Paediatric Prescribing Error: An Audit Tool

Through critical incident reporting, we were aware that prescribing practice in the NICU was sub-optimal. We hoped to produce a tool that would allow us to assess the volume and nature of prescribing errors that were occurring in the NICU.

Aims

• To facilitate detection of clinically significant prescribing errors in paediatric settings.
• To offer a useful means of categorisation of errors, to facilitate presentation of data, and to guide potential interventions.

Description

The tool comprises two parts. The first is a classification of errors, divided into categories. The list of errors is based on a published article by Ghaleb et al which established an expert consensus definition of prescribing error. This approach has the advantage that detected errors based on this classification have been agreed as clinically significant. Use of this classification would encourage standardisation in quality improvement work relating to paediatric prescribing error. We divided the errors into categories based on their type, which makes them easier to present. We believe the categories may correspond to some underlying factors in why particular errors occur.

The second part is an audit tool designed to detected the errors included in the classification. This allows for the details of the error to be described and for the error to be coded.

Implementation

We used the tool to collect and classify prescribing errors in the NICU on a weekly basis for approximately 10 weeks. We found that it was an effective way to detect errors, with rate of error corresponding to what we would have expected from the literature.

The classification proved useful in grouping together similar errors and directing interventions; we found the highest rate of error in category B, followed by category A. However, category A errors corresponded to a greater potential for harm.

Suggestions for Further Implementation

The tool is free to be adapted and modified for use in individual practice areas. It would be interesting and useful to introduce interventions and assess whether these reduced errors in different categories in different ways.

The main challenge we encountered was in the time-intensive nature of data collection by auditing prescription sheets, although this leads to a more complete and accurate sense of prescribing errors than use of other methods, such as using critical incident reporting alone. We feel audit work in this area would be best undertaken by a multidisciplinary team, including medical and pharmacy staff.

The tool was approved for use within the NICU by the consultant responsible for overseeing the project. It has not been formally validated, although we have used the tool successfully within our unit, as described above. We feel that the tool is readily applicable to other settings, whether in paediatrics or neonates. The tool may be modified for use in specific settings, and we have included explanatory notes to facilitate this. We plan to update the tool based on our personal experience of using it, and will endeavour to respond to any feedback from others.

Tool lead: Richard Conn, Academic Clinical Fellow; Clifford Mayes, Consultant Neonatologist

Organisation: Belfast Health & Social Care Trust