A New Verification Process to Reduce Dispensing Error Rate in Community Pharmacie

DESCRIPTION (provided by applicant): The ultimate goal of the proposed research is to develop a closed-loop process to significantly reduce the dispensing error rate for community pharmacies. In the proposed dispensing process, almost all initial medication filling errors will be caught by an independent cross-checking based verification process. In addition, the verification process will be automated so that the tedious and error prone manual check by pharmacists will be eliminated from current clinical pharmacy practice. Each year, medical errors cause more than one million injuries and 44,000 deaths in USA alone. Among them, medication errors are responsible for over 7,000 deaths. Medication dispensing error is a major component of overall medication errors. It is responsible for up to 24% of medication errors in community pharmacies and up to 12.5% of medication errors in hospital outpatient pharmacies. Current medication dispensing process design is based on an error prone open loop structure, where double-check based final checking/verifications are manually operated and lead to the weakest link in the process. An automated closed-loop verification process must be developed to close this gap. Since the part pack dispensing is the dominant dispensing form in community pharmacies, the automated pills identification capability is the key enabling technology for the closed loop dispensing process. In principle, it is feasible to identify any pill based on its imprinted codes, size, shape and color. But it has not been proven that such identification can be reliably operated in automated fashion, which is a crucial requirement for closed- loop dispensing verification process for community pharmacies. In the proposed research, we will demonstrate the feasibility of automated solid dosage medication identification by aiming to: (1) Demonstrate the capability of identifying pills when the size, shape and color measurement including errors due to unavoidable variations in clinical pharmacy practices; and (2) Develop image processing algorithm for engraved imprinted code recognition, which poses a great challenge for imprinted code recognitions. PUBLIC HEALTH RELEVANCE: Each year, medication errors cause more than 7000 deaths in USA alone. This research will result in a much lower dispensing error rate in community pharmacies. The new dispensing verification process will: (1) dramatically reduce the missed-catch rate in final dispensing verification process; (2) improve health care quality by allowing pharmacists to spend more time on patients rather than on tedious and error prone manual medication checking; (3) reduce the healthcare delivery cost by automating the verification portion of the dispensing process.