| Identifying the scope and determinants of misfills of ambulatory care prescriptions by community pharmacies is a daunting task, as it has previously required manual inspections on a prescription-by-prescription basis. Thus, such explorations have been rare, with the largest scale project to date reporting results on fewer than ten thousand prescriptions.;Practice-based research techniques taking advantage of prescription and fulfillment data contained within the Electronic Health Records (EHR) of clinical organizations allow for assessment at far greater scale. Our pilot study of five practice organizations, limited to only 3 categories of medications, Lipid Regulators, AntiHypertensives, and Opioid Analgesics examined over two million fulfillments of prescriptions generated through three distinct EHR products, AllScripts Enterprise, AllScripts Professional, and SmartClinic, from ambulatory care organizations of various sizes in five distinct markets.;Determinants looked at, Category of Drug, Availability of Drug as Generic Product, EHR product utilized, number of active ingredients in product, each appear to have a highly significant effect on rate of mismatch between prescriptions and fulfillments.;Due to some limitations in methods utilized, these numbers should be interpreted not as absolute rates, but rather as indications of where further investigation is needed. Mismatches varied by EHR product from 0.01% to 0.10%. The relationship between generic availability and putative error was not clear. Products with a single active ingredient showed a rate of 0.036% while multiple ingredient drugs had a rate of 0.102%.;Perhaps most interesting was the 0.121% rate for Opioid Analgesics, which, when compared to the rates of 0.041% for Lipid Regulators and 0.033 for AntiHypertensives, may suggest a possibility of widespread fraud.;This exploration has shown that research into medication fulfillment errors through extraction of prescription and fulfillment records resident in EHRs is practical and presents opportunities for exploration of relationships between prescriptions and their fulfillment for purposes of research and quality assurance. With proper masking of patient identity, it could also be used by regulatory and enforcement agencies interested in patterns of discrepancies. |
