A prospective hazard and improvement analytic approach to predicting the effectiveness of medication error interventions

Abstract

Medication errors are an important problem for the UK National Health Service (NHS). The aim of this study was to implement a novel quantitative modelling method to predict rates of preventable adverse drug events (ADEs) and identify interventions with the greatest potential for reducing the burden of medication errors in secondary care. A generic model structure was developed to describe the medication process in secondary care. The model followed pathways from error points through to the outcomes of undetected errors. The model was populated using quantitative estimates and calibration methods to describe the incidence and impacts of medication errors. The effectiveness of potential interventions was estimated by describing the impact of the interventions at different stages of the medication process. The model predicts the range of preventable adverse drug events that occur annually in a 400-bed hospital in the UK to be between 200 and 700. Of the interventions evaluated, computerised physician order entry systems and increased numbers of ward pharmacists are predicted to have the greatest impact on the number of preventable ADEs. The analysis provides a relative analysis of the interventions, and indicates priorities for research allocation decisions. The model highlights the complexity of the relationship between medication errors and adverse events, and the extreme attention to detail required in the development of interventions, and in their evaluation.