The slippery slope for patients, nurses, and doctors posed by robots in healthcare

By: RoseAnn DeMoro, Executive Director of National Nurses United

For patients needing dialysis or care for acute kidney failure, there’s a new doctor in the nephrology ward at St. Joseph Hospital in Eureka, Calif. Meet the doc on a stick.

It’s not a scene from “Star Trek” or the latest X-box video game. And, like the smooth-sounding, but ominous “Hal” computer running the spaceship in “2001: A Space Odyssey,” those side effects might be a killer.

Eureka RNs have noticed an immediate impact. Dr. “doctor on the lamp post” can sort of see the patient, but can’t offer hands-on care, has trouble getting around, and doesn’t hear so well. So it’s up to the RN to babysit the machine, wheel it around, and clinically assist the remote doctor with patient communication and physical assessments. (The company name “InTouch Health” should be listed in the dictionary somewhere between “deceptive” and “Are you kidding me with this?”)

For RNs, it means more time away from other patients on often already short-staffed units. Involvement with this process at the patient’s bedside will take considerable time, “during which a nurse’s other assigned patients have no nurse available to meet their needs,” says St. Joseph RN Katherine Donahue

Further, “the robot doctor is very impersonal for the patient. It undermines the hands-on ability for the doctor with the patient, and if the electronic equipment malfunctions it can compromise the medical record,” Donahue added.

Computer-driven errors from the trillions of dollars the healthcare industry is spending to develop, market, and profiteer from in the not-so-brave new world of medical technology. That’s not a concern, right?

Or as Hal the computer states right before killing the human passengers and pilots, he is “foolproof and incapable of error.”

Problems with the machines might be a surprise to the consultants and industry executives, not to mention the politicians who enacted financial incentives to promote rapid expansion of medical technology as key components of the Affordable Care Act, and the 2009 budget stimulus bill before that.

But nurses have long been aware of the downsides for quality care and human healthcare employment, deriving from computerized diagnostic and prognostic protocols, and other skill-debasing and displacing mechanical overlords.

At the dawn of the healthcare restructuring wave of the early 1990s, the California Nurses Association’s research department was sounding the first warnings of the implications of turning nurses and other hospital personnel and the nursing process itself into digital bits, of how individual patients need individualized care, and of how the machines can and do actually fail.

We cited the example, told in a PBS special, “The Thinking Machine,” of artificial intelligence researcher Doug Lenat describing his rusting 1980 Chevy to a skin disease diagnostic system as a lark. It concluded that the patient had measles.

For a more recent anecdote, consider Hal’s great-grandfather, Watson, developed by an IBM research team and rolled out with great fanfare on the quiz show “Jeopardy.”

Viewers watched in awe as Watson steamrolled its human competitors, until they got to the “Final Jeopardy” question: “What U.S. city’s largest airport was named for a World War II hero; its second largest, for a World War II battle.” Watson answered promptly, “Toronto,” (uh, that’s in Canada), while the overmatched humans got it right, “Chicago.”

That story made lots of news. But less recalled is follow-up explanation from the lead IBM research investigator who concluded that the “category names were tricky,” only minimally suggesting an “expected answer,” and “the way the language was parsed provided an advantage for the humans and a disadvantage for Watson,” as reported by Steve Hamm in a blog for “Building a Smarter Planet.”

Or, in a nutshell, all the problems nurses, CNA and NNU have reported with computerized protocols – substituting digitalized systems that don’t necessarily respond to the diverse, complex health problems faced by real patients for human professional judgment, the ability to think and analyze.

Watson went on to a better life in, you guessed it, healthcare. Last month IBM announced that Watson will be making utilization management decisions for lung cancer treatment at Memorial Sloan–Kettering Cancer Center in conjunction with insurance giant WellPoint. IBM Watson’s business chief boasted that 90 percent of nurses in the field who use Watson will follow its guidance, reports Forbes magazine.

While RNs have long been the Cassandras on technology, many doctors have been late to the game and been in the forefront of cheerleading for how the Watsons and telemedicine practitioners are the solution for medical errors, improving overall quality, and cutting healthcare costs.

A physician walking the rounds in a hospital might only see five or 10 patients a day. Put that same doctor at a desk with a computer monitor miles or continents away, and they might see 300 patients a day. How many doctors do you think the CEOs will need in this future?

One last example from the retail grocery industry. When product scanners were introduced, a lot of checkers thought it would make their jobs easier. It did, required fewer of them and downsized their skills as well.

Today, walk into any Safeway and notice the growing number of fully automated check-out registers with no live workers and no ability to respond to individual problems – in other words, just another grocery commodity. That’s what the doc behind the doc on the stick might give a little more thought to.