Quality advocates back computer orders, despite flaws

Two studies in past months have raised concerns among medical leaders that computerized prescribing systems are failing to prevent many medication-related problems.

A study in the March 9 Journal of the American Medical Association found that a computerized physician order entry system facilitated 22 types of medication error risks at a Pennsylvania teaching hospital. It was followed by a May 23 Archives of Internal Medicine study that found that 27% of adverse drug events at a highly computerized VA medical center in Utah were caused by medication errors.

Many medical leaders have pointed to computerized systems as a key way to decrease medication mistakes and bolster patient safety. But the studies are a warning, physicians said, that some systems are not yet sophisticated enough to reduce errors and that more improvements are needed to make them more effective.

The computerized system "did very little to address the errors in the ordering and monitoring process. These have always been the biggest problem for adverse drug events," said Jonathan Nebeker, MD, lead author of the VA study.

Proponents of computerized medicine said the studies are reason for caution but not cause to give up on computers as significant players in preventing errors.

"We have a lot to learn about how the functional characteristics of these systems can be used for safety. But at the end of the day, we're not going to get safer health care without information technology," said Carolyn Clancy, MD, director of the Agency for Healthcare Research and Quality.

Michael S. Barr, MD, vice president of practice advocacy and improvement for the American College of Physicians, said the studies provided information about the present state of computerized-entry systems and where they need to improve.

"There are a lot of lessons to be learned," Dr. Barr said. "At least people are looking at it and not assuming that it is a magic bullet."

Improvements needed

In the JAMA study, researchers examined use of the CPOE system at the Hospital of the University of Pennsylvania. The study identified 22 situations in which computerized order entry increased the probability of prescribing errors. Examples included fragmented CPOE displays that prevented a coherent view of patient medications and inflexible ordering formats that generated wrong orders.

Researchers recommended that physicians and organizations aggressively examine the technology they are using and fix any problems, because failing to do so could lead to persistent errors.

The Archives of Internal Medicine study looked at 937 admissions to the VA Medical Center in Salt Lake City during a 20-week period in 2000. Researchers found 483 clinically significant adverse drug events, although 73% of them were caused by drug reactions, not medication mistakes.

Even though transcription errors were eliminated, medication errors contributed to 27% of the drug-related problems. Of those errors, 61% occurred in ordering prescriptions and 25% were during monitoring.

Dr. Nebeker said computerized systems were reducing medication-caused problems but that they should be improved to provide more sophisticated approaches on dosing and drug monitoring. "Computers do what they're designed to do, not what you hope they do," he said.

Mark Leavitt, MD, PhD, medical director of the Healthcare Information and Management Systems Society, said effective CPOE systems must include thorough clinical decision support interventions such as alerts and dose suggestions. He said the VA study underscored the urgency for continued development and adoption of health care information technology.

"You can't say, 'Well, we put in a computer, so everything should work better,' " he said. "We have been moving too slowly. The health care industry is behind every other American industry in IT."

David C. Kibbe, MD, director of the Center for Health Information Technology at the American Academy of Family Physicians, said more studies need to examine the merits of CPOE, and physicians should learn more about how to best use computerized systems.

"Information technology just by itself doesn't necessarily lead to better outcomes. It's about how you use it and a whole host of factors."