Little over a month ago, IBM and WellPoint announced an agreement wherein WellPoint will deploy IBM’s latest and greatest super computer and artificial intelligence mega-mind Watson. Watson’s claim to fame was its ability to beat the human Jeopardy champions much like Big Blue beat reigning chess champion Garry Kasparov in 1997. Since that Jeopardy match, IBM has been quite vocal about its desire to apply Watson in the medical arena, we’ve been buried in press releases and briefings, but the WellPoint announcement is the first one of any real consequence. Having interviewed both IBM and WellPoint, following is our review and assessment.
Background:
Watson is a relatively new form of artificial intelligence, based to some extent on neural networks. What is unique about Watson is that it has been developed (trained) to understand the nuances of language. It is a question & answer system that uses among other techniques, natural language processing, to extract meaning out of unstructured data. In developing Watson for the Jeopardy challenge, one of the key design parameters was for Watson to answer a question in under three seconds – plenty fast enough in a diagnosis/treatment decision scenario. This is a key reason why Watson may have enormous utility in the healthcare sector where so much data is unstructured, the pace of change is so high and the ability to chose the optimum treatment patient plan for a given diagnosis is less than ideal today.
WellPoint is the largest payer in the US with some 34.2M members and 14 Blues across the country. Despite this impressive size (or maybe because of it), WellPoint has been far less aggressive than others in the HIT realm, especially for those systems used by providers. In signing this deal with IBM, WellPoint is signaling to the market, and likely those on Wall Street, that they intend not to be left behind. In asking WellPoint about their HIT strategy, WellPoint CIO Andrew J. Lang told Chilmark that WellPoint’s intent is not to create a new line of business (as is the case with UHG and Aetna) but to improve the quality of health delivered to their members by providing physicians the best tools possible. Certainly a noble goal, but only time will tell as to how closely they adhere to such a goal.
What it is:
While IBM is pursuing a number of potential vertical markets for Watson, including financial services, this is the first actual “Win” for IBM’s Watson. Money is changing hands with WellPoint paying an undisclosed amount to deploy Watson.
Watson will be deployed as a “cloud-based” service. Actual pricing for accessing this service has yet to be determined but as WellPoint put it to us, they do not want to create barriers to physician use. Thus, don’t be too surprised if it is offered virtually free to in-network providers, clinics and hospitals.
The Watson intelligence service will focus first on three oncology domains (breast, lung and colon cancer) that WellPoint’s internal clinical staff have determined most promising. These three were chosen for they are areas where WellPoint’s claims data shows high variability in treatment; there are significant, on-going advances in research and treatment; and a high likelihood for demonstrating the utility of a system such as Watson.
WellPoint does not intend to displace physician decision making, but augment it. WellPoint states that physicians will still be able to ultimately make their own decision as to the best course of action (treatment) for a given patient. As CIO Lang stated:
Watson is intended to be a doctor’s assistant, the doctor is still in the driver’s seat.
Watson is currently undergoing “training” with reams of data (research, claims data, etc.) in the three oncology domains being fed into it, questions being posed, answers evaluated against real clinical evidence to bring Watson up to a significant “confidence level.” The Watson service will be released in the first quarter of 2012.
Implications:
Among the multitude of announcements that pass across our computer screen on a daily basis, this is one that really piques our interest. Like any exceedingly powerful technology, Watson has the potential for good and likewise the potential for harm. If WellPoint follows its stated intent of deploying Watson as a service, as an assistant to the practicing physician, facilitating the care process with more rapid and accurate first time diagnosis and suggesting a treatment plan that is most relevant to that specific patient, then Wow, this could be truly game changing and far in excess of what other payers have done to date in the HIT realm. If, however, WellPoint’s deployment of Watson becomes prescriptive wherein physicians are no longer in the driver’s seat, then “Houston, we have a problem.”
Having personally seen what can happen when someone is misdiagnosed, when appropriate treatment is delayed, particularly for an aggressive form of cancer, the anguish and subsequent anger is nearly unfathomable. If Watson can indeed short-circuit the diagnosis and guide physicians to the most appropriate treatment in an expeditious manner, well then hat’s off to IBM for developing Watson and WellPoint for taking the risk to be the first to deploy Watson in the healthcare sector.
Dr. Watson, do you accept house calls?
Game changing? Maybe, but Aetna acquired an equally amazing technology in 2005 (est.): Active Health Management.
http://www.activehealth.com/
20 MDs analyzing dozens of medical journals full time, for years. Loading thousands of validated clinical data points into a decision support tool and actually comparing, real time, to claims data and lab values – to be sure care matches evidence, then messaging directly to the attending MD, pharmacy and patient. A powerful tool accessible in an ACO’s HIE.
They may be playing a bit of catch up but it will be interesting to follow Wellpoint and Watson in the coming years.
ps. I enjoy reading your research.
Paul,
Thanks for your comment and insight regarding ActiveHealth. Do agree that ActiveHealth had some interesting technology under the hood but part of its problem, and you infer such in your comments is that it was heavily reliant on human resources and thus not easily scalable.
Watson is a completely different take on this need to provide physicians with robust decision support tools. As WellPoint informed us, oncology claims are going up a whooping 23%/yr, far in excess of other diseases and the number of mis-diagnoses is significant. Watson may offer a new approach – we’ll have to wait and see just how successful that new approach will be.
Based on the recent FDA workshop on mobile apps (the second of two days devoted to CDSS) the FDA is working on a guidance document that lays out how they will regulate CDSS. Watson could easily end up regulated as a medical device.