- Email: [email protected]
Safety of Direct Oral Anticoagulants and Vitamin K Antagonists in Oldest Old Patients: A Prospective Study
JAMDA xxx (2018) 1e6
JAMDA journal homepage: www.jamda.com
Original Study
Safety of Direct Oral Anticoagulants and Vitamin K Antagonists in Oldest Old Patients: A Prospective Study Carlos Rodríguez-Pascual MD, PhD a, b, *, Ivett Torres-Torres MD a, Alejandro Gómez-Quintanilla MD a, Ana Isabel Ferrero-Martínez NP a, Jagdish Sharma MD, PhD b, Alba Guitián c, Maria Carmen Basalo MD c, Marina Montero-Magan MD a, Arturo Vilches-Moraga MD a, d, Maria-Teresa Olcoz-Chiva MD a, b, Emilio Paredes-Galán MD c, Fernando Rodríguez-Artalejo MD, PhD e a
Geriatric Medicine Department, Complejo Hospitalario Universitario de Vigo, Spain University of Lincoln and Lincoln County Hospital, Lincoln, Lincolnshire, United Kingdom c Cardiology Department, Complejo Hospitalario Universitario de Vigo, Spain d Ageing and Complex Medicine Department, Salford Royal NHS Foundation Trust, Manchester, United Kingdom e Department of Preventive Medicine and Public Health, Universidad Autónoma de Madrid /IdiPaz, CIBERESP, and IMDEA-Food Institute, Madrid, Spain b
a b s t r a c t Keywords: Atrial fibrillation elderly anticoagulants safety
Objective: The safety of direct oral anticoagulants (DOACs) in oldest old patients with nonvalvular atrial fibrillation (NVAF) in daily clinical practice has not been systematically assessed. This study examined the safety of DOACs and dicumarol (a vitamin K antagonist) in NVAF geriatric patients. Design: Prospective study from January 2010 through June 2015, with follow-up through January 2016. Setting: Geriatric medicine department at a tertiary hospital. Participants: A total of 554 outpatients, 75 years or older, diagnosed of NVAF and starting oral anticoagulation. Measurements: The main outcome was bleeding, which was classified into major (including those life-threatening) and nonmajor episodes. Statistical analyses were performed with Cox regression. Results: A total of 351 patients received DOACs and 193 dicumarol. Patients on DOACs were older, with more frequent comorbidities, mobility limitation and disability in activities of daily living, as well as higher mortality, than those treated with dicumarol. The incidence of any bleeding was 19.2/100 personyears among patients on DOACs and 13.7/100 person-years on dicumarol; corresponding figures for major bleeding were 5.2 for those on DOACs, and 3.3 for those on dicumarol. In crude analyses, hazard ratios (95% confidence intervals) for any bleeding, and for mayor bleeding in patients on DOACs vs dicumarol were 1.60 (1.04-2.44) and 2.22 (0.88-5.59), respectively. Excess risk of bleeding associated with DOACs vs dicumarol disappeared after adjustment for clinical characteristics, so that corresponding figures were 1.19 (0.68-2.08) and 1.01 (0.35-2.93). Results did not vary across subgroups of high-risk patients. Conclusion: In very old patients with NVAF, the higher risk of bleeding associated with DOACs vs dicumarol could be mostly explained by the worse clinical profile of patients receiving DOACs. Risk of bleeding was rather high, and warrants close clinical monitoring. Ó 2018 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
The prevalence of atrial fibrillation (AF) increases with age, and reaches 10% in people 80 years or older living in the community.1e3
The authors declare no conflicts of interest. * Address correspondence to Carlos Rodríguez-Pascual, University of Lincoln, Joseph Banks Laboratories, Green Lane, Lincoln LN6 7DL, United Kingdom. E-mail address: [email protected] (C. Rodríguez-Pascual). https://doi.org/10.1016/j.jamda.2018.04.017 1525-8610/Ó 2018 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Atrial fibrillation is a clinically significant disorder that is associated with increased mortality and an important use of health care resources.4 Specifically, compared to patients without nonvalvular atrial fibrillation (NVAF), those with NVAF show a 4 to 5 times higher risk of thromboembolic stroke, which is even higher in the oldest patients.5e7 To address this risk, these patients should receive anticoagulant therapy.8e10
2
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
Vitamin K antagonists (VKAs) reduce stroke incidence by 64%, and this benefit is particularly important in older patients because of their higher basal stroke risk.11 However, VKAs have a narrow therapeutic window, a long half-life, and a variable dose-response relation. In addition, patients require regular monitoring, which may be a limiting factor for some older patients to commence this valuable therapeutic option. The annual rate of major bleeding in patients on VKAs is 2%-3%, and it is even higher in the oldest patients.12e14 As a result, fear of bleeding, particularly intracranial hemorrhage, has limited the use of VKAs in the oldest old and frail patients. Although 50% to 60% of patients with AF receive warfarin treatment, only 35% of those aged 85 years and older receive it.15e17 Among older patients participating in clinical trials, direct oral anticoagulants (DOACs) have shown similar or even greater efficacy than VKAs.18e21 However, results on safety are less certain. Specifically, in older patients there seems to be a higher risk of gastrointestinal bleeding with dabigatran22e25 and rivaroxaban19 than with warfarin; by contrast, compared with VKAs, apixaban and edoxaban have shown a significant reduction in major bleeding episodes, whereas bleeding risk might be similar with rivaroxaban.24 Moreover, it is not clear if older patients included in clinical trials had better health and lower risk of bleeding than those treated in routine clinical practice. Consequently, extrapolation of results from clinical trials to routine care of very old patients should be done with caution.26 To our knowledge, no previous prospective study has investigated the safety of DOACs in very old patients with NVAF and substantial comorbidity and disability who are managed in routine clinical practice. This is important because these patients represent a substantial daily practice challenge, as higher age and comorbidity are linked to increased thromboembolic and bleeding risk, as estimated by risk scores for stroke (eg, CHADS2, CHA2DS2VASC) and for bleeding complications of anticoagulation (eg, HASBLED, ATRIA, HEMORR2HAGES). Accordingly, this study assessed the safety of DOACs and dicumarol (a vitamin K antagonist) in NVAF geriatric patients. Methods Study Design and Participants We registered prospectively all patients who attended the outpatient clinic of the Geriatric Medicine Department at a tertiary hospital from January 1, 2010, to June 30, 2015. Inclusion criteria for this study were age over 74 years; evidence