September, 12th 2011
For those individuals who missed the first broadcast or news about the contest on TV’s Jeopardy! between past champions Ken Jennings and Brad Rutter and IBM’s Watson computer, I don’t want to give away the outcome before this week’s rebroadcast. Suffice it to say that executives at WellPoint, the largest health insurer in the United States, were sufficiently impressed with Watson’s performance that they’ve made a deal with IBM to develop two of programs that will leverage the machine’s question-and-answer technology. This news broke earlier today.
The big challenge–for WellPoint and IBM–is winning over a healthcare establishment that is rightly cautious and skeptical about new technologies and new approaches to medicine. So the two companies plan on using small-scale tests to prove the effectiveness of the technology before WellPoint considers rolling it out to its network of physicians who treat 34 million individual members nationwide. “This could be game changing. It changes the dynamics of healthcare,” says Dr. Anthony Nguyen, senior vice president of care management at WellPoint. “But people are skeptical and we have to work with them to win over some thought leaders.”
The first of the pilot projects, which will begin by the end of the year, will provide hundreds of WellPoint’s internal clinicians–doctors, nurses and health counselors–with a version of Watson tailor made for healthcare. The Watson technology, which combines natural language processing with machine learning, will be fed a huge volume of medical literature together with WellPoint’s clinical treatment guidelines, and will tap into individual medical records. When WellPoint gets a request for approval of a treatment or procedure, the physicians and nurses considering the request will be able to determine the best treatment not just for the patient’s condition but for that individual patient–and recommend that approach to the physicians, backed with the best scientific data available.
Nguyen hopes that the test will be so successful that he’ll quickly be able to roll the program out to the 3,000 WellPoint clinicians who staff the company’s utilization management unit, and, ultimately to all 5,000 internal clinicians.
The second pilot project, targeted to begin in 2012, will enlist several hundred oncologists who are not WellPoint employees. The doctors will be able to query the Watson-based application to help them decide on the best treatment for individual patients. In addition, Nguyen hopes the application will eventually essentially read over the physicians shoulders when they’re researching a case and offer additional information based on what they’re looking at.
WellPoint chose oncology because diagnoses and treatment are so critical in this specialty. “If we can prove it with the oncology space, which can be life and death, we can then go out and offer it to other specialties,” he says.
Nguyen is by profession a hospitalist, the kind of physician who treats patients in emergency rooms and intensive care units. He says he was thrilled when he first watched Watson in action on Jeopardy! via YouTube clips. “I’m impressed with the capability of Watson to process so much information in such a small amount of time,” he says. “A doctor doesn’t have to perform a plethora of tests. This is like having a room full of the top experts to help him diagnose. It puts him on par with the world’s experts, which gives him greater confidence that he’s doing the right thing for his patients.”
The timing is very good for this initiative. The government’s healthcare reform law provides grants and other assistance for physicians to convert to electronic medical records. So the technology and data are becoming available to help create a revolution in health care. The hurdle now is getting people to use the new intelligence that they might soon have at their fingertips.