IBM sale of Watson Health reveals healthcare’s AI conflict

The reported exploration to sell Watson Health, the AI-driven health business of IBM, underscores deep challenges with technology in solving healthcare problems, according to the Wall Street Journal.

WSJ, which broke the anonymously sourced news that IBM might be looking to unload Watson Health, notes that massive datasets likely can’t solve healthcare’s biggest issues, as some tech and business solutions companies have touted. When it comes to complex medical needs, AI “can be difficult to apply,” the paper points out.

That’s because healthcare professionals may not have all the answers, with gaps in knowledge for some health conditions. AI is trained on data to help healthcare professionals make diagnoses and care decisions, and if the data is imperfect, as it often is with complex health conditions, the technology solutions are imperfect too. 

Further, it is difficult––or even impossible––to compute patient preferences. For example, the behavior of a young adult who refuses a COVID-19 vaccine isn’t predictable, and therefore computable, according to the report. That issue just can’t be solved with AI. 

However, there are a number of healthcare areas where AI can solve problems and improve efficiencies. Helping doctors diagnose patients is one area, as AI can pick up repetitive patterns in images to identify and classify MRI or X-rays. Health insurance giant Humana recently teamed up with IBM Watson Health to introduce a conversational AI solution for its employer group members, enabling an AI assistant to answer questions for customers. 

Neural networks and wearable devices are also gaining traction as successful ways to implement AI in healthcare, collecting vast amounts of data that can be used to identify early health risks and potential interventions.

As AI continues to expand in healthcare, its uses will continue to develop. But solving healthcare’s biggest and most complex challenges, like curing cancer, may still be a long way off. 

“I believe that we’re many years away from AI products that really positively impact clinical care for many patients,” Bob Kocher, MD, partner at Venrock and a former White House health adviser under President Barack Obama, tells the WSJ.

Amy Baxter

Amy joined TriMed Media as a Senior Writer for HealthExec after covering home care for three years. When not writing about all things healthcare, she fulfills her lifelong dream of becoming a pirate by sailing in regattas and enjoying rum. Fun fact: she sailed 333 miles across Lake Michigan in the Chicago Yacht Club "Race to Mackinac."

Around the web

The final list also included diabetes drugs sold by Boehringer Ingelheim and Merck. The first round of drug price negotiations reduced the Medicare prices for 10 popular drugs by up to 79%. 

HHS has thought through the ways AI can and should become an integral part of healthcare, human services and public health. Last Friday—possibly just days ahead of seating a new secretary—the agency released a detailed plan for getting there from here.

Philips is recalling the software associated with its Mobile Cardiac Outpatient Telemetry devices after certain high-risk ECG events were never routed to trained cardiology technicians as intended. The issue, which lasted for two years, has been linked to more than 100 injuries.