US study finds limits to bar code security

Bar coding patients and medicines does not result in a reduction in errors or improved safety, according to the surprising results of a survey carried out at five US hospitals and checks on about half a million bar code scans by a team from the University of Pennsylvania School of Medicine. The report, which was published in the Journal of the American Medical Informatics Association, is claimed to be the first of its kind.

The safety procedure is supposed to be operated by a health professional who scans the bar code on a medication as well as a similar code generated for each individual patient before each drug or treatment dose is given. However, the research team found that a number of practical complications can prevent the procedures from being properly carried out.

For example, restrictions on the use of electronic equipment in some parts of hospitals due to interference means that the scan cannot take place at the point of drug delivery (Marketletter June 30). Another problem is where the bar code for the patient is not physically located on the patient, for example on a wrist bracelet. In 4.2% of cases, nurses overrode the scan, by typing in a serial number manually - which creates opportunities for error - because the bar code label was unreadable.