to me — and to lots of them — that the business’s tendency towards “extra is best” relating to information assortment is essentially undermining trials, not enhancing them.