Objectives: To explore the combined use of a critical incident database and work domain analysis to understand patient safety issues in a health-care setting.
Method: A retrospective review was conducted of incidents reported to the UK National Reporting and Learning System (NRLS) that involved community pharmacy between April 2005 and August 2010. A work domain analysis of community pharmacy was constructed using observational data from 5 community pharmacies, technical documentation, and a focus group with 6 pharmacists. Reports from the NRLS were mapped onto the model generated by the work domain analysis.
Results: Approximately 14,709 incident reports meeting the selection criteria were retrieved from the NRLS. Descriptive statistical analysis of these reports found that almost all of the incidents involved medication and that the most frequently occurring error types were dose/strength errors, incorrect medication, and incorrect formulation. The work domain analysis identified 4 overall purposes for community pharmacy: business viability, health promotion and clinical services, provision of medication, and use of medication. These purposes were served by lower-order characteristics of the work system (such as the functions, processes and objects). The tasks most frequently implicated in the incident reports were those involving medication storage, assembly, or patient medication records.
Conclusions: Combining the insights from different analytical methods improves understanding of patient safety problems. Incident reporting data can be used to identify general patterns, whereas the work domain analysis can generate information about the contextual factors that surround a critical task.
From the *Centre for Pharmacoepidemiology and Drug Safety Research, Manchester Pharmacy School, and †NIHR Greater Manchester Patient Safety Translational Research Centre, The University of Manchester, Manchester, UK.
Correspondence: Denham L. Phipps, PhD, Room 1.134 Stopford Building, Oxford Road, Manchester M13 9PT, UK (e-mail: Denham.Phipps@manchester.ac.uk).
The authors disclose no conflict of interest.