Abstract
Background
Oral anticoagulants are the cornerstone for the management of atrial fibrillation (AF) to reduce cardioembolic stroke
Randomized controlled trials of anticoagulants have shown non-inferiority of direct oral anticoagulants (DOACs) compared to warfarin
Most DOACs represent an advance in therapeutic safety when compared to warfarin for prevention of thromboembolism in patients with AF.
Objectives
Determine long term survival, total mortality rates and mortality cause between patients with non-valvular atrial fibrillation (AF) receiving anticoagulants (warfarin, dabigatran and rivaroxaban)
Methods
Retrospective analysis of consecutive patients with AF receiving anticoagulants in two Hospitals in Montevideo, using electronic registries. Demographics, co-morbidities, CHA2DS2VASc scores and mortality cause were annotated. Follow-up started on Jan 2011 and finished on Dec 2017. Anticoagulation quality was expressed as the standard deviation of INRs (SD-INRs). We performed global mortality and mortality cause analysis on patients with anti-VitK versus direct anticoagulants.
Statistical analysis: Survival analysis was performed using Kaplan-Meier (log rank) and Cox regression model. All differences between groups were considered significant if the p value was <0.001.
Results
We studied 4501 pts., 3627 patients were on warfarin (80.6%), 456 (10.1%) were on dabigatran and 418 (9.3%) on rivaroxaban. Those receiving direct anticoagulants were older, 79±9 vs 77±11 years, (p=0.0001), 51.3% were female, with a significantly higher prevalence of HTN; 93.7% vs 88.8% and a CHA2DS2VASc score ≥2 (96% vs 91%), and a lower prevalence of CHD (5.8% vs 10.4%), CHF (3.7% vs 9.5%) and CKD (2.3% vs 6.3%).Total mortality was 818 (18%); patients receiving warfarin had significantly higher mortality rates, 727 (20.1%) vs 91 (10.4%); 63 and 28 (13.8%, 6.7% dabigatran and rivaroxaban respectively) Kaplan-Meier curves were significantly different (Figure 1) showing higher survival rates for those on DOACs. The SD-INRs were 0.85±0.47 (n=1726 alive) vs 1.05±0.46 (n=548 dead), mean difference 0.2 (99% CI 0.14–0.26). Mortality could be analysed in 759 patients (92,7%). The most important cause of death was cardiovascular disease in 26.5%. We could not find significant differences in the cause of death between groups. Using Cox regression model, variables with significant increased mortality were HTN, CHD, CHF, CKD and history of previous CVA. The only variable with a significant decrease in mortality was the use of dabigatran or rivaroxaban; HR 0.55 (95% CI 0.44–0.69)
Figure 1
Conclusions
In this large cohort of patients, those receiving warfarin have significantly higher mortality rates. Mortality differences were not related to stroke or major bleeding but could be explained by a higher prevalence of CHD, CHF and CKD in the warfarin group despite a significant lower CHA2DS2VASc score.
Bioprosthetic valves and atrial fibrillation: Direct anticoagulants or warfarin
