COMPUTERS AND MEDICINE I: JAMA Takes A Close Look
Today's Journal of the American Medical Association is devoted to various intersections of the worlds of computing and medicine, paying special attention to computerized systems in clinical care, patient-physician communication online, and using computerized medical data to support clinical decision-making.
Catching Errors
A nine-month study of a computer system installed to monitor medication at Harvard's Brigham and Women's Hospital found that the system reduced prescription errors by more than half, putting a "major dent in a problem that causes an estimated 1.3 million patient injuries" and "results in death for three of every 1,000 hospital patients." Serious errors in medication fell from 11 per 1,000 days of hospital care to five per 1,000 patient-days, with actual patient injuries decreasing 17%. The $1.9 million system saves an estimated $480,000 annually in medical injury costs and $10 million in drug and lab test efficiency. Dr. Lucian Leape of the Harvard School of Public Health, author of a JAMA study of the program, said, "I think it's the most important single systems change that hospitals could make." The Boston Globe reports that under the program, Brigham & Women's doctors "are required in every case to enter drug prescriptions and orders for tests through a central computer. The computer offers physicians a menu of drug options and dose ranges, displays up-to-date lab results for the patient, ensures that the doctor's order is legible to pharmacists and nurses, and intercepts orders that are for the wrong drug, an incorrect does, or a drug to which the patient is allergic." The study authors had expected to find a greater decrease in patient injuries, but benefits were offset by a puzzling rise in sedative-related errors for critically ill patients and hindered by doctors that "often neglected to record new allergic reactions" in the system.
Incompatible Systems
In a related JAMA article, a computer system installed at Good Samaritan Regional Medical Center in Phoenix caught nearly 600 potentially fatal drug interactions, 44% of which would have gone unnoticed if the system had not been in place (Knox, 10/21). Dr. Robert Raschke, co-author of the study, said the "rudimentary logic program" that the hospital installed should be utilized by more facilities. "A lot of hospitals have computer systems that would do this, but they are not using them that way," he said. "A lot of hospitals bought one system for their labs, another for demographics, and the systems don't talk to one another" (Van Der Werf, Arizona Republic, 10/21). The authors of the study that evaluated the Good Samaritan system concluded, "Improvements in hospital information systems and increasing utilization of this powerful tool by physicians should have an enormous beneficial impact on the quality of medical care" (Raschke et al, JAMA, 10/21 issue).