Abstract
This cross-sectional study evaluates the degree of anticoagulant use among nursing home residents with advanced dementia and atrial fibrillation at the end of life.
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.
Methods
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.
Results
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.
Table. Demographic and Clinical Characteristics by Anticoagulant Use.
| Characteristic | No. (%) | Adjusted OR (95% CI) | |
|---|---|---|---|
| Anticoagulation in the last 6 mo of life (n = 5033) | No anticoagulation in the last 6 mo of life (n = 10 184) | ||
| Age (overall), y | |||
| <80 | 785 (15.6) | 1142 (11.2) | 1 [Reference] |
| 80 to 89 | 2560 (50.9) | 4541 (44.6) | 0.82 (0.74-0.92) |
| ≥90 | 1688 (33.5) | 4501 (44.2) | 0.59 (0.52-0.66) |
| Female sex | 3259 (64.8) | 7125 (70.0) | 0.88 (0.81-0.95) |
| Race/ethnicity | |||
| White | 4325 (85.9) | 8839 (86.8) | 1 [Reference] |
| Black | 473 (9.4) | 861 (8.5) | 0.97 (0.85-1.10) |
| Asian | 77 (1.5) | 173 (1.7) | 0.80 (0.60-1.07) |
| Hispanic | 101 (2.0) | 208 (2.0) | 0.97 (0.75-1.25) |
| Other/missinga | 57 (1.1) | 103 (1.0) | 1.06 (0.75-1.50) |
| Non-Medicaid status | 1916 (38.1) | 3287 (32.3) | 1.59 (1.45-1.69) |
| >1 y in nursing home | 2154 (42.8) | 2399 (23.6) | 2.68 (2.48-2.89) |
| CHA2DS2VASC scoreb | |||
| <4 | 681 (13.5) | 1661 (16.3) | 1 [Reference] |
| 5-6 | 1914 (38.0) | 4242 (41.7) | 1.10 (0.99-1.23) |
| >7 | 2438 (48.4) | 4281 (42.0) | 1.38 (1.23-1.54) |
| ATRIA scorec | |||
| <3 | 796 (15.8) | 2205 (21.7) | 1 [Reference] |
| 4-6 | 1690 (33.6) | 3584 (35.2) | 1.19 (1.07-1.32) |
| >7 | 2547 (50.6) | 4395 (43.2) | 1.25 (1.13-1.39) |
| Other clinical factors | |||
| Rejection of care | 454 (9.0) | 844 (8.3) | 1.03 (0.91-1.17) |
| Falls | 1690 (33.6) | 3214 (31.6) | 1.04 (0.96-1.12) |
| Weight loss | 1856 (37.0) | 3363 (33.5) | 1.09 (1.01-1.18) |
| Pressure ulcer | 2052 (40.8) | 3228 (31.7) | 1.37 (1.27-1.48) |
| Difficulty swallowing | 1136 (22.6) | 2008 (19.8) | 1.12 (1.02-1.22) |
| Restraint use | 189 (3.8) | 433 (4.3) | 0.79 (0.66-0.95) |
| Hospice use | 1296 (25.8) | 3375 (33.1) | 0.76 (0.70-0.83) |
Abbreviation: OR, odds ratio.
Other includes American Indian, Alaskan Native, Native Hawaiian, or Pacific Islander.
The CHA2DS2VASC score predicts annual stroke risk among persons with atrial fibrillation using 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], and vascular disease.
The ATRIA (Anticoagulation and Risk Factors In Atrial Fibrillation) bleeding risk score predicts annual bleeding risk among persons anticoagulated for atrial fibrillation using the following risk factors: age ≥75 years (2 points), anemia (3 points), renal disease (3 points), prior bleeding (1 point), and hypertension (1 point).
Discussion
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.
References
- 1.Subic A, Cermakova P, Religa D, et al. Treatment of atrial fibrillation in patients with dementia: a cohort study from the Swedish Dementia Registry. J Alzheimers Dis. 2018;61(3):1119-1128. doi: 10.3233/JAD-170575 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 2.Lip GYH, Nieuwlaat R, Pisters R, Lane DA, Crijns HJGM. Refining clinical risk stratification for predicting stroke and thromboembolism in atrial fibrillation using a novel risk factor-based approach: the Euro Heart Survey on Atrial Fibrillation. Chest. 2010;137(2):263-272. doi: 10.1378/chest.09-1584 [DOI] [PubMed] [Google Scholar]
- 3.January CT, Wann LS, Alpert JS, et al. ; ACC/AHA Task Force Members . 2014 AHA/ACC/HRS guideline for the management of patients with atrial fibrillation: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on practice guidelines and the Heart Rhythm Society. Circulation. 2014;130(23):2071-2104. doi: 10.1161/CIR.0000000000000040 [DOI] [PubMed] [Google Scholar]
- 4.Mitchell SL, Teno JM, Kiely DK, et al. The clinical course of advanced dementia. N Engl J Med. 2009;361(16):1529-1538. doi: 10.1056/NEJMoa0902234 [DOI] [PMC free article] [PubMed] [Google Scholar]
- 5.Morris JN, Fries BE, Mehr DR, et al. MDS Cognitive Performance Scale. J Gerontol. 1994;49(4):M174-M182. doi: 10.1093/geronj/49.4.M174 [DOI] [PubMed] [Google Scholar]
- 6.Fang MC, Go AS, Chang Y, et al. A new risk scheme to predict warfarin-associated hemorrhage: The ATRIA (Anticoagulation and Risk Factors in Atrial Fibrillation) Study. J Am Coll Cardiol. 2011;58(4):395-401. doi: 10.1016/j.jacc.2011.03.031 [DOI] [PMC free article] [PubMed] [Google Scholar]