Customize your JAMA Network experience by selecting one or more topics from the list below.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Ouellet GM, Fried TR, Gilstrap LG, et al. Anticoagulant Use for Atrial Fibrillation Among Persons With Advanced Dementia at the End of Life. JAMA Intern Med. 2021;181(8):1121–1123. doi:10.1001/jamainternmed.2021.1819
Atrial fibrillation (AF) affects almost 20% of individuals with dementia.1 By virtue of their age and comorbidities, nearly all patients with AF and dementia meet the threshold stroke risk, as estimated by the CHA2DS2VASC score (a score to predict annual stroke risk in persons with AF based on the following risk factors: congestive heart failure, hypertension, age [>65 years = 1 point; >75 years = 2 points], diabetes, previous stroke or transient ischemic attack [2 points], vascular disease),2 for which guidelines recommend anticoagulation.3 As dementia progresses, function is gradually but irretrievably lost, so that the potential benefits of preventing a stroke become increasingly attenuated. In advanced dementia, patients develop profound cognitive deficits; require help with basic self-care activities, including eating; and have a very limited life expectancy.4 Our objective was to determine how often anticoagulation is continued among nursing home residents in this final stage of life and to examine clinical associations of its use.
In this cross-sectional study, we used Medicare data to identify nursing home residents 65 years or older with advanced dementia and AF who had at least moderate stroke risk (CHA2DS2VASC score ≥2), and who died between January 1, 2014, and December 31, 2017. Advanced dementia was defined as a diagnosis of Alzheimer disease or another dementia, Cognitive Performance Score of 5 or 6,5 and dependence in all activities of daily living on 2 Minimum Data Set assessments within the last 6 months of life. We used Chronic Condition Warehouse flags to ascertain AF. We excluded residents not enrolled in fee-for-service Medicare and those with claims for venous thromboembolism and valvular heart disease (including mechanical valves) in the 2 years before death. This study was approved by the institutional review boards at Yale University and the VA Connecticut Healthcare System, which waived the need for informed consent owing to the use of deidentified data. This study followed the Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) reporting guideline.
Data were analyzed from October 1 to December 30, 2020. The CHA2DS2VASC score (for stroke risk) and the ATRIA (Anticoagulation and Risk Factors In Atrial Fibrillation) score, which is used to predict annual bleeding risk among anticoagulated persons with AF (based on age ≥75 years [2 points], anemia [3 points], renal disease [3 points], prior bleeding [1 point], and hypertension [1 point]), were calculated using Chronic Condition Warehouse flags to identify comorbidities in each.2,6 Bleeding history, part of the ATRIA score, was determined using inpatient billing codes from the 2 years prior to death.6 We ascertained anticoagulant use from Minimum Data Set item N0410E on assessments in the last 6 months of life. Other clinical factors were obtained from Minimum Data Set assessments during the same period. We modeled the associations between anticoagulant use and patient characteristics (selected a priori) using multivariable logistic regression in Stata, version 15 (StataCorp LLC). A 2-sided P < .05 in the multivariable logistic regression model as determined by a Wald χ2 test, denoted statistical significance.
Among 15 217 nursing home residents with AF and advanced dementia (mean [SD] age, 87.5 [6.76] years; 10 384 women [68.2%]), 5033 (33.1%) received an anticoagulant in the last 6 months of life. The Table shows patient characteristics by anticoagulation status, along with adjusted odds ratios (ORs) for anticoagulant use. In multivariable logistic regression, higher CHA2DS2VASC (score >7, OR, 1.38 [95% CI, 1.23-1.54]) and ATRIA (score >7, OR, 1.25 [95% CI, 1.13-1.39]) scores, nursing home length of stay of at least 1 year (OR, 2.68 [95% CI, 2.48-2.89]), not having Medicaid (OR, 1.59 [95% CI, 1.45-1.69]), weight loss (OR, 1.09 [95% CI, 1.01-1.18]), pressure ulcers (OR, 1.37 [95% CI, 1.27-1.48]), and difficulty swallowing (OR, 1.12 [95% CI, 1.02-1.22]) were associated with greater odds of anticoagulant use. Conversely, older age (80-89 y, OR, 0.82 [95% CI, 0.74-0.92]; ≥90 y, OR, 0.59 [95% CI, 0.52-0.66]), female sex (OR, 0.88 [95% CI, 0.81-0.95]), requiring restraints (OR, 0.79 [95% CI, 0.66-0.95]), and being enrolled in hospice (OR, 0.76 [95% CI, 0.70-0.83]) were associated with lesser odds of anticoagulant use.
