[mattkrause]

Academia also isn't a great home for large-scale trials, especially long ones.

Most academic "things" (grants, hiring/promotion/graduation criteria) expect people to produce first-author papers, so projects are usually set up to involve 1-3 people over 1-3 years.

A trial, on the other hand, needs lots of people, ideally at many different sites, and will hopefully produce a single, well-defined outcome. Somebody's also got to make the drug--and under conditions where the end product can be given to humans (i.g., GMP). None of this is cheap either, and the NIH budget wouldn't stretch to too many large Phase III trials.