Tumblr Icon RSS Icon

Could IBM’s ‘Watson’ Supercomputer Be The Future Of U.S. Health Care Information Technology?

By Sy Mukherjee  

"Could IBM’s ‘Watson’ Supercomputer Be The Future Of U.S. Health Care Information Technology?"

Share:

google plus icon

The quest to improve patient care, maximize medical efficiency, and curb wasteful spending by digitizing Americans’ patient records, insurers’ claims, and providers’ treatment requests just gained a powerful new ally: “Watson,” IBM’s revolutionary data-mining supercomputer that made national waves when it defeated reigning Jeopardy! champion Ken Jennings at his own game.

American Medical News reports that health insurance giant WellPoint has struck up a deal with IBM and Memorial Sloan-Kettering Cancer Center in New York to use the supercomputer — which has spent its post-Jeopardy days amassing and “learning” massive amounts of data about the American health care, insurance, and public health industries — for two pioneer programs to automatically process, review, and pre-authorize medical claims and treatment requests, as well as a third program dubbed “Interactive Care Insights for Oncology” that will “identify individualized treatment options for cancer patients, starting with lung cancer” in order to advise oncologists on the latest and most effective treatment regimens by incorporating up-to-the minute longitudinal medical studies and cancer data into its suggestions.

In an email to ThinkProgress, Cindy Wakefield, a Regional Director for Public Relations at WellPoint, pointed out that the new technology has the potential to have a big impact on the health care industry. “We believe the IBM Watson technology can improve the efficiency and quality of treatment, potentially eliminating unnecessary testing, enhancing the consistency of actions, and accelerating the time to treatment via expedited decision-making processing,” Wakefield explained. “We are continuing to train Watson, and we are teaching Watson by ‘feeding’ it information such as our medical policies and clinical guidelines.”

Using Watson’s technology to automate claims processes could be a potent catalyst for a more efficient American health care industry — which is often bogged down by poor inter-provider communication, incomplete and non-centralized data, and archaic paper records. The supercomputer could also advise providers on the most efficient and appropriate use of treatments based on each individual medical claim, patients’ specific insurance benefits, and patients’ medical histories by analyzing health care data from across the country.

Interestingly, if Watson concludes that a physician or provider’s treatment request is not the most effective one based on a patient’s history and medical benefits, the computer can register its disagreement — but as Wakefield explained to ThinkProgress, it cannot override the provider’s decision or deny treatment requests. Instead, a human nurse would have to review Watson’s alternative suggestion, and then make a judgment call along with the provider on whether or not to comply with it.

That particular Watson capability could end up being quite significant given the initial missteps that hospitals and providers have experienced in instituting new electronic health records. Centralizing health records has long been an essential goal for health care reform advocates aiming to curb redundancy, waste, and miscommunication between providers, insurers, and patients — President Obama even made it a central feature of his landmark health reform law. But there have been several reports of widespread abuse in the nascent systems.

While the use of electronic health records (EHRs) has been rising sharply in recent years, transitioning to the systems has produced several unintended consequences — such as making it easier for some providers to engage in medical billing fraud. “The electronic medical record was meant to make the process of documentation easier, but I think it’s perpetuated copying,” said Daryl Thornton, assistant professor at Case Western Reserve University School of Medicine. That sort of fraudulent behavior can allow providers to bill patients — particularly Medicare beneficiaries — at higher rates than they should by “upcoding” the price of services rendered.

Such unethical billing trends are widespread — and that has Obama Administration officials worried. According to the New York Times, “hospitals that received government incentives to adopt electronic records showed a 47 percent rise in Medicare payments at higher levels from 2006 to 2010, the latest year for which data are available, compared with a 32 percent rise in hospitals that have not received any government incentives.” While the Obama Administration has been cracking down on such fraud, catching every medical billing fraudster is a tall order, especially in the digital age.

That’s where Watson’s new claims-processing and treatment recommendation system could, one day, end up making a difference through its perpetual data collection. While WellPoint’s Wakefield cautioned that “at this time, Watson is not being developed to detect fraud or abuse in the system, or to track fraudulent claims and requests,” all of the information that Watson disseminates and gathers regarding alternative treatments — and whether or not its suggestions are accepted — could end up providing future health care researchers with an invaluable resource for examining potentially fraudulent and wasteful trends in provider practices. And if more providers join forces with advanced and integrated computer systems such as Watson, the future of centralized electronic records could be bright — and even lead to a better, smarter American health care system.

‹ Five Ways The Sequester Will Harm Women

Indiana GOP Scales Back Ultrasound Bill To Include Just One Transvaginal Probe ›

By clicking and submitting a comment I acknowledge the ThinkProgress Privacy Policy and agree to the ThinkProgress Terms of Use. I understand that my comments are also being governed by Facebook, Yahoo, AOL, or Hotmail’s Terms of Use and Privacy Policies as applicable, which can be found here.