- Email: [email protected]
Dementia correlates with anticoagulation underuse in older patients with atrial fibrillation
Archives of Gerontology and Geriatrics 72 (2017) 108–112
Contents lists available at ScienceDirect
Archives of Gerontology and Geriatrics journal homepage: www.elsevier.com/locate/archger
Dementia correlates with anticoagulation underuse in older patients with atrial fibrillation
MARK
⁎
Giovanni Viscogliosia,b, , Evaristo Ettorrea, Iulia Maria Chiriacc a b c
Division of Gerontology, Department of Cardiovascular, Nephrologic, Anesthesiologic and Geriatric Sciences, Sapienza University, Rome, Italy Department of Epidemiology, Surveillance and Promotion of Health, National Institute of Health, Rome, Italy Geriatric Rehabilitation, Salus Infirmorum Clinic, Italy
A R T I C L E I N F O
A B S T R A C T
Keywords: Anticoagulants Atrial fibrillation Dementia Underuse
Objectives: Stroke prevention in older atrial fibrillation (AF) patients remains a challenge. This study aimed to investigate whether a dementia diagnosis is an independent correlate of lower prescription rate of oral anticoagulant treatment (OAT) in a sample of older AF patients. Methods: Cross-sectional retrospective study. Consecutive older community-dwelling AF patients referred for a comprehensive geriatric assessment, were considered. Evaluation of physical, social and mental health, and administration of the Cumulative Illness Rating Scale (CIRS) and Barthel Index were performed. Dementia cases were ascertained by consensus of 2 experienced geriatricians. Dementia severity was assessed using the Clinical Dementia Rating scale (CDR). Results: 316 AF patients (ages 74.7 ± 7.0 years, 55.7% women) with high stroke risk (77.5% had a CHA2DS2VASC score ≥3), low bleeding and falling risk, and no neuropsychiatric/behavioral symptoms, were included. 60.1% were prescribed with OAT. Among patients with dementia (n = 86, 27.2%), 22.0% received inadequate antithrombotic prophylaxis (i.e. antiplatelet) and 38.5% no treatment. Proportion of those receiving inadequate or no prophylaxis increased at increasing CDR score. By multiple regression models, either dementia (yes vs no), OR = 1.33, 95%CI = 1.11–1.46, p < 0.001, and dementia severity (CDR > 1), OR = 2.38, 95%CI = 2.19–2.60, p < 0.001, were associated with lack of OAT prescription independently of age, paroxysmal AF, and comorbidity burden. Conclusions: Dementia might be associated with underuse of OAT in older AF patients even in the absence of established contraindications. Future studies are needed to assess the real dimension of the problem and clinician’s barriers to prescribing OAT in demented patients.
1. Introduction Atrial fibrillation (AF), the most common arrhythmia encountered in the clinical practice, affects 10% of people older than 80 years, and confers 4–5 fold increased risk of ischemic stroke (Camm et al., 2012; Go et al., 2001). Stroke risk associated with AF can be reduced by 64%–70% with use of oral anticoagulants (Camm et al., 2012; Go et al., 2001). Current evidence suggests that oral anticoagulant treatment (OAT) should be recommended at ≥75 years of age regardless of additional risk factors for stroke (Camm et al., 2012). However, many older AF patients are not prescribed with OAT (Steinberg et al., 2015; Zarraga & Kron, 2013). Although age itself is a risk factor for bleeding (Go et al., 2001), older individuals seldom have absolute contraindications to OAT (Steinberg et al., 2015). The benefit of OAT for stroke prevention in AF has been demonstrated at any age (Go et al.,
⁎
Correspondence to: Viale del Policlinico 155, Rome, Italy. E-mail address: [email protected] (G. Viscogliosi).
http://dx.doi.org/10.1016/j.archger.2017.05.014 Received 9 December 2016; Received in revised form 4 May 2017; Accepted 27 May 2017 Available online 08 June 2017 0167-4943/ © 2017 Published by Elsevier Ireland Ltd.
2001; Mant et al., 2007). Dementia might be associated with lower prescription rate of OAT in older AF patients (Bahri et al., 2015; Dreischulte et al., 2014; Holt et al., 2012; Löppönen et al., 2006; Tanislav et al., 2014). However, it remains unclear the extent to which a dementia diagnosis is associated with lower OAT prescription in older AF patients independently of concomitant factors that might contraindicate OAT. Decision to prescribe or not to prescribe OAT to older subjects is indeed based on several issues. Demented subjects might be more likely to carry more risk factors for bleeding, e.g. greater comorbidity, polypharmacy, disability and risk of falling, when compared to age-matched controls (Steinberg et al., 2015). This study sought to assess whether dementia is independently associated with lower rate of OAT prescription in a sample of older AF patients with low bleeding risk.
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
2. Methods
neuropsychological charts. Self-reported diagnoses were confirmed by 2 experienced geriatricians. Then, in order to identify non-diagnosed dementia cases, all subjects were administered the Mini-mental State Examination (MMSE) test. Those scoring ≤24/30 and/or those who reported complaints of memory loss underwent more in-depth evaluation of cognitive function. Dementia was defined using the DSM-IV-TR criteria (American Psychiatric Association, 2000). Diagnoses were made by consensus of 2 experienced geriatricians. Dementia etiology was not considered for the purposes of this study. Either demented and non-demented participants were administered the Clinical Dementia Rating (CDR) scale (O’Bryant et al., 2008), a composite measure of cognitive and general function that quantifies dementia severity. A CDR score of 0 indicates no dementia; 0.5 questionable dementia; 1 mild dementia; 2 moderate dementia; > 2 severe dementia.