of AF in an electrocardiogram or 24-hour electrocardiographic recording, which was considered nonvalvular in origin according to a transthoracic echocardiography; and having started on anticoagulation in the previous 3 months. Exclusion criteria were (1) rheumatic mitral valve disease; (2) mechanical cardiac prosthesis; (3) previous treatment with oral anticoagulants; (4) treatment with aspirin or low-molecular-weight heparin for preventing stroke and systemic embolism; (5) estimated life expectancy shorter than 6 months; and (6) inability to attend follow-up visits at the outpatient clinic. Patients were followed until death or January 31, 2016, whichever came first. Study Variables Sociodemographic and biomedical variables were collected at the time of patient inclusion. Study outcomes were obtained during follow-up. Sociodemographic and biomedical variables Sociodemographic variables consisted of age, place of residence, cohabitation, and level of education. Among the biomedical variables was a geriatric assessment, which was performed by geriatricians or specialist nurses; it included assessment of cardiovascular risk factors, heart failure, ischemic heart disease, percutaneous transluminal
coronary angioplasty, previous heart surgery, pacemaker implantation, anemia, chronic obstructive pulmonary disease, chronic kidney disease, chronic liver disease, peptic ulcer, neoplasm, depression, osteoarthritis, fractures and falls in the last year, previous diagnosis of dementia, and the Charlson comorbidity index.27 Assessment also included disability in basic and instrumental activities of daily living using the Katz index28 and the Lawton and Brody index,29 respectively, cognitive function according to the Global Deterioration Scale for dementia,30 and mobility limitations per the following scale: 0 ¼ fully independent; 1 ¼ no walking limitation; 2 ¼ use of a stick; 3 ¼ use of a Zimmer frame or aid of one person; and 4 ¼ wheelchair bound or needing help from 2 or more people to walk. Biomedical variables also consisted of a physical examination to measure height, weight, blood pressure, heart rate, and presence of heart murmurs. Blood tests were performed at the time of study inclusion (before the initiation of anticoagulant therapy), at 6 and 12 months of treatment and included determinations of blood cells, coagulation time, and renal, liver, and thyroid function. The glomerular filtration rate was estimated using the Modified Diet in Renal Disease formula.31 Risks of thrombotic and hemorrhagic events were estimated using CHADS2,32 CHA2DS2VASc,33 and HASBLED34 scores. Lastly, we collected data on current drug treatments, with particular attention to antiplatelet agents, nonsteroidal anti-inflammatory drugs, steroids, and other drugs with potential interaction with anticoagulant therapy. Study outcomes Outcomes included vital status, bleeding episodes, and thrombotic events. This information was obtained through interviews with patients and a family member in regular outpatient clinical appointments, through review of the electronic medical chart, and telephone contact if necessary. Bleedings were grouped into intracranial, gastrointestinal, epistaxis, hematuria, and other hemorrhages. Bleeding episodes were also classified as major and nonmajor; major bleeding was defined as that associated with a drop of 2 to 4 g/dL in serum hemoglobin, or requiring transfusion of 1 to 2 bags of red blood cells, or lifethreatening bleeding (drop of >4 g/dL in hemoglobin, >2 bags of red blood cells, intracranial hemorrhage, or bleeding-related death). Thrombotic events included stroke, transient ischemic attack, acute coronary syndrome, deep venous thrombosis, and pulmonary embolism. Primary hemorrhagic strokes were classified under major bleeding episodes, but hemorrhagic transformation of cerebral infarct was included under thrombotic complications. Discontinuation of treatment was identified from patient and/or family interviews on follow-up appointments or telephone calls, from medication updates in the electronic chart and/or the electronic prescription system, or when medication was not retrieved from the pharmacy office. Statistical Analysis Data on continuous variables were reported as mean standard deviation, and for categorical variables as percentages. Student t test was used to assess differences between means, and the chi-square or Fisher exact test for differences between percentages. The risk of bleeding or thrombotic events was expressed as the number of individuals with those events per 100 patient-years of follow-up. The association between the type of anticoagulant treatment and bleeding during follow-up was summarized with hazard ratios (HRs) and their 95% confidence interval (CI) obtained from Cox regression, using patients treated with dicumarol as the referent group. We built 2 regression models: the first one was a crude model; and the second model was adjusted for the following variables that, in previous publications or in our study sample, were related to both anticoagulant treatment and risk of bleeding: age, sex, Charlson
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
comorbidity index, tobacco smoking, chronic renal failure, chronic obstructive pulmonary disease, anemia at inclusion, stroke, heart failure, previous falls, albumin, disability in activities of daily living (ADL), limitations in mobility, HASBLED score, and CHA2DS2VASC score. We also tested whether the association between DOACs vs dicumarol and any bleeding varied with age (85 years, <85 years), chronic renal failure, heart failure, HASBLED score (3 vs <3), Charlson comorbidity index (3 vs <3), treatment with antiplatelet agents, dependence in ADL (2 vs <2), limitation in mobility (severe vs none or moderate), and with the diagnosis of dementia. This was tested by assessing the statistical significance of interaction terms constructed as the product of DOAC treatment by those variables. Informed written consent was obtained from study participants. Statistical analyses were performed with SPSS v22 (IBM SPSS version 22.0, IBM Corp). Results Among the 554 participants, 193 received treatment with dicumarol, 203 with dabigatran, 112 with rivaroxaban, and 46 with apixaban. Prescribed doses for DOACs were the following: (1) 110 mg twice daily in 182 (89.6%) patients and 150 mg twice daily in 21 (10.4%) patients for dabigatran; (2) 15 mg once daily in 88 (78.6%) patients and 20 mg once daily in 24 (21.4%) patients for rivaroxaban; and (3) 2.5 mg twice daily in 34 (73.9%) patients and 5 mg twice daily in 12 (26.1%) patients for apixaban. Compared to patients on dicumarol, those under DOAC treatment were older, and had higher frequency of smoking, comorbidities, disability in ADL and instrumental ADL (only in