Reducing Medical Errors Information technology can help

by David Orenstein freelance writer

Two years after the Institute of Medicine’s famous estimate that as many as 100,000 Americans die each year from medical errors, many medical professionals are tired of people incessantly citing the report or can recite parts of it like a mantra. While the report heightened awareness of medical errors, many researchers are just getting started on ways to prevent errors. A lot of them have high hopes that information technology will help.

In a recent paper, “Reducing the Frequency of Errors in Medicine Using Information Technology” (J Am Med Inform Assoc. 2001;8:299-308), Bates et al asserts, “IT is not a panacea [but] the informatics community should make it a high priority to assess the effects of IT on patient safety.”

The authors mention a few ways IT already helps. At Cedars-Sinai Medical Center in Los Angeles, physicians carry pagers that alert them with on-screen messages and beeping when a wide variety of patient conditions (such as hematocrit, white blood cell count, medication levels) get out of whack. The authors cite two studies in which the quickness and reach of such automatic alerting systems have substantially sped up correction of these kinds of problems. In one case, an e-mail alerting system at an institution resulted in hospital staff correcting kidney medication levels an average of 21 hours earlier than before.

Another area the authors say IT can help is in prescribing and ordering medications. One author, Dr David W. Bates of Harvard Medical School, found in a 1998 controlled trial that computerized physician order entry systems reduced serious medication errors by 55 percent. The paper cites another study showing similar benefits. Well-designed computerized systems can make prescriptions legible (because they are not handwritten) and catch dangerous drug interactions or high doses. The group recommends computer-based prescribing (from whichever vendor you choose).

Of course IT may be able to help in other ways. In fact on Oct 11 the US Department of Health and Human Services awarded more than 94 grants, totaling $50 million, in a broad effort to research and reduce medical errors. Of these grants, 22 specifically are meant to discover ways that IT can help. Many fund studies of computer-based prescribing, but others will address areas of care ranging from surgery, to Web-based communicaton with elderly and underprivileged patients, to wireless collection of data from medical monitoring equipment.

In one project, Montefiore Medical Center in New York City is getting $1.3 million over the next three years to develop a virtual reality simulation of high-risk surgery to train surgeons. In another, Intelligent Automation, Inc, of Rockville, Md, is getting $100,000 to develop a system including smart cards and fingerprint reading to help emergency staff get patient information from the Internet. A third grant gives $356,000 to Columbia University researchers who propose to discover medical errors by analyzing the narrative notes of physicians with language processing software. (All the grants are described at www.ahrq.gov/qual/newgrants/it.htm.)

No one, especially Bates and his fellow authors, suggests that medical errors are mainly IT’s problem to solve. In fact their recommendations include stringent testing and monitoring of medical IT to ensure that it is not causing errors or failing to catch them. After all, before we put IT in place to reduce errors, we probably should remember that medical standby: First, do no harm.

David Orenstein is a technology and business writer in Silicon Valley. If you'd like to learn more about a certain computer topic, e-mail him at davealli@attbi.com.