Introduction: The use of warfarin to prevent ischemic stroke in patients with atrial fibrillation/flutter (AF) requires routine international normalized ratio monitoring and frequent dose adjustments to optimize time in therapeutic range (TTR). While TTR are generally low in usual care, anticoagulation services are associated with improved TTR and clinical outcomes. The direct oral anticoagulants (DOAC) are preferred to warfarin for AF as randomized controlled trial (RCT) data have shown they are as effective in preventing stroke with a lower risk of bleeding, particularly intracranial hemorrhage, compared to warfarin. However, the mean TTR for warfarin-treated patients enrolled in these RCT were relatively modest (55-65%); thus, it is unclear if clinical outcomes would differ if patients in the RCT control groups had been managed by an anticoagulation service. The purpose of this study was to evaluate the real-world effectiveness and safety of DOAC compared to warfarin among patients with AF followed by an anticoagulation service. Methods: This was a retrospective, longitudinal cohort study conducted at Kaiser Permanente Colorado, an integrated healthcare delivery system with a pharmacist-managed anticoagulation service. Adult patients diagnosed with AF, newly initiated on a DOAC or warfarin, and enrolled in the service between 11/01/2015 and 12/31/2018 were followed until anticoagulant switch/discontinuation, a study outcome, membership termination, or 01/05/2019, whichever came first. The index date was the date of service enrollment. TTR and percent of days covered (PDC, a measure of medication adherence) were calculated for the warfarin and DOAC patients, respectively. The primary outcome was a composite of all-cause mortality, clinically-relevant bleeding, ischemic stroke, or systemic embolism. Secondary outcomes included the individual components of the primary outcome as well as follow-up healthcare utilization. Multivariate Cox proportional hazards (reported as adjusted hazard ratio [aHR] with 95% confidence interval [CI]) and negative binomial (reported as adjusted incidence rate ratio [aIRR] with 95% CI) regression analyses were performed to assess differences between groups. All aspects of this study were reviewed and approved by the KPCO Institutional Review Board before data collection. Results: There were 2928 patients included with 1537 and 1391 warfarin and DOAC patients, respectively. The cohort was primarily older, male, and white with moderate burden of chronic disease and baseline healthcare utilization. Warfarin patients had a mean TTR of 72% and were more likely to be female and Medicaid beneficiaries and have had dementia, depression, and pulmonary disease comorbidities (all p<0.05). DOAC patients had a mean PDC of 83% and were more likely to have had a baseline hospitalization (p<0.05). With adjustment for age, sex, Charlson comorbidity index, race, Medicaid status, baseline healthcare utilization, dementia, depression and pulmonary disease comorbidities, patients on a DOAC had an associated lower risk of the primary composite outcome (aHR 0.75, 95% CI 0.62 - 0.91), any clinically-relevant bleeding (aHR 0.68, 95% CI 0.51 - 0.92), gastrointestinal bleeding (aHR 0.62, 95% CI 0.46 - 0.54), systemic embolism (aHR 0.59, 95% CI 0.36 - 0.95), and follow-up hospitalization (aIRR 0.76, 95% CI 0.65 - 0.89) (Table). DOAC patients had equivalent risks of ischemic stroke, all-cause mortality, intracerebral bleeding, and follow-up emergency department visits (all p>0.05). Conclusions: This retrospective analysis of a large sample of real-world patients with AF managed by an anticoagulation service identified that DOAC use was associated with improved safety compared to warfarin. Despite high-quality warfarin management, use of DOAC was associated with lower risks of clinically-relevant bleeding, systemic embolism, and hospitalization compared to patients receiving warfarin for AF which reinforce the application of RCT findings to health systems with available anticoagulation management services. Future real-world studies should be conducted to confirm our findings with multiple healthcare delivery systems, a larger sample size, and assessments of the individual DOAC versus warfarin to provide further evidence of the effectiveness and safety of DOAC. Figure 1View largeDownload PPTFigure 1View largeDownload PPT Close modal