men), and mobility limitations (Table 1). Patients on DOAC treatment also showed a higher mortality during follow-up (Table 1). Mean follow-up was 61.3 30.3 weeks, which resulted in 652.83 patient-years of observation; of these, 268.37 patient-years corresponded to dicumarol and 384.46 patient-years to DOACs (237.38 for dabigatran, 110.48 for rivaroxaban, and 36.60 for apixaban). The incidence of any bleeding was 19.2/100 person-years among patients on DOACs and 13.7/100 person-years on dicumarol; corresponding figures for major bleeding were 5.2 for those on DOACs and 3.3 for those on dicumarol (Table 2). Incidence of thrombotic events was low in patients on DOACs or dicumarol (Table 2). In crude analyses, hazard ratios (95% CI) for bleeding and for major bleeding in patients on DOACs versus dicumarol were 1.60 (1.04-2.44) and 2.22 (95% CI 0.88-5.59), respectively. Excess risk of bleeding associated with DOACs vs dicumarol disappeared after adjustment for clinical characteristics, yielding corresponding figures of 1.19 (0.682.08) and 1.01 (0.35-2.93) (Figure 1). These results did not vary across strata of high-risk patients, defined as age 85 years or older, chronic renal failure, heart failure, HASBLED score 3, Charlson comorbidity index 3, simultaneous treatment with antiplatelet agents, disability in 2 ADL, limitation in mobility, and diagnosis of dementia (P for interaction >.05 in all cases). Nonmajor bleeding with DOACs vs dicumarol showed a crude HR of 1.56 (0.99-2.46) and an adjusted HR of 1.46 (0.78-2.75); the corresponding figures for gastrointestinal bleeding were 2.17 (1.06-4.47) and 1.09 (0.46-2.62). However, dabigatran vs other anticoagulants showed a significant increase of gastrointestinal bleeding risk both in the crude (HR 3.25,1.73-6.12) and adjusted analysis (HR 2.27,1.10-7.70). Incidence (95% CI) of major bleeding was 3.3 (1.3%-5.3%), 5.4 (2.5%8.3%), 5.4 (1.1%-9.7%), and 2.7 (2.6%-8.2%) per 100 patient-years for dicumarol, dabigatran, rivaroxaban, and apixaban, respectively (Table 2). Figure 2 compares the incidence of major bleeding between our study and clinical trials with patients aged 75 years and older.18e20 Risk of major bleeding for dabigatran and rivaroxaban tended to be higher in our study than in trials, but lower for dicumarol (warfarin in
3
Table 1 Characteristics of Patients with Atrial Fibrillation, According to Treatment with Direct Oral Anticoagulants (DOAC) or Dicumarol All Patients DOAC (n ¼ 554) (n ¼ 361)
Dicumarol (n ¼ 193)
Age, y 84.3 4.9 85.4 4.6 82.3 4.8 Women 64.3 64.8 63.2 Classification of atrial fibrillation Paroxysmal 16.6 14.6 20.3 Persistent 5.2 5.1 5.3 Permanent 78.3 80.3 74.3 On antiplatelet agents 17.9 16.4 20.7 Charlson index 2.5 2.1 2.8 2.1 1.9 2 Current smoker 8.5 11.4 3.1 Dementia 9.4 10 8.3 Depression 14.4 16.3 10.9 Chronic renal disease 20.6 24.9 12.4 Hypertension 80.7 79.8 82.4 Chronic obstructive lung 21.1 24.4 15.0 disease Anemia 23.3 27.7 15 Diabetes 24.7 25.8 22.8 Peripheral vascular disease 7.4 8.3 5.7 Transient ischemic attack 6.3 7.5 4.1 Stroke 14.4 18.6 6.7 Coronary heart disease 15.7 16.6 14 Previous myocardial infarction 11.6 11.6 11.4 PCI 8.7 8.3 9.3 CABG 1.6 1.9 1 Heart failure 49.8 55.4 39.4 LVSD (LVEF 45%) 15.2 16.1 13.7 Previous falls 9.0 12.3 2.7 Hemoglobin, g/dL 12.7 1.7 12.5 1.7 13.1 1.6 Creatinine, mg/dL 1.1 0.3 1.1 0.3 1 0.3 MDRD, mL/min 62.4 23.3 61.1 23.4 65.2 23.0 Albumin, g/dL 3.8 0.4 3.8 0.4 4 0.4 Disability in ADL 45.8 52.9 29.8 Disability in IADL Women 82.6 84.7 78.6 Men 65.5 71.4 47.6 Need of mobility aids 49.3 56.9 32.4 CHADS2 2.9 1.1 3 1.1 2.6 1 CHA2DS2VASC 4.7 1.2 4.8 1.3 4.4 1.2 HASBLED 1.9 1 21 1.5 0.7 Transfusion during follow-up 7.2 8.4 4.8 Worsening renal function 12.1 12 12.2 during follow-up Mortality during follow-up 16.1 18.8 10.9
P Value <.001 .71
.22 .22 <.001 <.001 .52 .08 <.001 .50 .01 <.001 .44 .26 .12 <.001 .42 .93 .69 .42 <.001 .45 <.001 <.001 .07 .06 <.001 <.001 .38 .04 <.001 <.001 <.001 <.001 .16 1 .01
CABG, coronary artery bypass grafting; IADL, instrumental activities of daily living; LVEF, left ventricle ejection fraction; LVSD, left ventricle systolic dysfunction; MDRD, Modified Diet in Renal Disease; PCI, percutaneous coronary intervention. Data are mean standard deviation (SD) for continuous variables and percentages for categorical variables.
trials) and apixaban, although these differences did not reach statistical significance. The number of patients treated with apixaban in our study was rather small, and results might be difficult to interpret. Discussion In very old patients with NVAF, this study shows that the higher risk of bleeding associated with DOACs vs dicumarol could be mostly due to the worse clinical profile of patients receiving DOACs. Moreover, the risk of bleeding was rather high, and warrants close clinical monitoring. Older old and frail patients with AF receive oral anticoagulants less frequently than their younger and nonfrail counterparts. Among 207 patients with AF admitted to an acute care hospital and followed during 6 months, Perera et al35 showed that frail patients, as identified by the modified Edmonton frailty scale, received warfarin less frequently, and that frailty was associated with an increased risk of embolic stroke but
4
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
Table 2 Hemorrhagic and Thrombotic Complications According to Treatment with Direct Oral Anticoagulants (DOAC) or Dicumarol All Patients (n ¼ 554)
Any bleeding Life-threatening bleedingx Major bleedingk Nonmajor bleeding{ GI bleeding Intracranial bleeding TIA or stroke DVT or PE ACS
DOACs (n ¼ 361)
Dicumarol (n ¼ 193)
All
Dabigatran (n ¼ 203)*
Rivaroxaban (n ¼ 112)y
Apixaban (n ¼ 46)z
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
17 1.37 4.4 13.6 7.2 0.6 2 1.5 1.5
19.2 1.5 5.2 15.6 9.1 0.7 2.6 1.5 1.3
21.0 1.26 5.4 17.2 12.6 0.4 1.2 1.2 1.2
17.2 1.8 5.4 12.6 4.5 1.8 5.4 2.7 0.9
13.6 2.7 2.7 13.6 0.0 0.0 2.7 0.0 2.7
13.7 1.1 3.3 10.8 4.4 0.3 0.9 1.2 1.5
ACS, acute coronary syndrome; DVT, deep venous thrombosis; GI, gastrointestinal; p-y, patient-years; PE, pulmonary embolism TIA, transient ischemic attack. *Prescribed dose was 110 mg twice daily in 182 (89.6%) cases, and 150 mg twice daily in 21 (10.4%) cases. y Prescribed dose was 15 mg once daily in 88 (78.6%) cases and 20 mg once daily in 24 (21.4%) cases. z Prescribed dose was 2.5 mg twice daily in 34 (73.9%) cases and 5 mg twice daily in 12 (26.1%) cases. x Drop of >4 g/dL in serum hemoglobin, requiring transfusion of >2 bags of red blood cells, intracranial bleeding, or bleeding-related death. k Drop of 2-4 g/dL in serum hemoglobin, or requiring transfusion of 1-2 bags of red blood cells. { Epistaxis, hematuria, and other drop in serum hemoglobin or bleeding not requiring transfusion or hospital admission.