2.1. Study population This was a cross-sectional study with retrospective ascertainment of data. Participants were selected from among older community-dwelling subjects referred by their primary care provider to our outpatients clinic (January 2015–December 2015) for a comprehensive geriatric assessment. Inclusion criteria were: age ≥65 years; a documented diagnosis of paroxysmal, persistent or permanent AF. Medical records of health status, medication prescriptions, previous hospital discharge diagnoses, routine blood tests performed no earlier than 3 months before assessment, and the presence of a caregiver during the visit, were required. A written informed consent, or a proxy consent for those with severe cognitive impairment, was required. The local review board approved the study protocol. Subjects were consecutively included if none of the following criteria applied: previous anticoagulant treatment discontinued for any reason; recent (< 1 year) percutaneous coronary intervention or aortocoronary by-pass; heart valve prosthesis; history of falls or syncope; high risk of falling defined by a score of < 19 on the Tinetti mobility test (Tinetti, 1986); major bleeding episodes; peptic ulcer disease; history of aneurysm; hemostasis disorders; hemoglobin < 10 g/dl; endstage kidney disease; liver insufficiency; active malignancies; uncontrolled blood pressure; daily alcohol consumption > 13 g for women and > 26 g for men in the last 3 months, or any pattern of alcohol addiction; any condition limiting oral medications intake; inability to walk or being on a wheelchair; BMI < 20 kg/m2. The Neuropsychiatric Inventory (NPI) was administered to assess neuropsychiatric/behavioral symptoms (Cummings, 1997); each of the 12 NPI item is assigned a frequency by severity score. Patients who scored > 2 in 3 or more items, were excluded.
2.4. Statistical analysis All analyses were performed via Statistical Package for Social Sciences (SPSS), version 20.0 for Windows (Chicago, IL, USA). The associations between OAT prescription (yes vs no) and clinical variables were analyzed using binary logistic regression, adjusting for age. Significant variables were entered into multiple logistic regression models predicting lack of OAT prescription. Backward elimination of non-significant variables yielded the final models. Due to co-linearity, two separate models were constructed entering dementia (yes vs no) or dementia severity (CDR score > 1) as predicting variables. Statistical significance was set at 2-sided p values at ≤0.05. 3. Results Out of 358 consecutive AF patients, 42 were excluded. The study population consisted of 316 patients ages 74.7 ± 7.0 years, 55.7% women. 190 (60.1%) of them were on OAT. Dementia was ascertained in 86 (27.2%) participants, of whom 81 (94.2%) were already diagnosed. Stroke risk distribution was: CHA2DS2VASc = 1 in 3.5% of participants, score = 2 in 19.0%, ≥3 in 77.5%. Those not prescribed with OAT were more likely to have dementia (Table 1). 22.0% of demented participants received inadequate antithrombotic prophylaxis, i.e. antiplatelet agents only, and 38.5% were left untreated; prescription rate did not change according to disability status and comorbidity (Fig. 1). Proportion of subjects who received inadequate or no prophylaxis increased at increasing CDR score (Fig. 2). By multiple logistic regression analysis, either dementia and CDR > 1 were associated with significantly greater probability of not being prescribed with anticoagulation, after simultaneous adjustment for age, paroxysmal AF and CIRS-CI ≥ 3 (Table 2).
2.2. Clinical assessment Review of archived medical records was performed by experienced physicians. There were no missing data for the subjects eligible to this study. Data on physical, mental and social health were collected using standardized questionnaires. Clinical measurements included weight, height, and blood pressure. A 12-lead resting ECG was taken. Stroke risk was estimated by calculation of the CHA2DS2-VASc score (Camm et al., 2012). Chronic diseases were identified by interview, clinical evidence and review of medical records. The burden of multimorbidity was estimated using the Cumulative Illness Rating Scale (CIRS), whose details are described elsewhere (Parmalee et al., 1995). 14 domains are considered; each is scored from 0 to 4, the higher the score the greater the disease severity. The CIRS comorbidity index (CIRS-CI) is calculated as the sum of items for whom a score of 3 or 4 is obtained. By convention, the fourteenth item, i.e. neuropsychiatric diseases, e.g. dementia, is not considered for CIRS-CI calculation. Thus, comprehensive score ranges from 0 to 13. Difficulty in performing activities of daily living, a proxy of functional disability, was assessed using the Barthel Index (Mahoney & Barthel, 1965). The scale rates the level of dependency for specific ADLs, e.g. self-feeding, bathing, urinary and fecal continence, with score ranging 0–100, the lower the score the greater the functional dependency of the subject. The Tinetti Mobility Test (Tinetti, 1986), a valid instrument to assess balance and gait limitations in older people, was administered to assess fall risk. Comprehensive score (balance + gait subscales) ranges 0–28. High risk of falling is defined by score < 19/28.
4. Discussion This study indicates an underuse of OAT in a sample of older community-dwelling individuals with high stroke risk and low hemorrhagic risk, i.e. no recognized risk factors for bleeding other than old age. 12.0% received inadequate antithrombotic prophylaxis and 28.0% received no treatment. When analyses were limited to those with dementia, proportion of those appropriately prescribed with OAT fell to 38.5%. All AF patients aged ≥75 years have stroke risk as high as > 4%/year regardless of additional risk factors (Camm et al., 2012). Thus, all older AF patients should be evaluated to receiving OAT (Camm et al., 2012; Mant et al., 2007). Influential evidence indicates a favorable risk to benefit ratio of OAT at any age (Go et al., 2001; Mant et al., 2007). In particular, the BAFTA study, in which AF patients aged ≥75 years were randomly assigned to receive warfarin (INR 2–3) or aspirin (75 mg/day), has demonstrated lower incidence of stroke and less intracranial bleedings in those who received warfarin (Mant et al., 2007).