In this cross-sectional study, we found that almost one-third of nursing home residents with AF and advanced dementia remained on anticoagulation in the last 6 months of life. Nursing home length of stay at least 1 year and not having Medicaid were more strongly associated with anticoagulant use than CHA2DS2VASC score. Greater bleeding risk, counterintuitively, was associated with greater odds of anticoagulant use. With the notable exception of hospice use, most indicators of high short-term mortality, such as difficulty swallowing, weight loss, and pressure ulcers, were associated with greater odds of anticoagulant use.
These findings underscore the fact that, while practice guidelines contain a well-defined threshold for starting anticoagulation for AF, there is no clear standard for stopping it. Clinicians are instead asked to engage in shared decision-making with patients and their families.3 Data about the benefits and harms of therapy are essential to that process. For patients with dementia, little such evidence is available, although the magnitudes of benefits and harms are likely to change substantially as the disease progresses. This study is limited by its cross-sectional design and includes only persons with AF and advanced dementia in the nursing home setting. Nonetheless, our work points to the need for high-quality data to inform decision-making about anticoagulation in this population.
Accepted for Publication: March 16, 2021.
Published Online: May 10, 2021. doi:10.1001/jamainternmed.2021.1819
Corresponding Author: Gregory M. Ouellet, MD, MHS, 333 Cedar St, PO Box 208025, New Haven, CT 06520-8025 (firstname.lastname@example.org).
Author Contributions: Dr Ouellet had full access to all of the data in the study and takes responsibility for the integrity of the data and the accuracy of the data analysis.
Concept and design: Ouellet, Fried, Cohen.
Acquisition, analysis, or interpretation of data: Ouellet, Gilstrap, O’Leary, Austin, Skinner, Cohen.
Drafting of the manuscript: Ouellet, Cohen.
Critical revision of the manuscript for important intellectual content: Fried, Gilstrap, O’Leary, Austin, Skinner, Cohen.
Statistical analysis: Ouellet, Gilstrap, O’Leary, Skinner, Cohen.
Obtained funding: Ouellet.
Administrative, technical, or material support: O’Leary.
Supervision: Gilstrap, Skinner, Cohen.
Conflict of Interest Disclosures: Dr Ouellet reported receiving grants from the National Institutes of Health (NIH) during the conduct of the study. Dr Fried reported receiving grants from the NIH during the conduct of the study. Dr Gilstrap reported receiving grants from the NIH during the conduct of the study. Dr Skinner reported receiving grants from the NIH during the conduct of the study; personal fees from Sutter Health and the National Bureau of Economic Research; and equity from Dorsata, Inc. outside the submitted work. Dr Cohen reported receiving grants from the NIH during the conduct of the study. No other disclosures were reported.
Funding/Support: Support for data from the Centers for Medicare and Medicaid Services was provided by the Department of Veterans Affairs, VA Health Services Research and Development Service, VA Information Resource Center (project numbers SDR 02-237 and 98-004). Dr Ouellet was supported by a GEMSSTAR award from the National Institute on Aging (R03AG064255), a Pepper Scholars Award from the Yale Pepper Center (P30AG21342), and a Robert E. Leet and Clara Guthrie Patterson Mentored Research Award, Bank of America, N.A., Trustee. Dr Gilstrap was supported by a Patient-Oriented Career Development Award from NHLBI (K23HL142835). Dr Cohen was supported by a Beeson award (K76AG059987) from the National Institute on Aging. All authors were supported by the Yale Claude D. Pepper Older Americans Independence Center (P30AG21342).
Role of the Funder/Sponsor: The sponsors had no role in the study design, analysis, or preparation of this research letter.