In a 2005 paper (published in 2006), Eric von Hippel and his colleagues showed that the majority of new therapeutic applications approved one year by the FDA were due to user innovation by doctors and other clinicians. This comes through the practice of off-label prescription of existing therapies, which doctors vigorously defend as important both for scientific progress and helping individual patients.
The possibility of abuse has been used by the FDA to justify prosecuting doctors and pharma companies for off-label uses. A famous case is Xyrem, in which the FDA prosecuted a drug salesman, a doctor and the company. As I summarize the full story in my Bio Business blog, the company settled for $20 million, the doctor killed himself in 2011, but the salesman won his case last month on appeal.
At KGI, we often remark on how innovation is different in the heavily regulated field of human therapeutics. Getting FDA approval takes regulatory skills, time and money far beyond what’s required for radio handsets — let alone unregulated software or e-commerce. There’s no college students shipping products from their dorm rooms.
The life sciences map reasonably well onto open innovation. At last summer’s open innovation conference in London, we had two papers about open innovation by drug companies — one by Henry Chesbrough himself, and one by my KGI coworker Steven Casper. Even if these companies use OI, they still bring products to market through the same process of clinical trials, regulatory hoops, and expensive sales channels.
User innovation is another story. As with other user innovation, the off-label innovation is a decentralized, democratized innovation process. Consonant with von Hippel’s “sticky information” thesis, off-label decisions allow those closest to the point of need to find solutions to those needs. However, by their nature doctors (and nurses and patients) don’t have the hundreds of millions needed to go through normal regulatory channels.
Von Hippel and Carliss Baldwin hypothesized a division of labor between users and producers (that was later published in Org Science). Regulatory costs and restrictions clearly can distort this division of labor, favoring those (producers) that have access to capital over those who have scarce knowledge — suggesting an opportunity to modify and update their model.