Application of the Bow-Tie model in medication safety risk analysis: consecutive experience in two hospitals in the Netherlands.

BACKGROUND: To improve medication safety effectively, one should systematically analyse and assess the risks for medication errors and determine the possible causes. So far, no risk-analysis instrument exists in healthcare that can be used to analyse and visualize risks, causes and consequences of potential adverse events in a prospective manner. In high-risk industries such as petrochemistry and aviation, the Bow-Tie model is frequently used. This model combines causes, errors, preventive and recovery measures, and consequences in one model and gives insight into the magnitude and causes of existing safety risks. The aim of our project was to study the usefulness of the Bow-Tie model in the hospital setting for prospective analysis of risks in the medication process in order to develop a practicable method.
METHODS: The model was first adapted to the clinical setting. Thereafter, the risk-analysis model was applied in a large tertiary teaching hospital in multidisciplinary sessions. The sessions and risk-analysis method were evaluated on the following aspects: applicability, comprehensibility, creation of awareness in and motivation of participants, and the capability of the 'system approach' (the approach taken by the Bow-Tie model, which focuses on the conditions under which individuals work and tries to build defences to avert errors or mitigate their effects, in contrast to a 'person approach', which focuses on errors of individuals, blaming them for forgetfulness, inattention etc.). Based on this evaluation, the risk analysis method was adjusted and consecutively applied in a general teaching hospital. After evaluation of the sessions in the second hospital a recommended method for risk analysis with the Bow-Tie model was defined.
RESULTS: The risk-analysis method with the Bow-Tie model in the first hospital gave insight into many medication safety-related risks. However, the method was insufficient on comprehensibility and on the creation of awareness and motivation owing to a great number of determined risks which made thorough analysis, drawing of Bow-Ties and prioritizing difficult. The adjusted method in the second hospital focused more on the in-depth analysis of a small number of important safety issues of a department with specific attention for underlying causes. This approach was considered better in applicability, comprehensibility and the creation of awareness. Furthermore, by analyzing underlying causes, more attention could be paid to latent conditions (which can translate into error-provoking conditions) within the system.
CONCLUSION: We found the Bow-Tie to be an appropriate model for prospective risk analysis of medication safety in a hospital. By applying the model in two hospitals consecutively we developed a feasible method for risk-analysis sessions. Key factors of this recommended method are a focus on the prioritized selection of safety issues and specific attention to latent conditions within the system by analysing these safety issues in depth to the root causes with the help of the Bow-Tie model.