Article Text
Abstract
Background Automated drug dispensing systems (ADDSs) are designed for clinical drug management to support safe and efficient medicines management.
Purpose The objective of this study was to develop a strategy to improve the management of high-risk medicines available in ADDSs and therefore patient safety.
Materials and methods Taking into account clinical impact criteria, a list of high-risk medicines was developed based on recommendations of the ISMP and the Spanish Medicines Agency. Drugs with similar or look-alike names or appearance were included in the list to reduce medicines errors. The strategies to improve safety of the medicines were: to store medicines in Cubie pockets restricting access to only one medicine at a time during the removal and refill process, to store in different drawers, to label drawers with alert stickers and use the ‘tall man’ letter approach to distinguish between potentially dangerous look-alike drug names. An access application was designed to identify potential risk situations, cross matching data from ADDS inventories (drug physical location and type of drawer in each station) with the high-risk medicines list.
Results A total of 1056 medicines included in the ADDS database were reviewed, identifying 154 (15%) high-risk medicines. A total of 651 medicines pairs were identified as having a potentially dangerous similar appearance and 39 look-alike drug names were modified using a ‘tall man’ letter approach. 15 ADDSs station were reviewed using the Access application and a total of 489 potential risk situation were identified. 73% were resolved by labelling drawers with alert stickers and loading medicines in Cubie pockets. 27% were just labelled because they could not be loaded in a Cubie pocket. 157 drugs that looked similar to one another were also identified and loaded in different drawers in each ADDS.
Conclusions Applying security criteria in ADDS management increased medicines safety, reducing potential for medicines errors during the refill and removal process.