Document resource
Objective To evaluate the safety of implementing structured medication optimisation reviews (SMORs) for patients with atrial fibrillation (AF) prescribed direct oral anticoagulants (DOACs). SMORs aimed to improve quality of care and facilitate large-scale alignment with national prescribing guidance (to switch patients onto edoxaban).Intervention Complex intervention including a SMOR embedded within primary care electronic patient records alongside clinical decision support tools.Design Doubly robust difference-in-difference analysis using linked electronic health records, comparing changes in prescribing and bleeding admissions in patients undergoing SMOR with those in patients not reviewed.Setting Sefton (intervention area) and Liverpool (comparator area) in the Northwest of England.Participants All patients with AF prescribed a DOAC in 2022.Main outcomes and measures Proportion of patients prescribed apixaban, proportion of patients prescribed edoxaban and rate of emergency hospital admissions for bleeding-related events.Results The proportion of patients in Sefton prescribed edoxaban increased from 19% to 35%; 13% (95% CI 11% to 14%) of the increase was associated with the SMOR. There was an 11-percentage point decrease in patients prescribed apixaban (95% CI −12% to −10%). Undergoing review was associated with a non-significant reduction in the risk of bleeding admissions (eight fewer admissions per 1000 people reviewed per year; 95% CI −22 to 6).Conclusions SMORs can be delivered at scale and used to switch medications for a large proportion of people. There was no evidence of an increased risk of admissions for bleeding complications in AF patients following a large-scale switch from apixaban to edoxaban supported by SMORs. Such reviews could improve prescribing quality and patient safety by ensuring patients are on the most appropriate dose and choice of DOAC and lead to cost savings to health services (by facilitating a switch to a better value product) while not increasing risks for patients.