Fig. 1. Survival curves for any bleeding (left) and major bleeding (right) among those treated with direct oral anticoagulants (DOACs) or dicumarol. (A) Nonadjusted; (B) adjusted for age, sex, Charlson comorbidity index, tobacco smoking, chronic renal failure, chronic obstructive pulmonary disease, anemia at inclusion, stroke, heart failure, previous falls, albumin, disability in ADL, limitations in mobility, HASBLED score, and CHA2DS2VASC score. Please note that in adjusted analyses (B), the curves of survival without major bleeding overlap completely for DOACs and dicumarol.
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6 6%
5.40%
5.40%
5.19% 5%
4.37%
4.43% *
4.63%
4.40%
4%
3.33%
3.30% 3%
2.70% 2%
1%
0%
Warfarin * For 110 mg. dose
Dabigatran Current
RELY
Rivaroxaban ROCKET
Apixaban
ARISTOTLE
Fig. 2. Incidence of major bleeding (cases/100 patient-years) in the current study versus in clinical trials of patients aged 75 years and older.
not of bleeding. Fortunately, development of DOACs have made anticoagulant therapy easier for very old and frail patients; however, despite the fact that about 90%, 80%, and 75% of patients in our study were on low doses of dabigatran, rivaroxaban, and apixaban, respectively, the risk of bleeding was rather high, with 1 in 5 patients requiring medical attention, mainly because of nonmajor bleeding. This indicates that these very old patients deserve close clinical monitoring regardless of whether they are treated with low-dose DOACs or VKAs. Several randomized clinical trials and observational studies have examined the efficacy and bleeding risks of DOACs vs VKAs in older individuals. In a meta-analysis of the main phase-3 trials of DOACs vs warfarin in patients with AF, Ruff et al36 found that DOACs were more effective than warfarin in stroke prevention among patients aged 75 years, mainly because of reduction in hemorrhagic stroke, with no significant difference in major bleeding rates. Sardar et al,37 in a meta-analyses of 10 trials with patients aged 75 years with or without AF, reported that DOACs did not cause excess bleeding and were associated with equal or greater efficacy than conventional therapy. In a network meta-analysis of randomized and nonrandomized studies, Lin et al38 found a nonsignificant trend toward lower risk of stroke and of major bleeding with DOACs vs warfarin in patients aged 65-74 years; however, the efficacy of DOACs was attenuated, and there was an increased risk of major bleeding with dabigatran 110 mg, dabigatran 150 mg, and rivaroxaban, and also of gastrointestinal bleeding with dabigatran 150 mg, among patients aged 75 years. In another recent meta-regression analysis of studies in patients aged 65 years, Bai et al39 observed that DOACs were associated with a lower risk of major bleeding in 4 randomized trials (risk ratio 0.82, 95% CI 0.70e0.95), but not in 5 real-life studies (relative ratio 0.93, 95% CI 0.75e1.11), and that the benefit in major bleeding rates with DOACs in trials was attenuated by increasing age. Subgroup analyses in clinical trials and observational studies provide relevant information on variations in anticoagulation effectiveness and safety according to age. Nevertheless, exclusion criteria in clinical trials, such as reduced life expectancy in the investigator’s opinion, as well as the heterogeneity of older patients, limit generalization of results to very old and frail patients. Moreover, retrospective analyses of series of older patients from clinical practice have included
5
patients younger than ours and only provided limited data on relevant variables, such as physical function and cognition.38 Thus, our study is particularly relevant because it included very old patients with high prevalence of comorbidities and disability, who were attended in a routine clinical practice setting. Furthermore, we have adjusted the analyses for relevant variables in very old patients, such as ADL disability, mobility limitations, falls, previous stroke, renal dysfunction, or anemia. The only significant difference in bleeding episodes between individual DOACs in our patients was an increased risk of gastrointestinal bleeding with dabigatran, even though 90% of patients were on a low dose, that is, 110 mg twice daily. Although in the RELY trial the rate of gastrointestinal bleeding in patients >80 years under low-dose dabigatran was similar to that for warfarin,21 other studies have reported results in line with ours. Using administrative data, AvgilTsadok et al40 found a reduced risk of intracranial hemorrhage and higher rate of gastrointestinal bleeding with dabigatran vs warfarin in patients aged 75 years. Other analyses have also shown increased gastrointestinal bleeding risk of dabigatran and rivaroxaban vs VKAs in patients aged >75 years.41 Also using an administrative database, Abraham et al42 found that, compared with rivaroxaban and dabigatran, apixaban was the DOAC with the best gastrointestinal safety profile in patients aged 75 years. Most of our patients were on low doses of anticoagulants. This is relevant because the safety and efficacy of low doses of some DOACs have not been sufficiently assessed in very old patients. Indeed, only the lower dose of dabigatran (110 mg) was studied in the RELY trial. Also, doses of rivaroxaban and apixaban were reduced under predetermined criteria in the ROCKET and ARISTOTLE trials, respectively; although the safety was similar in patients 75 years and in the general trial population, only a small proportion of patients in the ARISTOTLE and ROCKET trials was on low-dose apixaban and rivaroxaban, respectively. A real world study in UK found that bleeding risk in patients 80 years was lower for apixaban 2.5 mg than for warfarin (HR: 0.78, 95% CI: 0.63-0.96) and not significantly different for dabigatran 110 mg (HR 0.89, 0.76-1.04) and rivaroxaban 15 mg (HR 1.13, 0.91-1.40).43 In this study, apixaban was associated with a nonsignificant higher risk of stroke or systemic embolism than warfarin (HR 1.15, 0.94-1.41) whereas the risk was lower for rivaroxaban 15 mg (HR 0.71, 0.52-0.95) and similar for dabigatran 110 mg (HR 0.98, 0.82-1.17). However, although patients on DOACs were older than those on warfarin, which was properly adjusted for in the data analysis, there was no information about the degree of frailty or disability in these patients. In our study, patients showed a nonstatistically significant higher risk of major bleeding than patients 75 years in major clinical trials.18e20 In fact, the rates of bleeding were high. Consequently, the introduction of anticoagulation in older patients with comorbidities and disabilities should be carefully considered. Our main study