2.3. Dementia cases definition Dementia cases were identified using a 2-step approach. First, selfreported diagnoses were ascertained by review of clinical and 109
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
Table 1 Characteristics of the sample according to anticoagulant treatment use. Anticoagulant treatment
Age > 80 years (%) Women (%) Education < 8 years (%) Paroxysmal atrial fibrillation (%) CHA2DS2VASc score ≥ 3 (%) Heart failure/LV dysfunction (%) Hypertension (%) Diabetes mellitus (%) Stroke/transient ischemic attack (%) Coronary heart disease (%) Peripheral artery disease (%) CIRS comorbidity index ≥ 3 (%) Barthel index score ≤ 75 (%) Dementia (%) Clinical Dementia Rating > 1 (%) Warfarin (%) Non-vitamin K agents (%) Aspirin or clopidogrel (%) β blockers (%) Digoxin (%) Antiarrhythmic agents (%)
Yes (n = 190)
No (n = 126)
OR (95%CI)
p
3.7 63.0 72.6 11.0 70.5 4.7 74.7 13.7 4.7 4.2 4.2 11.7 31.5 16.7 6.0 74.7 25.3 5.0 23.7 15.2 8.7
27.8 44.4 85.0 25.4 88.2 8.0 77.7 17.4 14.3 20.0 8.0 25.4 45.0 43.0 33.4 – – 30.0 25.4 12.0 8.0
0.60 1.06 0.87 0.78 0.90 0.90 0.92 0.97 0.86 0.73 0.86 0.88 0.92 0.73 0.78 – – 0.75 0.97 1.01 1.04
< 0.001 0.247 0.012 < 0.001 0.108 0.316 0.159 0.696 0.124 < 0.001 0.179 0.001 0.339 < 0.001 0.002 – – < 0.001 0.728 0.926 0.590
(0.52–0.70) (0.96–1.17) (0.78–0.97) (0.68–0.89) (0.80–1.02) (0.73–1.11) (0.81–1.03) (0.85–1.12) (0.72–1.04) (0.62–0.86) (0.70–1.07) (0.80–0.94) (0.79–1.08) (0.65–0.82) (0.67–0.91)
(0.66–0.85) (0.83–1.13) (0.87–1.15) (0.90–1.18)
OR and 95% CI for anticoagulant treatment use (yes vs no) after adjustment for age. *p values calculated using binary logistic regression analysis.
neuropsychiatric symptoms, making unlikely the hypothesis of traumaassociated bleeding risk. In addition, the fact that all patients were living with a family member or with a formal caregiver, makes the hypothesis of low treatment adherence less likely as well. Therefore, the finding that 61.5% of a sample of older demented AF patients with high stroke risk and without additional risk factors for bleeding, and without effective contraindications to anticoagulation, received inadequate (22%) or no (39.5%) antithrombotic prophylaxis, is disappointing. In the absence of other conceivable explanations, it seems reasonable to assume that physicians may perceive dementia itself as a factor limiting anticoagulation prescription. Clinical impression of cognitive impairment might particularly affect decision to treat or not to treat. Such hypothesis is corroborated by our finding that dementia severity (CDR > 1) was a stronger predictor of lack of OAT prescription than dementia (yes vs no) by itself. Prevention of thromboembolism in older AF patients by proper prophylaxis is of eminent importance for maintaining functional
Despite such evidence, either primary and secondary stroke prevention in older AF patients remains a challenge (Bahri et al., 2015; Zarraga & Kron, 2013). Clinician’s barriers to prescribing OAT mainly include older age, perceived bleeding risk, and need for frequent blood testing and dose adjustment for vitamin-K antagonists (Bahri et al., 2015; Zarraga & Kron, 2013). In addition, authors have indirectly indicated that impaired cognitive function might limit anticoagulation prescription (Bahri et al., 2015; Dreischulte et al., 2014; Löppönen et al., 2006). Undoubtedly subjects with dementia are more likely to have more severe comorbidity, greater disability, and behavioral/neuropsychiatric symptoms that might increase risk of trauma regardless of absolute fall risk. Furthermore, dementia might result in reduced adherence to therapies, a key factor affecting anticoagulant treatment efficacy. All such features could affect clinician’s perception of hemorrhagic risk, and discourage them from OAT prescription. However, the subjects of this study were tangibly free of recognized risk factors for bleeding, they had low risk of falling, and no relevant behavioral/
Fig. 1. Antithrombotic prophylaxis (%) in participants with dementia (n = 86), according to Barthel Index score and Cumulative Illness Rating Scale – Comorbidity Index (CIRS-CI). a 2 χ = 2.98, p = 0.225; b χ2 = 2.66, p = 0.246; c χ2 = 1.23, p = 0.540.
110
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
Fig. 2. Antithrombotic prophylaxis (%) in all participants (n = 316), according to Clinical Dementia Rating scale (CDR) score. χ2 = 30.68, p < 0.001.
population, the proportion of those unjustifiably not prescribed might be even underestimated. As the population aging grows fast, prevalence of either AF and dementia will increase exponentially over the forthcoming years. Even a modest association between dementia and lack of anticoagulation prescription in older AF patients could have huge effects due to high healthcare costs associated with underuse of OAT in AF (Casciano et al., 2013). Future studies are necessary to assess the real dimension of the problem.
Table 2 Multiple regression models. Factors associated with lack of anticoagulant treatment prescription. Model 1
Age > 80 years (yes/ no) Paroxysmal atrial fibrillation (yes/no) CIRS comorbidity index ≥ 3 (yes/no) Dementia (yes vs no) Clinical Dementia Rating > 1 (yes/ no)
Model 2
OR (95%CI)
p
OR (95%CI)
p
1.12 (1.01–1.24) 1.18 (1.03–1.36) 1.19 (1.08–1.31) 1.33 (1.11–1.46) –
0.043
1.19 (1.05–1.35) 1.19 (1.05–1.35) 1.20 (1.08–1.33) –
0.003
0.002 0.006 < 0.001 –
2.38 (2.19–2.60)
0.002
Author disclosure statement 0.003
No competing financial interests exist.
