Watson, the IBM supercomputer that recently generated headlines by beating top Jeopardy players, is being groomed to advise doctors about diagnoses and treatments. But even if Watson delivers more accurate answers faster than other decision support tools, there's no guarantee that physicians will heed its advice.
When it's ready, perhaps two years from now, Watson will analyze the answers to questions about a patient's symptoms and medical history and suggest the most likely diagnoses. It will use the same approach to recommend treatments.
The source material for Watson's new healthcare knowledge will include medical textbooks and journals, data from electronic health records and personal blogs about illnesses. The goal is to provide Watson with a vast fund of constantly updated information that is difficult--if not impossible--for doctors to keep in their heads.
Physicians eventually might connect to Watson's database by speaking into a hand-held device, using speech-recognition software and cloud computing, an IBM executive told the Associated Press. But the history of this kind of decision support does not suggest that many doctors are willing to rely on computer logic, even as a backup.
For example, a system named Isabel--which suggests diagnostic alternatives--has been around for some time, and is used in some hospitals. Yet, it hasn't taken the world by storm. The company that makes Isabel says that it will include the right diagnosis in a list of possible diagnoses 90 percent of the time, but doctors still have to decide which one is correct.
"Clinical-decision support is best for prompts and reminders about services due, and has never been that effective for practicing clinicians in terms of suggesting a diagnosis," commented Peter Basch, MD, an internist and informatics expert at MedStar Health in Washington, D.C., in a 2007 interview with Medical Economics. "Eventually, it may get there, but we're not there yet."