You may have been part of the audience in 2011 that watched IBM’s Watson trounce the year’s two dominant Jeopardy champs. Or maybe you heard the news of Watson introducing its own BBQ sauce—a recipe developed after analyzing hundreds of thousands of recipes and the basic parameters of human chemistry and psychology.
Exactly what or who is this Watson?
According to IBM, Watson represents a first step into cognitive systems, a new era of computing. The cognitive technology of Watson takes in information and processes it more like a human than a computer. It understands conversational language, generates hypotheses based on patterns and evidence, and learns as it goes. With Watson technology, we would move away from a keyword-based, Google-type search, which directs us to a list of fully realized sites, and towards a conversational question-and-answer type system that leads us to a series of confidence-ranked responses.
So what can we expect next from Watson? How about Doc Watson. At the 2014 Front End of Innovation conference, Gary Robinson, Program Director of IBM Big Data Organization, laid out the very real need for Watson’s services in healthcare.
- According to physician surveys, less than 20% of their day-to-day decisions are based on clinical evidence.
- Every year in the US, more than 1.5 million errors occur in how medicines are prescribed.
- Additionally, a recent New York Times article stated that preventable medical error is the third-leading cause of death in the US, accounting for hundreds of thousands of deaths each year.
At the same time, the amount of medical information that’s readily available on the Web is amassing at a staggering rate, doubling every 5 years. This rapid pace of new information would require a physician to spend 160 hours every week reading clinical papers just to keep up, all in a 168-hour week.
Enter superprocessing, superevaluating Watson.
As part of an EMR system, Watson would work alongside physicians, accessing and filtering hundreds of thousands of articles on any given topic, helping them find the evidence-based answers they need. During an office visit, Watson would “sit” beside the physician, “listen” to symptoms directly from the patient, all the while sorting through its massive clinical database. Based on patient input, Watson would respond very much like a human—probe for additional information, recommend particular tests, and provide a variety of treatment options, each with accompanying confidence scores and clinical trial data support.
In the role of patient, having Watson in my physician’s corner gives me an additional level of comfort. As an advertising professional and pharma marketing person, this clinical-trial-based decision-making process raises some questions.
First, creatively. With Watson providing an evidence-based perspective on efficacy, will we no longer be in the trenches fighting for nuanced expressions of powerful efficacy, redefining efficacy, transformative efficacy? Perhaps instead, we can focus on owning emotional territory around such aspects as a valuable support system, a distinct patient type, a unique aspect of administration, or a particular attitude that will engender confidence and lead to greater adherence.
Second, beyond-the-pill initiatives. As we know, so much goes into a product’s effectiveness that is not reflected in clinical trials. Will Watson’s approach lead to greater focus on designing impactful adherence programs and patient preference studies that can be clinically tested and validated?
Third, regulatory. With very few categories engaging in head-to-head studies, Watson’s promised treatment rankings are a bit tricky. Certainly study designs would have to be ruthlessly consistentized. According to Robinson, Watson’s database is expected to be populated by articles published in medical journals. However in today’s environment, promotional materials are required to stick quite close to FDA-approved labels. Will every published article need to receive FDA approval? Where would the EMEA come in? Would Watson need to take out malpractice insurance?
Fourth, unimagined possibilities. In other industries, such as farming, Watson has uncovered some unexpected benefits. Along with evaluating soil quality and projected rainfall, Watson takes into consideration current and future market conditions to enable it to predict over-supply and unmet needs. Watson then advises on what nutrients can be added to speed up or slow down crop maturation to align with the market.
In healthcare, will Watson be able to predict unmet needs that we aren’t even aware of? Will Watson have the ability to recommend a particular adherence program, an emotion the physician should assume when discussing a particular diagnosis, the types of support conversations that need to take place, the level of urgency that is necessary to change behavior?
If you recall early human vs machine chess battles, Watson’s predecessor, Deep Blue. consistently beat its human competitor. But the good news is, when human and machine worked together, they consistently beat Deep Blue alone. The possibilities of human and machine working together to diagnose, treat, even predict and prevent illness could open up a true revolution in healthcare and in our overall approach to health.
Obviously there are many issues that need to be worked through before Watson starts helping our HCPs make treatment decisions. But the potential benefits are enormous. And we on the marketing/advertising side can’t sit by and wait until these issues are resolved to start exploring the challenges and potential it could open up for us.
Check back here regularly for the latest ideas and inspiration.