–
References
< 0.001
American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders (4th ed. text revision). Washington, DC: American Psychiatric Association. Bahri, O., Roca, F., Lechani, T., Druesne, L., Jouanny, P., Serot, J. M., & et al (2015). Underuse of oral anticoagulation for individuals with atrial fibrillation in a nursing home setting in France: comparisons of resident characteristics and physician attitude. J Am Geriatr Soc, 63, 71–76. Camm, A. J., Lip, G. Y., De Caterina, R., Savelieva, I., Atar, D., Hohnloser, S. H., & et al (2012). 2012 focused update of the ESC Guidelines for the management of atrial fibrillation: an update of the 2010 ESC Guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association. Eur Heart J, 33, 2719–2747. Casciano, J. P., Dotiwala, Z. J., Martin, B. C., & Kwong, W. J. (2013). The costs of warfarin underuse and nonadherence in patients with atrial fibrillation: a commercial insurer perspective. J Manag Care Pharm, 19, 302–316. Cummings, J. L. (1997). The neuropsychiatric inventory: Assessing psychopathology in dementia patients. Neurology, 48, S10–S16. Dreischulte, T., Barnett, K., Madhok, V., & Guthrie, B. (2014). Use of oral anticoagulants in atrial fibrillation is highly variable and only weakly associated with estimated stroke risk: cross-sectional population database study. Eur J Gen Pract, 20, 181–189. Eikelboom, J. W., Wallentin, L., Connolly, S. J., Ezekowitz, M., Healey, J. S., Oldgren, J., & et al (2011). Risk of bleeding with 2 doses of dabigatran compared to warfarin in older and younger patients with atrial fibrillation: an analysis of the randomized evaluation of long-term anticoagulant therapy (RE-LY) trial. Circulation, 123, 2363–2371. Go, A. S., Hylek, E. M., Phillips, K. A., Chang, Y., Henault, L. E., Selby, J. V., et al. (2001). Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study. JAMA, 285, 2370–2375. Holt, T. A., Hunter, T. D., Gunnarsson, C., Khan, N., Cload, P., & Lip, G. Y. (2012). Risk of stroke and oral anticoagulant use in atrial fibrillation: a cross-sectional survey. Br J Gen Pract, 62, e710–e717. Löppönen, M., Räihä, I., Isoaho, R., Vahlberg, T., Puolijoki, H., & Kivela, S. L. (2006). Dementia associates with undermedication of cardiovascular diseases in the elderly. Dement Geriatr Cogn Disord, 22, 132–141. Mahoney, F. I., & Barthel, D. W. (1965). Functional evaluation: The Barthel index. Maryland State Medical Journal, 14, 61–65. Mant, J., Hobbs, F. D., Fletcher, K., Roalfe, A., Fitzmaurice, D., Lip, G. Y., et al. (2007). Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomized controlled trial. Lancet, 370, 493–503. O’Bryant, S. E., Waring, S. C., Cullum, C. M., Hall, J., Lacritz, L., Massman, P. J., et al. (2008). Staging dementia using Clinical Dementia Rating Scale Sum of Boxes score: a Texas Alzheimer’s research consortium study. Arch Neurol, 65, 1091–1095.
independency. In particular, AF is a major risk factor for dementia. AF subjects have 2.3-fold increased dementia risk when compared to agematched controls (O’Bryant et al., 2008). Ischemic stroke and chronic cerebral hypoperfusion are the fundamental underlying mechanisms (Udompanich et al., 2013). Thus, OAT in older AF patients is fundamental in preventing both de novo and further cognitive decline. Dementia by itself should not be considered a contraindication to anticoagulation in older AF patients. Subjects with dementia surely represent a frail population. A tailored management of thromboembolic risk in such patients should always be implemented. A multi-disciplinary approach involving gerontologists and cardiologists, and periodical re-evaluation of the benefit to risk ratio should always be guaranteed. Of note, non-vitamin K oral anticoagulants (NOACs) have more predictable pharmacological profile and lesser drug interactions than warfarin, especially at old age (Eikelboom et al., 2011). Future studies are needed to demonstrate whether NOACs could represent the first-line pharmacological approach for thromboembolism prevention in the frail demented population. This study has potential limitations. The cross-sectional design does not allow to elucidate whether the low prescription rate of OAT in demented AF patients is the cause or the consequence of dementia. The fact that participants were selected from among older individuals referred for a comprehensive geriatric assessment could represent a bias. However the proportion of those with dementia and functional disability is similar than that reported for age-matched population-based samples. Furthermore, the sample was composed of subjects with low risk of bleeding and falling. Such features could make the study sample not completely representative of the older non-institutionalized population as a whole. If we assume that such participants are somewhat healthier than those from the general older 111
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
clinical practice. BMC Neurol, 14, 195. Tinetti, M. E. (1986). Performance-oriented assessment of mobility problems in elderly patients. Journal of the American Geriatrics Society, 34, 119–126. Udompanich, S., Lip, G. Y., Apostolakis, S., & Lane, D. A. (2013). Atrial fibrillation as a risk factor for cognitive impairment: a semi-systematic review. QJM, 106, 795–802. Zarraga, I. G., & Kron, J. (2013). Oral anticoagulation in elderly adults with atrial fibrillation: Integrating new options with old concepts. Journal of the American Geriatrics Society, 61, 143–150.