results did not vary with age or decreased renal function, heart failure, dementia, antiplatelet treatment, ADL disability, or severe mobility limitations. This indicates that a broad spectrum of patients could benefit from DOACs. The risk for bleeding seems to be similar with all types of oral anticoagulants and supports the need for individualized treatment.44 Some methodological issues in this study warrant comment. First, the study was conducted among outpatients from a Geriatric Medicine Department in a single hospital; hence, the results should be confirmed in other settings and hospitals. Second, we did not obtain information about time in therapeutic range among patients on dicumarol; thus, we could not adjust bleeding risk for time in therapeutic range. Third, we did not include patients without antithrombotic treatment or those on aspirin or low-weight heparins, who might have the highest risk of bleeding. In fact, the study included only patients whose physician in charge believed they could benefit
6
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
from oral anticoagulants. Notwithstanding this, the clinical characteristics of these patients corresponded to those routinely attended in geriatric departments. Finally, an administrative permit is required in Spain to prescribe DOACs for each individual patient, so DOAC treatment is restricted to patients with limitations for VKAs or with high risk of intracranial bleeding; thus, a subset of patients with a severe clinical profile may have been selected for DOACs. Future research in very old patients should focus on better defining risk factors for anticoagulant-associated bleeding. Also, research in daily clinical practice should establish the risk-benefit ratio of anticoagulation and define the limit where risks outweigh benefits. Finally, the type of DOAC that is most appropriate according to the clinical characteristics of each patient should be defined. References 1. Go AS, Hylek EM, Phillips KA, et al. 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 2001;285:2370e2375. 2. Lloyd-Jones DM, Wang TJ, Leip EP, et al. Lifetime risk for development of atrial fibrillation: The Framingham Heart Study. Circulation 2004;110: 1042e1046. 3. Miyasaka Y, Barnes ME, Gersh BJ, et al. Secular trends in incidence of atrial fibrillation in Olmsted County, Minnesota, 1980 to 2000, and implications on the projections for future prevalence. Circulation 2006;114:119e125. 4. Zoni-Berisso M, Lercari F, Carazza T, Domenicucci S. Epidemiology of atrial fibrillation: European perspective. Clin Epidemiol 2014;6:213e220. 5. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: The Framingham study. Stroke 1991;22:983e988. 6. Go AS, Mozaffarian D, Roger VL, et al. American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statisticsd2014 update: A report from the American Heart Association. Circulation 2014;129:e28ee292. 7. Coppens M, Eikelboom JW, Hart RG, et al. The CHA2DS2-VASc score identifies those patients with atrial fibrillation and a CHADS2 score of 1 who are unlikely to benefit from oral anticoagulant therapy. Eur Heart J 2013;34(3): 170e176. 8. Lip GY, Clementy N, Pericart L, et al. Stroke and major bleeding risk in elderly patients aged >75 years with atrial fibrillation: The Loire Valley atrial fibrillation project. Stroke 2015;46:143e150. 9. Singer DE, Chang Y, Fang MC, et al. The net clinical benefit of warfarin anticoagulation in atrial fibrillation. Ann Intern Med 2009;151:297e305. 10. Friberg L, Rosenqvist M, Lip GY. Net clinical benefit of warfarin in patients with atrial fibrillation: A report from the Swedish Atrial Fibrillation Cohort Study. Circulation 2012;125:2298e2307. 11. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 2007;146:857e867. 12. Van der Meer FJ, Rosendaal FR, Vandenbroucke JP, Briët E. Bleeding complications in oral anticoagulant therapy. An analysis of risk factors. Arch Intern Med 1993;153:1557e1562. 13. Hutten BA, Lensing AW, Kraaijenhagen RA, Prins MH. Safety of treatment with oral anticoagulants in the elderly. A systematic review. Drugs Aging 1999;14: 303e312. 14. Dodson JA, Petrone A, Gagnon DR, et al. Incidence and determinants of traumatic intracranial bleeding among older veterans receiving warfarin for atrial fibrillation. JAMA Cardiol 2016;1:65e72. 15. Go AS, Hylek EM, Borowsky LH, et al. Warfarin use among ambulatory patients with nonvalvular atrial fibrillation: The AnTicoagulation and Risk factors In Atrial fibrillation (ATRIA) study. Ann Intern Med 1999;131: 927e934. 16. Hylek EM, D’Antonio J, Evans-Molina C, et al. Translating the results of randomized trials into clinical practice: the challenge of warfarin candidacy among hospitalized elderly patients with atrial fibrillation. Stroke 2006;37: 1075e1080. 17. Nieuwlaat R, Capucci A, Lip GY, et al. Euro Heart Survey Investigators. Antithrombotic treatment in real-life atrial fibrillation patients: a report from the Euro Heart Survey on Atrial Fibrillation. Eur Heart J 2006;27: 3018e3026. 18. Eikelboom JW, Wallentin L, Connolly SJ, et al. Risk of bleeding with 2 doses of dabigatran compared with warfarin in older and younger patients with atrial fibrillation: an analysis of the Randomized Evaluation of Long-term anticoagulant therapY (RE-LY) trial. Circulation 2011;123:2363e2372.
19. Halperin JL, Hankey GJ, Wojdyla DM, et al. ROCKET AF Steering Committee and Investigators. Efficacy and safety of rivaroxaban compared with warfarin among elderly patients with nonvalvular atrial fibrillation in the Rivaroxaban Once Daily, Oral, Direct Factor Xa Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF). Circulation 2014;130:138e146. 20. Halvorsen S, Atar D, Yang H, et al. Efficacy and safety of apixaban compared with warfarin according to age for stroke prevention in atrial fibrillation: observations from the ARISTOTLE trial. Eur Heart J 2014;35:1864e1872. 21. Lauw MN, Eikelboom JW, Coppens M, et al. Effects of dabigatran according to age in atrial fibrillation. Heart 2017;103:1015e1023. 22. Lip GY, Lane DA. Stroke prevention with oral anticoagulation therapy in patients with atrial fibrillationdfocus on the elderly. Circ J 2013;77:1380e1388. 23. Barco S, Cheung YW, Eikelboom JW, Coppens M. New oral anticoagulants in elderly patients. Best Pract Res Clin Haematol 2013;26:215e224. 24. Sharma M, Cornelius VR, Patel JP, et al. Efficacy and harms of direct oral anticoagulants in the elderly for stroke prevention in atrial fibrillation and secondary prevention of venous thromboembolism: Systematic review and meta-analysis. Circulation 2015;132:194e204. 25. Romanelli RJ, Nolting L, Dolginsky M, et al. Dabigatran versus warfarin for atrial fibrillation in real-world clinical practice: A systematic review and metaanalysis. Circ Cardiovasc Qual Outcomes 2016;9:126e134. 