UIC SPEECH EXPERT SOLVING MEDICATION ERRORS
(DECEMBER 1997)
University of Illinois at Chicago researchers are using computer models and psychological testing to help reduce the number of medication errors caused by look-alike and sound-alike medication names. They recently demonstrated that their computer model can distinguish between pairs of medication names reported to have been involved in medication errors and medication names randomly selected from a large data base. With new grants from several government agencies, the researchers are well on their way to predicting memory errors in humans based on look-alike and sound-alike medication names.
Experts estimate that look-alike and sound-alike medication names play a part in roughly one-quarter of the hundreds of thousands of medication errors made by health professionals in the United States each year. Clonodine and clonopin, for example, are names that look and sound alike. Published reports of errors made by health care practitioners show that when a patient received the former instead of the latter, the results were severe abdominal pain, nausea, palpitations, dizziness and flushing. Those same reports show that a patient died when a provider confused the medicines vinblastine and vincristine.
Medication errors not only cause inconvenience, suffering and death, but also have high economic costs. Putting safeguards in place to prevent medication errors is the goal of the UIC team's research, said lead researcher Bruce Lambert.
Lambert, assistant professor of pharmacy administration, holds a doctorate in speech communication and may be the only person in academia using both computer models and psychological testing to predict look-alike and sound-alike medication errors. The tests Lambert recently designed will measure the effect of phonological similarity on the memories of pharmacist test subjects. The researchers will use lists of similar looking and sounding medication names to replicate the pharmacy environment.
"We want to bring more scientific rigor to the name approval process," Lambert said. The FDA, United States Adopted Names Council and the United States Patent and Trademark Office are the primary agencies responsible for approving trademarks and established names for new drug products. To assess the confusion potential of new trademarks and the more than 15,000 existing medication names (in the United States alone), they rely primarily on a panel of experts completing a variety of rating scales.
The Labeling and Nomenclature Committee of the FDA already is using Lambert's computer model on a trial basis and the FDA and United States Pharmacopeia expect to adopt Lambert's methods on a long-term basis.
(Editor: For more information, call Jody Oesterreicher, 312/996-8277; [email protected])
RSS