The Pharmaceutical Journal
Vol 279 No 7482 p668
15 December 2007

This article
Reprint   Photocopy

  Acrobat Reader

News summary

Authentication systems could reduce error rates

Authentication of dispensed items, using barcodes or radio-frequency identification tags, could have a positive impact on dispensing accuracy, according to researchers who have conducted the most comprehensive study of dispensing errors in UK community pharmacies to date.

The study is published in the International Journal of Pharmacy Practice this month (2007;15:273).

Bryony Dean Franklin, director, and Kara O’Grady, research pharmacist, at the Centre for Medication Safety and Service Quality, Imperial Healthcare NHS Trust, London, aimed to describe the incidence of dispensing errors in UK community pharmacies and to assess the likely impact on error rates of authentication at the point of dispensing.

Ms O’Grady checked dispensed items waiting for collection and compared them with the original prescription in 11 community pharmacies in England and Wales. An expert panel assessed the potential clinical significance of the errors identified.

For each error, the likelihood that it would have been prevented by three types of authentication systems was assessed. The hypothetical systems were stand alone, linked to patient medication records or linked to electronic transfer of prescriptions. The study was funded by Aegate, a company that is developing authentication systems.

The researchers identified a content error in 49 of 2,859 dispensed items (1.7 per cent) and a labelling error in 46 items (1.6 per cent). Of the labelling errors, incorrect instructions and incorrect drug name were most common; of the content errors, those involving too many or too few dose units were most common.

Of the 95 errors identified, 64 were considered minor (67 per cent), 30 were considered moderate (32 per cent) and one severe (1 per cent). The single serious error involved a labelling error.

Using key assumptions about the systems, the researchers predicted that a stand-alone authentication system would have prevented 8.5 of the content errors identified (17 per cent) and one of the labelling errors (2 per cent).

A PMR-linked system would have prevented 12.1 content errors (25 per cent) and 16.5 labelling errors (36 per cent). And an ETP-linked system would have prevented 22 content errors (45 per cent) and 22 labelling errors (48 per cent).

None of the three systems would have prevented the one serious error, say the researchers. But they predict that 22–60 per cent of the moderate errors would have been prevented depending on the system used.

“An experimental study is now needed to find out whether or not these predictions are borne out in practice,” they conclude.