Failure mode and effect analysis applied to improve the medication management process in a pharmacy of a teaching hospital and a proposal for a simplified rating system

Background Healthcare is not as safe as it should be and medication error remains a significant source of preventable morbidity and mortality among patients. Objectives To present a failure mode and effect analysis (FMEA) of the medication management process in the pharmacy of the largest teaching hospital in Tunisia. Secondly, to examine the validity of a proposed simplified risk rating method by comparing the degree of concordance with the FMEA rating system in classifying failure modes related to the studied process. Methods The FMEA method was applied to the medication management process in the pharmacy for 5 months from January 2020. For the traditional FMEA rating system, failure modes were prioritised according to the risk priority number, which considers severity, occurrence and non-detectability. Failure modes were classified for the traditional method considering three categories: accepted, requiring control and critical. The proposed rating system was based on two indices: the number of parts, which reflected severity, and the number of causes according to the 5M method (manpower, machines, material, methods and medium), which reflected occurrence. Failure modes were classified for the proposed method considering three categories: low, medium and high. Failure modes were independently analysed to determine the degree of agreement in ranking of risk between the two studied methods. Prioritised failure modes were targeted by decisions and solutions aiming to reduce risk and enhance safety. Results Twenty-four failure modes were identified for the six-step process of medication management in a pharmacy (overall criticality=2607). The most critical failure modes were: data error in drugs reception (risk priority number (RPN)=432), break in the cold chain (RPN=320) and non-optimal pharmaceutical analysis (RPN=280). A good agreement was found between the classic FMEA and the proposed rating methods (kappa=0.795). A high correlation was shown between the two scorings (r=0.785). Three failure modes were underestimated by the proposed rating method. Conclusions An FMEA study on the medication management process in a teaching pharmacy showed that FMEA is an effective, proactive risk assessment that enables a better understanding of the studied process. The proposed risk scoring permits a good concordance with the classic method, with the advantage of being fast. Targeting the identified risks will allow integration into a continuous process of improvement and increase patient safety.