Probabilistic risk analysis of medication error

Over the years, the United States has spent billions of dollars in its quest to improve the quality and safety of health care through the development of new drugs and technologies. Using the probabilistic risk analysis of medication error process, we demonstrate the application of Bayesian Causal Network Model to assess the probabilities of occurrence of rare events related to medication errors. This article summarizes the methodology involved in the process.

A convenience sample of annual incident reports from a Northeast acute care community hospital was used for the study. Importance sampling was used to improve the accuracy of estimates of the rare events so that we can better understand the relationship of causes within the reported events.

The Bayesian Causal Network Model provided contextual maps of the behaviors and errors that lead to medication delivery process failures, including unanticipated risks associated with actual errors as well as near misses and common deviations from standard procedures and policies. Health care administrators, clinicians, regulators, and educators can prospectively identify and prioritize risk reduction interventions using the Bayesian Causal Network Model.

The Bayesian Causal Network Model can identify behavioral and systemic factors that can enhance or reduce the risk of wrong drug, wrong frequency, wrong dose, omitted dose, drug interactions, and wrong patient medication errors in hospitals.