Parmalee, P. A., Thuras, P. D., Katz, I. R., & Lawton, M. P. (1995). Validation of the Cumulative Illness Rating Scale in a geriatric residential population. J Am Geriatr Soc, 43, 130–137. Steinberg, B. A., Greiner, M. A., Hammill, B. G., Curtis, L. H., Benjamin, E. J., Heckbert, S. R., et al. (2015). Contraindications to anticoagulation therapy and eligibility for novel anticoagulants in older patients with atrial fibrillation. Cardiovasc Ther, 33, 177–183. Tanislav, C., Milde, S., Schwartzkopff, S., Sieweke, N., Krämer, H. H., Juenemann, M., et al. (2014). Secondary stroke prevention in atrial fibrillation: a challenge in the
112
Contents lists available at ScienceDirect
Archives of Gerontology and Geriatrics journal homepage: www.elsevier.com/locate/archger
Dementia correlates with anticoagulation underuse in older patients with atrial fibrillation
MARK
⁎
Giovanni Viscogliosia,b, , Evaristo Ettorrea, Iulia Maria Chiriacc a b c
Division of Gerontology, Department of Cardiovascular, Nephrologic, Anesthesiologic and Geriatric Sciences, Sapienza University, Rome, Italy Department of Epidemiology, Surveillance and Promotion of Health, National Institute of Health, Rome, Italy Geriatric Rehabilitation, Salus Infirmorum Clinic, Italy
A R T I C L E I N F O
A B S T R A C T
Keywords: Anticoagulants Atrial fibrillation Dementia Underuse
Objectives: Stroke prevention in older atrial fibrillation (AF) patients remains a challenge. This study aimed to investigate whether a dementia diagnosis is an independent correlate of lower prescription rate of oral anticoagulant treatment (OAT) in a sample of older AF patients. Methods: Cross-sectional retrospective study. Consecutive older community-dwelling AF patients referred for a comprehensive geriatric assessment, were considered. Evaluation of physical, social and mental health, and administration of the Cumulative Illness Rating Scale (CIRS) and Barthel Index were performed. Dementia cases were ascertained by consensus of 2 experienced geriatricians. Dementia severity was assessed using the Clinical Dementia Rating scale (CDR). Results: 316 AF patients (ages 74.7 ± 7.0 years, 55.7% women) with high stroke risk (77.5% had a CHA2DS2VASC score ≥3), low bleeding and falling risk, and no neuropsychiatric/behavioral symptoms, were included. 60.1% were prescribed with OAT. Among patients with dementia (n = 86, 27.2%), 22.0% received inadequate antithrombotic prophylaxis (i.e. antiplatelet) and 38.5% no treatment. Proportion of those receiving inadequate or no prophylaxis increased at increasing CDR score. By multiple regression models, either dementia (yes vs no), OR = 1.33, 95%CI = 1.11–1.46, p < 0.001, and dementia severity (CDR > 1), OR = 2.38, 95%CI = 2.19–2.60, p < 0.001, were associated with lack of OAT prescription independently of age, paroxysmal AF, and comorbidity burden. Conclusions: Dementia might be associated with underuse of OAT in older AF patients even in the absence of established contraindications. Future studies are needed to assess the real dimension of the problem and clinician’s barriers to prescribing OAT in demented patients.
1. Introduction Atrial fibrillation (AF), the most common arrhythmia encountered in the clinical practice, affects 10% of people older than 80 years, and confers 4–5 fold increased risk of ischemic stroke (Camm et al., 2012; Go et al., 2001). Stroke risk associated with AF can be reduced by 64%–70% with use of oral anticoagulants (Camm et al., 2012; Go et al., 2001). Current evidence suggests that oral anticoagulant treatment (OAT) should be recommended at ≥75 years of age regardless of additional risk factors for stroke (Camm et al., 2012). However, many older AF patients are not prescribed with OAT (Steinberg et al., 2015; Zarraga & Kron, 2013). Although age itself is a risk factor for bleeding (Go et al., 2001), older individuals seldom have absolute contraindications to OAT (Steinberg et al., 2015). The benefit of OAT for stroke prevention in AF has been demonstrated at any age (Go et al.,
⁎
Correspondence to: Viale del Policlinico 155, Rome, Italy. E-mail address: [email protected] (G. Viscogliosi).
http://dx.doi.org/10.1016/j.archger.2017.05.014 Received 9 December 2016; Received in revised form 4 May 2017; Accepted 27 May 2017 Available online 08 June 2017 0167-4943/ © 2017 Published by Elsevier Ireland Ltd.
2001; Mant et al., 2007). Dementia might be associated with lower prescription rate of OAT in older AF patients (Bahri et al., 2015; Dreischulte et al., 2014; Holt et al., 2012; Löppönen et al., 2006; Tanislav et al., 2014). However, it remains unclear the extent to which a dementia diagnosis is associated with lower OAT prescription in older AF patients independently of concomitant factors that might contraindicate OAT. Decision to prescribe or not to prescribe OAT to older subjects is indeed based on several issues. Demented subjects might be more likely to carry more risk factors for bleeding, e.g. greater comorbidity, polypharmacy, disability and risk of falling, when compared to age-matched controls (Steinberg et al., 2015). This study sought to assess whether dementia is independently associated with lower rate of OAT prescription in a sample of older AF patients with low bleeding risk.
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
2. Methods
neuropsychological charts. Self-reported diagnoses were confirmed by 2 experienced geriatricians. Then, in order to identify non-diagnosed dementia cases, all subjects were administered the Mini-mental State Examination (MMSE) test. Those scoring ≤24/30 and/or those who reported complaints of memory loss underwent more in-depth evaluation of cognitive function. Dementia was defined using the DSM-IV-TR criteria (American Psychiatric Association, 2000). Diagnoses were made by consensus of 2 experienced geriatricians. Dementia etiology was not considered for the purposes of this study. Either demented and non-demented participants were administered the Clinical Dementia Rating (CDR) scale (O’Bryant et al., 2008), a composite measure of cognitive and general function that quantifies dementia severity. A CDR score of 0 indicates no dementia; 0.5 questionable dementia; 1 mild dementia; 2 moderate dementia; > 2 severe dementia.