26. Ho P, Brooy BL, Hayes L, et al. Direct oral anticoagulants in frail older adults: A geriatric perspective. Semin Thromb Hemost 2015;41:389e394. 27. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373e383. 28. Katz S, Down TD, Cash HR, Grotz RC. Progress in the development of the index of ADL. Gerontologist 1970;10:20e30. 29. Lawton MP, Brody EM. Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist 1969;9:179e186. 30. Reisberg B, Ferris SH, De Leon MJ, Crook T. The global deterioration scale for assessment of primary degenerative dementia. Am J Psychiatry 1982;139: 1136e1139. 31. Levey AS, Coresh J, Greene T, et al. Chronic kidney disease epidemiology collaboration. Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate. Ann Intern Med 2006;145:247e254. 32. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification schemes for predicting stroke: Results from the National Registry of Atrial Fibrillation. JAMA 2001;285:2864e2870. 33. Lip GYH, Nieuwlaat R, Pisters R, et al. 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;137:263e272. 34. Pisters R, Lane DA, Nieuwlaat R, et al. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: The Euro Heart Survey. Chest 2010;138:1093e1100. 35. Perera V, Bajorek BV, Matthews S, Hilmer SN. The impact of frailty on the utilization of antithrombotic therapy in older patients with atrial fibrillation. Age Ageing 2009;38:156e162. 36. Ruff CT, Giugliano RP, Braunwald E, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: A meta-analysis of randomised trials. Lancet 2014;383(9921):955e962. 37. Sardar P, Chatterjee S, Chaudhari S, Lip GY. New oral anticoagulants in elderly adults: Evidence from a meta-analysis of randomized trials. J Am Geriatr Soc 2014;62:857e864. 38. Lin L, Lim WS, Zhou HJ, et al. Clinical and safety outcomes of oral antithrombotics for stroke prevention in atrial fibrillation: A systematic review and network meta-analysis. J Am Med Dir Assoc 2015;16:1103. e1e1103.e19. 39. Bai Y, Guo SD, Deng H, et al. Effectiveness and safety of oral anticoagulants in older patients with atrial fibrillation: A systematic review and meta-regression analysis. Age Ageing 2017;5:767e773. 40. Avgil-Tsadok M, Jackevicius CA, Essebag V, et al. Dabigatran use in elderly patients with atrial fibrillation. Thromb Haemost 2016;115:152e160. 41. Abraham NS, Singh S, Alexander GC, et al. Comparative risk of gastrointestinal bleeding with dabigatran, rivaroxaban, and warfarin: Population based cohort study. BMJ 2015;350:h1857. 42. Abraham NS, Noseworthy PA, Yao X, et al. Gastrointestinal safety of direct oral anticoagulants: A large population-based study. Gastroenterology 2017;152: 1014e1022.e1. 43. Nielsen PB, Skjøth F, Søgaard M, et al. Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study. BMJ 2017; 356:j510. 44. Voukalis C, Lip GYH. Oral anticoagulation for elderly patients with non-valvular atrial fibrillation: Recent insights from randomised trials and the “real world”. Heart 2017;103:977e978.
JAMDA journal homepage: www.jamda.com
Original Study
Safety of Direct Oral Anticoagulants and Vitamin K Antagonists in Oldest Old Patients: A Prospective Study Carlos Rodríguez-Pascual MD, PhD a, b, *, Ivett Torres-Torres MD a, Alejandro Gómez-Quintanilla MD a, Ana Isabel Ferrero-Martínez NP a, Jagdish Sharma MD, PhD b, Alba Guitián c, Maria Carmen Basalo MD c, Marina Montero-Magan MD a, Arturo Vilches-Moraga MD a, d, Maria-Teresa Olcoz-Chiva MD a, b, Emilio Paredes-Galán MD c, Fernando Rodríguez-Artalejo MD, PhD e a
Geriatric Medicine Department, Complejo Hospitalario Universitario de Vigo, Spain University of Lincoln and Lincoln County Hospital, Lincoln, Lincolnshire, United Kingdom c Cardiology Department, Complejo Hospitalario Universitario de Vigo, Spain d Ageing and Complex Medicine Department, Salford Royal NHS Foundation Trust, Manchester, United Kingdom e Department of Preventive Medicine and Public Health, Universidad Autónoma de Madrid /IdiPaz, CIBERESP, and IMDEA-Food Institute, Madrid, Spain b
a b s t r a c t Keywords: Atrial fibrillation elderly anticoagulants safety
Objective: The safety of direct oral anticoagulants (DOACs) in oldest old patients with nonvalvular atrial fibrillation (NVAF) in daily clinical practice has not been systematically assessed. This study examined the safety of DOACs and dicumarol (a vitamin K antagonist) in NVAF geriatric patients. Design: Prospective study from January 2010 through June 2015, with follow-up through January 2016. Setting: Geriatric medicine department at a tertiary hospital. Participants: A total of 554 outpatients, 75 years or older, diagnosed of NVAF and starting oral anticoagulation. Measurements: The main outcome was bleeding, which was classified into major (including those life-threatening) and nonmajor episodes. Statistical analyses were performed with Cox regression. Results: A total of 351 patients received DOACs and 193 dicumarol. Patients on DOACs were older, with more frequent comorbidities, mobility limitation and disability in activities of daily living, as well as higher mortality, than those treated with dicumarol. The incidence of any bleeding was 19.2/100 personyears among patients on DOACs and 13.7/100 person-years on dicumarol; corresponding figures for major bleeding were 5.2 for those on DOACs, and 3.3 for those on dicumarol. In crude analyses, hazard ratios (95% confidence intervals) for any bleeding, and for mayor bleeding in patients on DOACs vs dicumarol were 1.60 (1.04-2.44) and 2.22 (0.88-5.59), respectively. Excess risk of bleeding associated with DOACs vs dicumarol disappeared after adjustment for clinical characteristics, so that corresponding figures were 1.19 (0.68-2.08) and 1.01 (0.35-2.93). Results did not vary across subgroups of high-risk patients. Conclusion: In very old patients with NVAF, the higher risk of bleeding associated with DOACs vs dicumarol could be mostly explained by the worse clinical profile of patients receiving DOACs. Risk of bleeding was rather high, and warrants close clinical monitoring. Ó 2018 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
The prevalence of atrial fibrillation (AF) increases with age, and reaches 10% in people 80 years or older living in the community.1e3
The authors declare no conflicts of interest. * Address correspondence to Carlos Rodríguez-Pascual, University of Lincoln, Joseph Banks Laboratories, Green Lane, Lincoln LN6 7DL, United Kingdom. E-mail address: [email protected] (C. Rodríguez-Pascual). https://doi.org/10.1016/j.jamda.2018.04.017 1525-8610/Ó 2018 AMDA e The Society for Post-Acute and Long-Term Care Medicine.
