COMMUNITY PRACTICE

New study sheds light on medication errors

Community pharmacies could help reduce medication errors by simply providing better illumination of their facilities, according to a new study by Auburn University School of Pharmacy.

Installing brighter lights, cutting down on the decibels, and having more employees on hand are ways to reduce medication errors, according to Elizabeth Flynn, Ph.D., R.Ph., associate research professor, center of pharmacy operations and design, department of pharmacy care systems, at the pharmacy school. She recently presented the findings at the annual meeting of the Human Factors and Ergonomics Society in Baltimore.

An error rate of 1.56% per 100 scripts was recorded among the 50 community pharmacies situated in six large cities that participated in the Auburn study. Half of the pharmacies were chain drugstores, about 20% were independents, and the rest were health-system outpatient or health maintenance organization (HMO) pharmacies. Among the 5,784 scripts examined, 91 contained errors. Extrapolating that rate to the estimated three billion prescriptions filled annually, there are about 60 million medication errors per year.

Flynn said that pharmacists trained as observers spent one day in each pharmacy—or long enough to watch the staff dispense 100 scripts, which included new Rxs, refills, and scripts in the will-call area. The trained R.Ph.s tracked real errors—those the staff failed to detect by the time the Rx was intercepted by the investigating pharmacist before it was given to the patient—and near errors, which were those caught by the staff before they were checked by the pharmacist investigator. The study found 74 near errors.

Lighting levels above 94 foot- candles (a measurement of light quantity) were found to help prevent and detect errors and to reduce near errors. The findings of the observers confirm an earlier Auburn study concluding that a lighting level of 146 foot-candles can lead to significant reductions in dispensing errors, Flynn said.

When it came to noise in the pharmacy, less was better. When sound rose above the 75-decibel level, it was harder for pharmacy staffers to detect and prevent errors. "When the source of the sound was a radio or TV, the staff detected a significantly higher proportion of errors, but when that sound got louder (above 75 decibels), the staff was less likely to catch the errors," Flynn told Drug Topics. "For comparison, a normal conversation is about 50 decibels," she said.

Bar-code scanning had a significant impact on the detection and prevention of medication errors in the eight study pharmacies using the technology. More than 75% of errors of content, such as wrong drug or wrong strength, occurred in pharmacies using manual inspection of Rxs. But Flynn cautioned that bar-coding technology is only as good as the people using it. For example, someone's failure to scan a script at some point in the dispensing process could lead to an error.

The Auburn researchers hope to bring the field data into the laboratory setting for further study and to finalize their recommendations about ways pharmacies can reduce medication errors, said Flynn. "For example, we'll try to narrow down what lighting level is best and look at bar-code systems," she added.

The Auburn medication error study was a collaborative effort between the pharmacy school and the department of systems and industrial engineering. In addition to Flynn, the team included Kenneth N. Barker, Brian J. Carnahan, Nathan T. Dorris, and Grady T. Holman. They hope to publish full results of the study next spring.

Carol Ukens

Carol Ukens. New study sheds light on medication errors.

Drug Topics

2002;21:33.