When it comes to innovative companies disrupting the way we think about health care through artificial intelligence and machine learning, few companies, if any, come to mind before IBM and Watson for Oncology.

But a recent STAT news investigation by Casey Ross and Ike Swetlitz finds that “the supercomputer isn’t living up to the lofty expectations IBM created for it.”

One of the key takeaways from the exhaustive report is that the system still has trouble with just “learning about different forms of cancer.”

In fact, the reports found that rather than Watson figuring out how to treat cancer automatically from its access to data, a “couple of dozen physicians at a single, though highly respected, U.S. hospital, Memorial Sloan Kettering Cancer Center in New York” are inputting their own advice into the system to help to guide its decision-making.

The reporters also noted that IBM “in its rush to bolster flagging revenue, unleashed a product without fully assessing the challenges of deploying it in hospitals globally. While it has emphatically marketed Watson for cancer care, IBM hasn’t published any scientific papers demonstrating how the technology affects physicians and patients.”

For example, its advice about treatments, which is based on American practices, does not take into account different approaches to care in different countries. It also doesn't consider the economics, and whether a procedure or approach is going to get coverage from national or other health care in different nations.

IBM told STAT that Watson for Oncology was making progress “rapidly” and that it would be able to provide treatment guidance for a dozen cancers responsible for about 80 percent of global cancers by year's end.

Health IT Homepage