2.1. Study population This was a cross-sectional study with retrospective ascertainment of data. Participants were selected from among older community-dwelling subjects referred by their primary care provider to our outpatients clinic (January 2015–December 2015) for a comprehensive geriatric assessment. Inclusion criteria were: age ≥65 years; a documented diagnosis of paroxysmal, persistent or permanent AF. Medical records of health status, medication prescriptions, previous hospital discharge diagnoses, routine blood tests performed no earlier than 3 months before assessment, and the presence of a caregiver during the visit, were required. A written informed consent, or a proxy consent for those with severe cognitive impairment, was required. The local review board approved the study protocol. Subjects were consecutively included if none of the following criteria applied: previous anticoagulant treatment discontinued for any reason; recent (< 1 year) percutaneous coronary intervention or aortocoronary by-pass; heart valve prosthesis; history of falls or syncope; high risk of falling defined by a score of < 19 on the Tinetti mobility test (Tinetti, 1986); major bleeding episodes; peptic ulcer disease; history of aneurysm; hemostasis disorders; hemoglobin < 10 g/dl; endstage kidney disease; liver insufficiency; active malignancies; uncontrolled blood pressure; daily alcohol consumption > 13 g for women and > 26 g for men in the last 3 months, or any pattern of alcohol addiction; any condition limiting oral medications intake; inability to walk or being on a wheelchair; BMI < 20 kg/m2. The Neuropsychiatric Inventory (NPI) was administered to assess neuropsychiatric/behavioral symptoms (Cummings, 1997); each of the 12 NPI item is assigned a frequency by severity score. Patients who scored > 2 in 3 or more items, were excluded.
2.4. Statistical analysis All analyses were performed via Statistical Package for Social Sciences (SPSS), version 20.0 for Windows (Chicago, IL, USA). The associations between OAT prescription (yes vs no) and clinical variables were analyzed using binary logistic regression, adjusting for age. Significant variables were entered into multiple logistic regression models predicting lack of OAT prescription. Backward elimination of non-significant variables yielded the final models. Due to co-linearity, two separate models were constructed entering dementia (yes vs no) or dementia severity (CDR score > 1) as predicting variables. Statistical significance was set at 2-sided p values at ≤0.05. 3. Results Out of 358 consecutive AF patients, 42 were excluded. The study population consisted of 316 patients ages 74.7 ± 7.0 years, 55.7% women. 190 (60.1%) of them were on OAT. Dementia was ascertained in 86 (27.2%) participants, of whom 81 (94.2%) were already diagnosed. Stroke risk distribution was: CHA2DS2VASc = 1 in 3.5% of participants, score = 2 in 19.0%, ≥3 in 77.5%. Those not prescribed with OAT were more likely to have dementia (Table 1). 22.0% of demented participants received inadequate antithrombotic prophylaxis, i.e. antiplatelet agents only, and 38.5% were left untreated; prescription rate did not change according to disability status and comorbidity (Fig. 1). Proportion of subjects who received inadequate or no prophylaxis increased at increasing CDR score (Fig. 2). By multiple logistic regression analysis, either dementia and CDR > 1 were associated with significantly greater probability of not being prescribed with anticoagulation, after simultaneous adjustment for age, paroxysmal AF and CIRS-CI ≥ 3 (Table 2).
2.2. Clinical assessment Review of archived medical records was performed by experienced physicians. There were no missing data for the subjects eligible to this study. Data on physical, mental and social health were collected using standardized questionnaires. Clinical measurements included weight, height, and blood pressure. A 12-lead resting ECG was taken. Stroke risk was estimated by calculation of the CHA2DS2-VASc score (Camm et al., 2012). Chronic diseases were identified by interview, clinical evidence and review of medical records. The burden of multimorbidity was estimated using the Cumulative Illness Rating Scale (CIRS), whose details are described elsewhere (Parmalee et al., 1995). 14 domains are considered; each is scored from 0 to 4, the higher the score the greater the disease severity. The CIRS comorbidity index (CIRS-CI) is calculated as the sum of items for whom a score of 3 or 4 is obtained. By convention, the fourteenth item, i.e. neuropsychiatric diseases, e.g. dementia, is not considered for CIRS-CI calculation. Thus, comprehensive score ranges from 0 to 13. Difficulty in performing activities of daily living, a proxy of functional disability, was assessed using the Barthel Index (Mahoney & Barthel, 1965). The scale rates the level of dependency for specific ADLs, e.g. self-feeding, bathing, urinary and fecal continence, with score ranging 0–100, the lower the score the greater the functional dependency of the subject. The Tinetti Mobility Test (Tinetti, 1986), a valid instrument to assess balance and gait limitations in older people, was administered to assess fall risk. Comprehensive score (balance + gait subscales) ranges 0–28. High risk of falling is defined by score < 19/28.
4. Discussion This study indicates an underuse of OAT in a sample of older community-dwelling individuals with high stroke risk and low hemorrhagic risk, i.e. no recognized risk factors for bleeding other than old age. 12.0% received inadequate antithrombotic prophylaxis and 28.0% received no treatment. When analyses were limited to those with dementia, proportion of those appropriately prescribed with OAT fell to 38.5%. All AF patients aged ≥75 years have stroke risk as high as > 4%/year regardless of additional risk factors (Camm et al., 2012). Thus, all older AF patients should be evaluated to receiving OAT (Camm et al., 2012; Mant et al., 2007). Influential evidence indicates a favorable risk to benefit ratio of OAT at any age (Go et al., 2001; Mant et al., 2007). In particular, the BAFTA study, in which AF patients aged ≥75 years were randomly assigned to receive warfarin (INR 2–3) or aspirin (75 mg/day), has demonstrated lower incidence of stroke and less intracranial bleedings in those who received warfarin (Mant et al., 2007).