Atrial fibrillation is a clinically significant disorder that is associated with increased mortality and an important use of health care resources.4 Specifically, compared to patients without nonvalvular atrial fibrillation (NVAF), those with NVAF show a 4 to 5 times higher risk of thromboembolic stroke, which is even higher in the oldest patients.5e7 To address this risk, these patients should receive anticoagulant therapy.8e10
2
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
Vitamin K antagonists (VKAs) reduce stroke incidence by 64%, and this benefit is particularly important in older patients because of their higher basal stroke risk.11 However, VKAs have a narrow therapeutic window, a long half-life, and a variable dose-response relation. In addition, patients require regular monitoring, which may be a limiting factor for some older patients to commence this valuable therapeutic option. The annual rate of major bleeding in patients on VKAs is 2%-3%, and it is even higher in the oldest patients.12e14 As a result, fear of bleeding, particularly intracranial hemorrhage, has limited the use of VKAs in the oldest old and frail patients. Although 50% to 60% of patients with AF receive warfarin treatment, only 35% of those aged 85 years and older receive it.15e17 Among older patients participating in clinical trials, direct oral anticoagulants (DOACs) have shown similar or even greater efficacy than VKAs.18e21 However, results on safety are less certain. Specifically, in older patients there seems to be a higher risk of gastrointestinal bleeding with dabigatran22e25 and rivaroxaban19 than with warfarin; by contrast, compared with VKAs, apixaban and edoxaban have shown a significant reduction in major bleeding episodes, whereas bleeding risk might be similar with rivaroxaban.24 Moreover, it is not clear if older patients included in clinical trials had better health and lower risk of bleeding than those treated in routine clinical practice. Consequently, extrapolation of results from clinical trials to routine care of very old patients should be done with caution.26 To our knowledge, no previous prospective study has investigated the safety of DOACs in very old patients with NVAF and substantial comorbidity and disability who are managed in routine clinical practice. This is important because these patients represent a substantial daily practice challenge, as higher age and comorbidity are linked to increased thromboembolic and bleeding risk, as estimated by risk scores for stroke (eg, CHADS2, CHA2DS2VASC) and for bleeding complications of anticoagulation (eg, HASBLED, ATRIA, HEMORR2HAGES). Accordingly, this study assessed the safety of DOACs and dicumarol (a vitamin K antagonist) in NVAF geriatric patients. Methods Study Design and Participants We registered prospectively all patients who attended the outpatient clinic of the Geriatric Medicine Department at a tertiary hospital from January 1, 2010, to June 30, 2015. Inclusion criteria for this study were age over 74 years; evidence of AF in an electrocardiogram or 24-hour electrocardiographic recording, which was considered nonvalvular in origin according to a transthoracic echocardiography; and having started on anticoagulation in the previous 3 months. Exclusion criteria were (1) rheumatic mitral valve disease; (2) mechanical cardiac prosthesis; (3) previous treatment with oral anticoagulants; (4) treatment with aspirin or low-molecular-weight heparin for preventing stroke and systemic embolism; (5) estimated life expectancy shorter than 6 months; and (6) inability to attend follow-up visits at the outpatient clinic. Patients were followed until death or January 31, 2016, whichever came first. Study Variables Sociodemographic and biomedical variables were collected at the time of patient inclusion. Study outcomes were obtained during follow-up. Sociodemographic and biomedical variables Sociodemographic variables consisted of age, place of residence, cohabitation, and level of education. Among the biomedical variables was a geriatric assessment, which was performed by geriatricians or specialist nurses; it included assessment of cardiovascular risk factors, heart failure, ischemic heart disease, percutaneous transluminal
coronary angioplasty, previous heart surgery, pacemaker implantation, anemia, chronic obstructive pulmonary disease, chronic kidney disease, chronic liver disease, peptic ulcer, neoplasm, depression, osteoarthritis, fractures and falls in the last year, previous diagnosis of dementia, and the Charlson comorbidity index.27 Assessment also included disability in basic and instrumental activities of daily living using the Katz index28 and the Lawton and Brody index,29 respectively, cognitive function according to the Global Deterioration Scale for dementia,30 and mobility limitations per the following scale: 0 ¼ fully independent; 1 ¼ no walking limitation; 2 ¼ use of a stick; 3 ¼ use of a Zimmer frame or aid of one person; and 4 ¼ wheelchair bound or needing help from 2 or more people to walk. Biomedical variables also consisted of a physical examination to measure height, weight, blood pressure, heart rate, and presence of heart murmurs. Blood tests were performed at the time of study inclusion (before the initiation of anticoagulant therapy), at 6 and 12 months of treatment and included determinations of blood cells, coagulation time, and renal, liver, and thyroid function. The glomerular filtration rate was estimated using the Modified Diet in Renal Disease formula.31 Risks of thrombotic and hemorrhagic events were estimated using CHADS2,32 CHA2DS2VASc,33 and HASBLED34 scores. Lastly, we collected data on current drug treatments, with particular attention to antiplatelet agents, nonsteroidal anti-inflammatory drugs, steroids, and other drugs with potential interaction with anticoagulant therapy. Study outcomes Outcomes included vital status, bleeding episodes, and thrombotic events. This information was obtained through interviews with patients and a family member in regular outpatient clinical appointments, through review of the electronic medical chart, and telephone contact if necessary. Bleedings were grouped into intracranial, gastrointestinal, epistaxis, hematuria, and other hemorrhages. Bleeding episodes were also classified as major and nonmajor; major bleeding was defined as that associated with a drop of 2 to 4 g/dL in serum hemoglobin, or requiring transfusion of 1 to 2 bags of red blood cells, or lifethreatening bleeding (drop of >4 g/dL in hemoglobin, >2 bags of red blood cells, intracranial hemorrhage, or bleeding-related death). Thrombotic events included stroke, transient ischemic attack, acute coronary syndrome, deep venous thrombosis, and pulmonary embolism. Primary hemorrhagic strokes were classified under major bleeding episodes, but hemorrhagic transformation of cerebral infarct was included under thrombotic complications. Discontinuation of treatment was identified from patient and/or family interviews on follow-up