Analysing Care Home Medication Errors: A Comparison of the London Protocol and Work Domain Analysis

Abstract

Although methods for investigating safety in complex systems have been widely applied, few studies have made explicit comparisons of two or more methods used to study one system problem. This paper presents a comparison of two methods that were used to analyse medication errors identified in UK care homes namely, The London Protocol and Work Domain Analysis (WDA). The London Protocol is based on an organisational accident causation model and WDA is the first phase of analysis in Cognitive Work Analysis (CWA) that models work domain constraints. Nine medication administration errors identified in seven care homes involving thirty-one residents were analysed. The analyses resulted in the identification of contributory factors to medication errors and changes that would improve system safety. The London Protocol was comparatively easier to use and required less time than the WDA. In The London Protocol, contributory factors were identified from a context-free list of error contributory conditions. From the identified error contributory factors, recommendations to improve system safety were made. However, it was difficult to relate these recommendations to the wider context of the system being studied. The WDA used a contextual model of the care home medication system in the form of an abstraction hierarchy (AH) to analyse errors. Factors contributing to medication errors were specific to the work domain and the resultant recommendations had broad applications, extending beyond the context of the specific errors analysed and were relevant to the whole system. The London Protocol would be a suitable analysis tool if the aim of the analysis was to generate short-term solutions in a short time and with limited financial resources. To make a long-term impact on system safety, the WDA is a useful tool. The AH can be used to analyse the current system, evaluate recommendations for system improvement, design interventions, implement strategies and used throughout the system's life cycle.