2.3. Dementia cases definition Dementia cases were identified using a 2-step approach. First, selfreported diagnoses were ascertained by review of clinical and 109
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
Table 1 Characteristics of the sample according to anticoagulant treatment use. Anticoagulant treatment
Age > 80 years (%) Women (%) Education < 8 years (%) Paroxysmal atrial fibrillation (%) CHA2DS2VASc score ≥ 3 (%) Heart failure/LV dysfunction (%) Hypertension (%) Diabetes mellitus (%) Stroke/transient ischemic attack (%) Coronary heart disease (%) Peripheral artery disease (%) CIRS comorbidity index ≥ 3 (%) Barthel index score ≤ 75 (%) Dementia (%) Clinical Dementia Rating > 1 (%) Warfarin (%) Non-vitamin K agents (%) Aspirin or clopidogrel (%) β blockers (%) Digoxin (%) Antiarrhythmic agents (%)
Yes (n = 190)
No (n = 126)
OR (95%CI)
p
3.7 63.0 72.6 11.0 70.5 4.7 74.7 13.7 4.7 4.2 4.2 11.7 31.5 16.7 6.0 74.7 25.3 5.0 23.7 15.2 8.7
27.8 44.4 85.0 25.4 88.2 8.0 77.7 17.4 14.3 20.0 8.0 25.4 45.0 43.0 33.4 – – 30.0 25.4 12.0 8.0
0.60 1.06 0.87 0.78 0.90 0.90 0.92 0.97 0.86 0.73 0.86 0.88 0.92 0.73 0.78 – – 0.75 0.97 1.01 1.04
< 0.001 0.247 0.012 < 0.001 0.108 0.316 0.159 0.696 0.124 < 0.001 0.179 0.001 0.339 < 0.001 0.002 – – < 0.001 0.728 0.926 0.590
(0.52–0.70) (0.96–1.17) (0.78–0.97) (0.68–0.89) (0.80–1.02) (0.73–1.11) (0.81–1.03) (0.85–1.12) (0.72–1.04) (0.62–0.86) (0.70–1.07) (0.80–0.94) (0.79–1.08) (0.65–0.82) (0.67–0.91)
(0.66–0.85) (0.83–1.13) (0.87–1.15) (0.90–1.18)
OR and 95% CI for anticoagulant treatment use (yes vs no) after adjustment for age. *p values calculated using binary logistic regression analysis.
neuropsychiatric symptoms, making unlikely the hypothesis of traumaassociated bleeding risk. In addition, the fact that all patients were living with a family member or with a formal caregiver, makes the hypothesis of low treatment adherence less likely as well. Therefore, the finding that 61.5% of a sample of older demented AF patients with high stroke risk and without additional risk factors for bleeding, and without effective contraindications to anticoagulation, received inadequate (22%) or no (39.5%) antithrombotic prophylaxis, is disappointing. In the absence of other conceivable explanations, it seems reasonable to assume that physicians may perceive dementia itself as a factor limiting anticoagulation prescription. Clinical impression of cognitive impairment might particularly affect decision to treat or not to treat. Such hypothesis is corroborated by our finding that dementia severity (CDR > 1) was a stronger predictor of lack of OAT prescription than dementia (yes vs no) by itself. Prevention of thromboembolism in older AF patients by proper prophylaxis is of eminent importance for maintaining functional
Despite such evidence, either primary and secondary stroke prevention in older AF patients remains a challenge (Bahri et al., 2015; Zarraga & Kron, 2013). Clinician’s barriers to prescribing OAT mainly include older age, perceived bleeding risk, and need for frequent blood testing and dose adjustment for vitamin-K antagonists (Bahri et al., 2015; Zarraga & Kron, 2013). In addition, authors have indirectly indicated that impaired cognitive function might limit anticoagulation prescription (Bahri et al., 2015; Dreischulte et al., 2014; Löppönen et al., 2006). Undoubtedly subjects with dementia are more likely to have more severe comorbidity, greater disability, and behavioral/neuropsychiatric symptoms that might increase risk of trauma regardless of absolute fall risk. Furthermore, dementia might result in reduced adherence to therapies, a key factor affecting anticoagulant treatment efficacy. All such features could affect clinician’s perception of hemorrhagic risk, and discourage them from OAT prescription. However, the subjects of this study were tangibly free of recognized risk factors for bleeding, they had low risk of falling, and no relevant behavioral/
Fig. 1. Antithrombotic prophylaxis (%) in participants with dementia (n = 86), according to Barthel Index score and Cumulative Illness Rating Scale – Comorbidity Index (CIRS-CI). a 2 χ = 2.98, p = 0.225; b χ2 = 2.66, p = 0.246; c χ2 = 1.23, p = 0.540.
110
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
Fig. 2. Antithrombotic prophylaxis (%) in all participants (n = 316), according to Clinical Dementia Rating scale (CDR) score. χ2 = 30.68, p < 0.001.
population, the proportion of those unjustifiably not prescribed might be even underestimated. As the population aging grows fast, prevalence of either AF and dementia will increase exponentially over the forthcoming years. Even a modest association between dementia and lack of anticoagulation prescription in older AF patients could have huge effects due to high healthcare costs associated with underuse of OAT in AF (Casciano et al., 2013). Future studies are necessary to assess the real dimension of the problem.
Table 2 Multiple regression models. Factors associated with lack of anticoagulant treatment prescription. Model 1
Age > 80 years (yes/ no) Paroxysmal atrial fibrillation (yes/no) CIRS comorbidity index ≥ 3 (yes/no) Dementia (yes vs no) Clinical Dementia Rating > 1 (yes/ no)
Model 2
OR (95%CI)
p
OR (95%CI)
p
1.12 (1.01–1.24) 1.18 (1.03–1.36) 1.19 (1.08–1.31) 1.33 (1.11–1.46) –
0.043
1.19 (1.05–1.35) 1.19 (1.05–1.35) 1.20 (1.08–1.33) –
0.003
0.002 0.006 < 0.001 –
2.38 (2.19–2.60)
0.002
Author disclosure statement 0.003
No competing financial interests exist.
