Right after IBM's Watson DeepQA super computer beat down its human contestants in Jeopardy! in February (and which is being rebroadcast this week), there was news that its next field of study - or conquest - was going to be in the area of medical diagnosis and recommended treatment.
"... IBM, along with Nuance Communications Inc. and the Columbia University and University of Maryland medical schools, announced that they are developing Watson as a diagnostic tool that can help doctors identify diseases and recommend treatments. They hope to begin lab tests as early as next year, with real world testing later in 2012."
You can also hear Spectrum Senior Associate Editor Steven Cherry interview IBM artificial intelligence researcher Chris Welty about using Watson-like programs to help doctors in hospitals in this podcast from March.
I guess 2012 has come early this year, as IBM and WellPoint, the largest health benefits company in the US, announced yesterday that they reached an agreement,
" ... to create the first commercial applications of the IBM Watson technology. Under the agreement, WellPoint will develop and launch Watson-based solutions to help improve patient care through the delivery of up-to-date, evidence-based health care for millions of Americans. IBM will develop the base Watson healthcare technology on which WellPoint's solution will run."
What are WellPoint's expectations of "Dr. Watson"?
"Watson's ability to analyze the meaning and context of human language, and quickly process vast amounts of information to suggest options targeted to a patient's circumstances, can assist decision makers, such as physicians and nurses, in identifying the most likely diagnosis and treatment options for their patients."
"... we envision that new applications will allow physicians to use Watson to consult patient medical histories, recent test results, recommended treatment protocols and the latest research findings loaded into Watson to discuss the best and most effective courses of treatment with their patients."
"Watson may help physicians identify treatment options that balance the interactions of various drugs and narrow among a large group of treatment choices, enabling physicians to quickly select the more effective treatment plans for their patients. It is also expected to streamline communication between a patient's physician and their health plan, helping to improve efficiency in clinical review of complex cases. It could even be used to direct patients to the physician in their area with the best success in treating a particular illness."
Sounds interesting. However, one operating assumption seems to be that for Watson to "consult patient medical histories" the patient's medical record will need to be in the form of an electronic health record that Watson can easily access. While some progress has been made towards this objective, widespread use of EHRs in the US still has some way to go. Also accessing the relevant patient data may require a lot of interfaces to the various vendors' EHRs to be written.
Second, as noted in a Minneapolis - St. Paul Star Tribune article last month, doctors might balk at having to use Watson to help make diagnoses. No one likes their professional judgments to be second-guessed, especially by a computer.
In addition, doctors might worry that their diagnoses will be graded against that performed by Watson (or against whether they used Watson or not in making their diagnosis). Even if not used in this way, the use of Watson might be seen by insurance companies as a way to speed up diagnoses, meaning that you may get to spend even less time with your doctor than you do now discussing your illness (as well force doctors to see even more patients per day).
Further, if patients get direct access to Watson and its diagnosis differs from that of the doctor, things could get interesting. Doctors already complain of patients accessing the Internet and misinterpreting the medical information they find, or worse, becoming cyberchondriacs. Using Watson, with its perceived authoritative findings, creates a horse of a different color altogether. Watsonchondriacs, anyone?
[Update: 21 September 2011
There was a story in the Wall Street Journal about the US Agency for Healthcare Research and Quality and Ad Council's latest initiative "to encourage clinicians and patients to engage in effective two-way communication to ensure safer care and better health outcomes." Again, how communication between doctor and patient will be impacted by Watson is yet to be seen.]
Fifth, are there any legal ramifications to using Watson? Can IBM be sued for malpractice if the diagnosis Watson provides is incorrect, or will using Watson be seen as a good defense against litigation? What is the expectation that the diagnosis provided by Watson will always be correct? Will patients have even a greater expectation than they do now that medicine is a perfect science?
Sixth, if a health benefits company like WellPoint decides to use Watson for pre-authorizations of medical treatments, will it also be programmed to look for the cheapest treatment that works most of the time instead of the one that might be best for the individual patient but at a higher cost?
And finally, as noted in the Star Tribune story, there is projected to be a 20,000 or more shortfall in doctors, especially primary care doctors, in the US. Access to Watson by patients might be a way to help alleviate the shortfall, or even to reduce the use of emergency rooms which people often inappropriate use for non-emergency health care reasons.
The next few years in healthcare will be interesting to watch with the addition of Watson.
By the way, would you have any problems with your doctor using Watson to help diagnose your illness? Would you like access to Watson to help you diagnosis your own or your children's illnesses? Would you likely believe Watson if the diagnosis differs from that of your doctor?
Contributing Editor Robert N. Charette is an acknowledged international authority on information technology and systems risk management. A self-described “risk ecologist,” he is interested in the intersections of business, political, technological, and societal risks. Along with being editor for IEEE Spectrum’s Risk Factor blog, Charette is an award-winning author of multiple books and numerous articles on the subjects of risk management, project and program management, innovation, and entrepreneurship. A Life Senior Member of the IEEE, Charette was a recipient of the IEEE Computer Society’s Golden Core Award in 2008.