Oncology Without Oncologists, Brought to You by the Power of Supercomputers

IBM’s Watson is displayed at a press conferencebefore the “Man V. Machine” Jeopardy! competition at the IBM T.J. Watson Research Center on Jan. 13, 2011, in Yorktown Heights, N.Y.

Photo by Ben Hider/Getty Images

IBM brought a little Watson roadshow to Capitol Hill this week to show off some of its capabilities now that the Jeopardy!-playing supercomputer has been studying health care. IBM is a little cagey about what the specific plans and agenda are here, but if you look at this video of Watson as an oncology diagnostics tool, I think you’ll come to see very quickly that there’s an important public policy issue that legislators are going to have to confront:

The video talks about Dr. Mark Norton, a clinical oncologist, as the user of this software. And surely it’s true, the first users of Watson as a diagnostic tool will be doctors. But on some level having a doctor use Watson would be like hiring a skilled carpenter to work in the Ikea warehouse. The whole point of this machine is that it can amass all the background oncological knowledge of a medical doctor and then people who don’t have that knowledge can go use it.

That will make this kind of initial consultations cheaper and more widely available, and then doctors will spend their time focusing on things that Watson can’t do (yet), meaning those services will also become cheaper and more widely available.

Maybe the right person to go through this Watson exercise with you is a nurse or a nurse practitioner. Or maybe what’s really needed is a kind of professional we don’t have yet, someone whose training is primarily focused on a counseling and social work function—providing the “human element” that the medical database and digital diagnostic tool lacks. Medical science is about diagnoses and treatments, but there’s much more to the on-the-ground practice of effective medicine than that scientific core.

Where legislators come in is that health care is one of the most heavily licensed and heavily regulated fields of endeavor in the world. Already there are substantial state-to-state differences in what nurses are allowed to do without the supervision of doctors, differences that tend to reflect the entrenched political power of doctors rather than any sound public policy research. These “scope of practice” regulations are a state matter, but with the federal government picking up such a large share of the health care tab, congress really ought to take an interest.