–
References
< 0.001
American Psychiatric Association (2000). Diagnostic and statistical manual of mental disorders (4th ed. text revision). Washington, DC: American Psychiatric Association. Bahri, O., Roca, F., Lechani, T., Druesne, L., Jouanny, P., Serot, J. M., & et al (2015). Underuse of oral anticoagulation for individuals with atrial fibrillation in a nursing home setting in France: comparisons of resident characteristics and physician attitude. J Am Geriatr Soc, 63, 71–76. Camm, A. J., Lip, G. Y., De Caterina, R., Savelieva, I., Atar, D., Hohnloser, S. H., & et al (2012). 2012 focused update of the ESC Guidelines for the management of atrial fibrillation: an update of the 2010 ESC Guidelines for the management of atrial fibrillation. Developed with the special contribution of the European Heart Rhythm Association. Eur Heart J, 33, 2719–2747. Casciano, J. P., Dotiwala, Z. J., Martin, B. C., & Kwong, W. J. (2013). The costs of warfarin underuse and nonadherence in patients with atrial fibrillation: a commercial insurer perspective. J Manag Care Pharm, 19, 302–316. Cummings, J. L. (1997). The neuropsychiatric inventory: Assessing psychopathology in dementia patients. Neurology, 48, S10–S16. Dreischulte, T., Barnett, K., Madhok, V., & Guthrie, B. (2014). Use of oral anticoagulants in atrial fibrillation is highly variable and only weakly associated with estimated stroke risk: cross-sectional population database study. Eur J Gen Pract, 20, 181–189. Eikelboom, J. W., Wallentin, L., Connolly, S. J., Ezekowitz, M., Healey, J. S., Oldgren, J., & et al (2011). Risk of bleeding with 2 doses of dabigatran compared to warfarin in older and younger patients with atrial fibrillation: an analysis of the randomized evaluation of long-term anticoagulant therapy (RE-LY) trial. Circulation, 123, 2363–2371. Go, A. S., Hylek, E. M., Phillips, K. A., Chang, Y., Henault, L. E., Selby, J. V., et al. (2001). Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the Anticoagulation and Risk Factors in Atrial Fibrillation (ATRIA) study. JAMA, 285, 2370–2375. Holt, T. A., Hunter, T. D., Gunnarsson, C., Khan, N., Cload, P., & Lip, G. Y. (2012). Risk of stroke and oral anticoagulant use in atrial fibrillation: a cross-sectional survey. Br J Gen Pract, 62, e710–e717. Löppönen, M., Räihä, I., Isoaho, R., Vahlberg, T., Puolijoki, H., & Kivela, S. L. (2006). Dementia associates with undermedication of cardiovascular diseases in the elderly. Dement Geriatr Cogn Disord, 22, 132–141. Mahoney, F. I., & Barthel, D. W. (1965). Functional evaluation: The Barthel index. Maryland State Medical Journal, 14, 61–65. Mant, J., Hobbs, F. D., Fletcher, K., Roalfe, A., Fitzmaurice, D., Lip, G. Y., et al. (2007). Warfarin versus aspirin for stroke prevention in an elderly community population with atrial fibrillation (the Birmingham Atrial Fibrillation Treatment of the Aged Study, BAFTA): a randomized controlled trial. Lancet, 370, 493–503. O’Bryant, S. E., Waring, S. C., Cullum, C. M., Hall, J., Lacritz, L., Massman, P. J., et al. (2008). Staging dementia using Clinical Dementia Rating Scale Sum of Boxes score: a Texas Alzheimer’s research consortium study. Arch Neurol, 65, 1091–1095.
independency. In particular, AF is a major risk factor for dementia. AF subjects have 2.3-fold increased dementia risk when compared to agematched controls (O’Bryant et al., 2008). Ischemic stroke and chronic cerebral hypoperfusion are the fundamental underlying mechanisms (Udompanich et al., 2013). Thus, OAT in older AF patients is fundamental in preventing both de novo and further cognitive decline. Dementia by itself should not be considered a contraindication to anticoagulation in older AF patients. Subjects with dementia surely represent a frail population. A tailored management of thromboembolic risk in such patients should always be implemented. A multi-disciplinary approach involving gerontologists and cardiologists, and periodical re-evaluation of the benefit to risk ratio should always be guaranteed. Of note, non-vitamin K oral anticoagulants (NOACs) have more predictable pharmacological profile and lesser drug interactions than warfarin, especially at old age (Eikelboom et al., 2011). Future studies are needed to demonstrate whether NOACs could represent the first-line pharmacological approach for thromboembolism prevention in the frail demented population. This study has potential limitations. The cross-sectional design does not allow to elucidate whether the low prescription rate of OAT in demented AF patients is the cause or the consequence of dementia. The fact that participants were selected from among older individuals referred for a comprehensive geriatric assessment could represent a bias. However the proportion of those with dementia and functional disability is similar than that reported for age-matched population-based samples. Furthermore, the sample was composed of subjects with low risk of bleeding and falling. Such features could make the study sample not completely representative of the older non-institutionalized population as a whole. If we assume that such participants are somewhat healthier than those from the general older 111
Archives of Gerontology and Geriatrics 72 (2017) 108–112
G. Viscogliosi et al.
clinical practice. BMC Neurol, 14, 195. Tinetti, M. E. (1986). Performance-oriented assessment of mobility problems in elderly patients. Journal of the American Geriatrics Society, 34, 119–126. Udompanich, S., Lip, G. Y., Apostolakis, S., & Lane, D. A. (2013). Atrial fibrillation as a risk factor for cognitive impairment: a semi-systematic review. QJM, 106, 795–802. Zarraga, I. G., & Kron, J. (2013). Oral anticoagulation in elderly adults with atrial fibrillation: Integrating new options with old concepts. Journal of the American Geriatrics Society, 61, 143–150.
Parmalee, P. A., Thuras, P. D., Katz, I. R., & Lawton, M. P. (1995). Validation of the Cumulative Illness Rating Scale in a geriatric residential population. J Am Geriatr Soc, 43, 130–137. Steinberg, B. A., Greiner, M. A., Hammill, B. G., Curtis, L. H., Benjamin, E. J., Heckbert, S. R., et al. (2015). Contraindications to anticoagulation therapy and eligibility for novel anticoagulants in older patients with atrial fibrillation. Cardiovasc Ther, 33, 177–183. Tanislav, C., Milde, S., Schwartzkopff, S., Sieweke, N., Krämer, H. H., Juenemann, M., et al. (2014). Secondary stroke prevention in atrial fibrillation: a challenge in the
112