No
Purpose - The purpose of this article is to examine a sample of paper-based incident reports concerning drug incidents to assess the utility of a reporting system. Design/methodology/approach - A 50 per cent random sample of drug-related incident reports between 1999 and 2003 (n=1,253) was reviewed. Details of the incident including error type and contributory factors were identified, as was status of the reporter. Content analysis of the free text established whether the data provided could promote medication safety and organisational learning. Findings The paper finds that all definitive drug errors (n=991) allowed an error type to be identified, but 276 (27.8 per cent) did not include the contributory factor(s) involved. Content analysis of the errors demonstrated an inconsistent level of completeness, and circumstances, causation and action taken were not always logically related. Inter-rater reliability scores were varied. There was sometimes a significant focus on the actions of one individual in comparison to other factors. Research limitations/implications - Incident reports can be biased by psychological phenomena, and may not be representative of the parent organisation other than those who report. This study was carried out in a single health care organisation and generalisability may be questioned. Practical implications - How health professionals interpret drug errors and their reporting could be improved. Reporting can be further developed by reference to taxonomies, but their validity should be considered. Incident report analysis can provide an insight into the competence of individual reporters and the organisation's approach to risk management. Originality/value - This paper highlights the various data that can be captured from drug error reports but also their shortfalls which include: superficial content, incoherence; and according to professional group - varied reporting rates and an inclination to target individuals.