Editor's Note: Commentary based on Orkaby AR, Ozonoff A, Reisman JI, Miller DR, Zhao S, Rose AJ. Continued use of Warfarin in Veterans with Atrial Fibrillation After Dementia Diagnosis. J Am Geriatr Soc 2017;65:249-56.

Rationale for Study/Background: Warfarin in patients with nonvalvular atrial fibrillation and a moderate risk for stroke provides >60% reduction in the risk of ischemic stroke. However, the benefits of warfarin in older adults with concomitant dementia are not well characterized, as such subjects were excluded from all major clinical trials. Given the increasing incidence of atrial fibrillation and dementia with advancing age, as well as the potential impact of cognitive dysfunction on international normalized ratio (INR) monitoring, this study addresses a critically important issue as to whether warfarin is safe and effective in patients with atrial fibrillation and newly diagnosed dementia.

Funding: VA Health Services Research and Development, American Federation for Aging Research, John A. Hartford Foundation and Centers of Excellence National Program.

Methods

Design: Retrospective study with inception cohort as defined below.

Inclusion Criteria: Subjects with dementia diagnosed by ICD-9 code in 2007-2008 who were greater than 65 years of age, receiving warfarin for nonvalvular atrial fibrillation for at least 6 months prior to a dementia diagnosis, and had at least two INRs measured 56 days or less apart.

Exclusion Criteria: Medicare Advantage program (because outcomes outside the VA system could not be captured) and subjects transferred to non-VA nursing homes.

Exposure: Warfarin therapy after dementia diagnosis.

Primary Outcome(s): Major hemorrhage, stroke and all-cause mortality over 4 years of follow-up.

Secondary Outcomes: Persistent use of warfarin, anticoagulation control.

Statistical Analysis: To control for confounding by indication and healthy user bias, two to one matching based on propensity scores was used. The propensity score model included demographic features, comorbid conditions and number of medications. Kaplan Meier survival curves for times to events of interest (death, stroke, major hemorrhage, or any of these events) were constructed comparing those who continued versus those who did not continue warfarin after dementia diagnosis, both for the cohort overall and in the matched dataset. Cox proportional hazards were used for adjusted outcomes.

Results: The average age was 80 and almost all subjects were male (98.6%) with an average CHADS2 score of 3.3±1.3. After a diagnosis of dementia, 405 individuals (16%) persisted on warfarin therapy and time in therapeutic range did not differ significantly after compared with before a dementia diagnosis. Unadjusted Cox proportional hazards analysis demonstrated a protective effect of warfarin in prevention of ischemic stroke (hazard ratio [HR] 0.64, 95% confidence interval [CI] 0.46–0.89, P = .008), major bleeding (HR 0.72, 95% CI 0.55– 0.94, P = .02), and all-cause mortality (HR 0.66, 95% CI 0.55–0.79, P < .001). Using propensity score matching, the protective effect of continuing warfarin persisted in prevention of stroke (HR 0.74, 95% CI 0.54–0.996, P = .047) and mortality (HR 0.72, 95% CI 0.60–0.87, P < .001), but the difference in risk of major bleeding was no longer significant (HR 0.78, 95% CI 0.61–1.01, P = .06)

Conclusion: Discontinuing warfarin after a diagnosis of dementia in older adults with non-valvular atrial fibrillation is associated with a significant increase in stroke and mortality.

Limitations of Study: This was a retrospective study with an inception cohort and residual confounding is likely. Generalizability is limited given the nearly exclusively Caucasian male population. Cohorts were defined and outcomes were assessed by ICD-9 codes, which likely led to an under-diagnosis of dementia. Confounding effect of aspirin use not studied. No formal measures of frailty or function were employed but rather surrogates such as recent cancer, ocular disorders, psychiatric disease, venous insufficiency and polypharmacy were used. Caregiver support for patients who continued to receive anticoagulation was not delineated but is likely a critical factor in successful long term use of warfarin in patients with cognitive impairment. Finally, the study did not include patients treated with NOACs.

Geriatric Perspective for the Cardiovascular Clinician: Assessment of cognitive function is an essential part of the clinical evaluation for all older adults with cardiovascular disease, as cognitive function is common, often under-recognized, and impacts on adherence to complex regimens.

This retrospective study with an inception cohort provides evidence of potential harms associated with discontinuation of warfarin therapy in older adults who have a new diagnosis of dementia. These data add to the growing evidence base supporting the benefits of chronic anticoagulation in older adults with atrial fibrillation with respect to important endpoints including stroke and mortality. In the absence of data from randomized controlled trials, observational data such as these can provide critical information about the value of various therapies for frail, older adults with cognitive dysfunction and other geriatric syndromes. Importantly, these data may be used by clinicians to enhance the process of shared decision making, thereby optimizing patient centered care for the growing population of older adults with atrial fibrillation and dementia who may benefit from continued anticoagulation. Full endorsement of continuing anticoagulation for atrial fibrillation with newly diagnosed dementia will require more robust data from RCTs, which would ideally include the use of NOACs as one of the study arms, but such data are unlikely to be forthcoming in the near future.

Keywords: Warfarin, International Normalized Ratio, Atrial Fibrillation, Frail Elderly, Stroke, Hemorrhage, Dementia, Risk, Risk Factors, Cognition, Venous Insufficiency, Geriatrics

< Back to Listings