The U.S. is going to spend in short order more than $30 billion on bringing automation to the health care industry. Listening to the health care experts at the XSITE 2009 event last week, I realized how much of this money is going to go into automating decision-making by health care professionals. Let the buyer beware! Today’s Washington Post has a great article on the perils of automation (in the context of the Metrorail crash), quoting John D. Lee, a professor of industrial and systems engineering at the University of Wisconsin at Madison:
“The better you make the automation, the more difficult it is to guard against these catastrophic failures in the future, because the automation becomes more and more powerful, and you rely on it more and more.”
This warning should accompany every check the Federal government writes for automating, I mean stimulating, health care. As ON magazine’s columnist Jim campy observed in the latest issue: “When I asked one major hospital how they were able to consistently reduce an astounding “near-miss” rate to a much lower prescription error rate, they responded that the quick action of competent, observant nurses allowed them to close the gap. Fortunately, a trained nurse knows when a drug or dosage is wrong for a patient. “
Technology can reduce unnecessary waste and may even help improve decision-making. But let’s not rely too much (or only) on technology to make decisions about our health.

July 1st, 2009 at 8:31 pm
What you say is true but the current approach of allowing medical staff to make decisions based on what they remember (by far the most prevalent approach) and to apply the practices they think best (not evidence based medicine) is pretty badly flawed. Automating decisions like dosing and drug interactions (if done in a way that brings medical staff into the management of the system rather than just letting IT code it) can save lives and costs.
In the Metrorail crash we should remember that drivers have crashed trains because they fell asleep or decided to text instead of watching - at least computer systems don’t do that…
JT