appointments or telephone calls, from medication updates in the electronic chart and/or the electronic prescription system, or when medication was not retrieved from the pharmacy office. Statistical Analysis Data on continuous variables were reported as mean standard deviation, and for categorical variables as percentages. Student t test was used to assess differences between means, and the chi-square or Fisher exact test for differences between percentages. The risk of bleeding or thrombotic events was expressed as the number of individuals with those events per 100 patient-years of follow-up. The association between the type of anticoagulant treatment and bleeding during follow-up was summarized with hazard ratios (HRs) and their 95% confidence interval (CI) obtained from Cox regression, using patients treated with dicumarol as the referent group. We built 2 regression models: the first one was a crude model; and the second model was adjusted for the following variables that, in previous publications or in our study sample, were related to both anticoagulant treatment and risk of bleeding: age, sex, Charlson
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
comorbidity index, tobacco smoking, chronic renal failure, chronic obstructive pulmonary disease, anemia at inclusion, stroke, heart failure, previous falls, albumin, disability in activities of daily living (ADL), limitations in mobility, HASBLED score, and CHA2DS2VASC score. We also tested whether the association between DOACs vs dicumarol and any bleeding varied with age (85 years, <85 years), chronic renal failure, heart failure, HASBLED score (3 vs <3), Charlson comorbidity index (3 vs <3), treatment with antiplatelet agents, dependence in ADL (2 vs <2), limitation in mobility (severe vs none or moderate), and with the diagnosis of dementia. This was tested by assessing the statistical significance of interaction terms constructed as the product of DOAC treatment by those variables. Informed written consent was obtained from study participants. Statistical analyses were performed with SPSS v22 (IBM SPSS version 22.0, IBM Corp). Results Among the 554 participants, 193 received treatment with dicumarol, 203 with dabigatran, 112 with rivaroxaban, and 46 with apixaban. Prescribed doses for DOACs were the following: (1) 110 mg twice daily in 182 (89.6%) patients and 150 mg twice daily in 21 (10.4%) patients for dabigatran; (2) 15 mg once daily in 88 (78.6%) patients and 20 mg once daily in 24 (21.4%) patients for rivaroxaban; and (3) 2.5 mg twice daily in 34 (73.9%) patients and 5 mg twice daily in 12 (26.1%) patients for apixaban. Compared to patients on dicumarol, those under DOAC treatment were older, and had higher frequency of smoking, comorbidities, disability in ADL and instrumental ADL (only in men), and mobility limitations (Table 1). Patients on DOAC treatment also showed a higher mortality during follow-up (Table 1). Mean follow-up was 61.3 30.3 weeks, which resulted in 652.83 patient-years of observation; of these, 268.37 patient-years corresponded to dicumarol and 384.46 patient-years to DOACs (237.38 for dabigatran, 110.48 for rivaroxaban, and 36.60 for apixaban). The incidence of any bleeding was 19.2/100 person-years among patients on DOACs and 13.7/100 person-years on dicumarol; corresponding figures for major bleeding were 5.2 for those on DOACs and 3.3 for those on dicumarol (Table 2). Incidence of thrombotic events was low in patients on DOACs or dicumarol (Table 2). In crude analyses, hazard ratios (95% CI) for bleeding and for major bleeding in patients on DOACs versus dicumarol were 1.60 (1.04-2.44) and 2.22 (95% CI 0.88-5.59), respectively. Excess risk of bleeding associated with DOACs vs dicumarol disappeared after adjustment for clinical characteristics, yielding corresponding figures of 1.19 (0.682.08) and 1.01 (0.35-2.93) (Figure 1). These results did not vary across strata of high-risk patients, defined as age 85 years or older, chronic renal failure, heart failure, HASBLED score 3, Charlson comorbidity index 3, simultaneous treatment with antiplatelet agents, disability in 2 ADL, limitation in mobility, and diagnosis of dementia (P for interaction >.05 in all cases). Nonmajor bleeding with DOACs vs dicumarol showed a crude HR of 1.56 (0.99-2.46) and an adjusted HR of 1.46 (0.78-2.75); the corresponding figures for gastrointestinal bleeding were 2.17 (1.06-4.47) and 1.09 (0.46-2.62). However, dabigatran vs other anticoagulants showed a significant increase of gastrointestinal bleeding risk both in the crude (HR 3.25,1.73-6.12) and adjusted analysis (HR 2.27,1.10-7.70). Incidence (95% CI) of major bleeding was 3.3 (1.3%-5.3%), 5.4 (2.5%8.3%), 5.4 (1.1%-9.7%), and 2.7 (2.6%-8.2%) per 100 patient-years for dicumarol, dabigatran, rivaroxaban, and apixaban, respectively (Table 2). Figure 2 compares the incidence of major bleeding between our study and clinical trials with patients aged 75 years and older.18e20 Risk of major bleeding for dabigatran and rivaroxaban tended to be higher in our study than in trials, but lower for dicumarol (warfarin in
3
Table 1 Characteristics of Patients with Atrial Fibrillation, According to Treatment with Direct Oral Anticoagulants (DOAC) or Dicumarol All Patients DOAC (n ¼ 554) (n ¼ 361)
Dicumarol (n ¼ 193)
Age, y 84.3 4.9 85.4 4.6 82.3 4.8 Women 64.3 64.8 63.2 Classification of atrial fibrillation Paroxysmal 16.6 14.6 20.3 Persistent 5.2 5.1 5.3 Permanent 78.3 80.3 74.3 On antiplatelet agents 17.9 16.4 20.7 Charlson index 2.5 2.1 2.8 2.1 1.9 2 Current smoker 8.5 11.4 3.1 Dementia 9.4 10 8.3 Depression 14.4 16.3 10.9 Chronic renal disease 20.6 24.9 12.4 Hypertension 80.7 79.8 82.4 Chronic obstructive lung 21.1 24.4 15.0 disease Anemia 23.3 27.7 15 Diabetes 24.7 25.8 22.8 Peripheral vascular disease 7.4 8.3 5.7 Transient ischemic attack 6.3 7.5 4.1 Stroke 14.4 18.6 6.7 Coronary heart disease 15.7 16.6 14 Previous myocardial infarction 11.6 11.6 11.4 PCI 8.7 8.3 9.3 CABG 1.6 1.9 1 Heart failure 49.8 55.4 39.4 LVSD (LVEF 45%) 15.2 16.1 13.7 Previous falls 9.0 12.3 2.7 Hemoglobin, g/dL 12.7 1.7 12.5 1.7 13.1 1.6 Creatinine, mg/dL 1.1 0.3 1.1 0.3 1 0.3 MDRD, mL/min 62.4 23.3 61.1 23.4 65.2 23.0 Albumin, g/dL 3.8 0.4 3.8 0.4 4 0.4 Disability in ADL 45.8 52.9 29.8 Disability in IADL Women 82.6 84.7 78.6 Men 65.5 71.4 47.6 Need of mobility aids 49.3 56.9 32.4 CHADS2 2.9 1.1 3 1.1 2.6 1 CHA2DS2VASC 4.7 1.2 4.8 1.3 4.4 1.2 HASBLED 1.9 1 21 1.5 0.7 Transfusion during follow-up 7.2 8.4 4.8 Worsening renal function 12.1 12 12.2 during follow-up Mortality during follow-up 16.1 18.8 10.9
P Value <.001 .71
.22 .22 <.001 <.001 .52 .08 <.001 .50 .01 <.001 .44 .26 .12 <.001 .42 .93 .69 .42 <.001 .45 <.001 <.001 .07 .06 <.001 <.001 .38 .04 <.001 <.001 <.001 <.001 .16 1 .01
CABG, coronary artery bypass grafting; IADL, instrumental activities of daily living; LVEF, left ventricle ejection fraction; LVSD, left ventricle systolic dysfunction; MDRD, Modified Diet in Renal Disease; PCI, percutaneous coronary intervention. Data are mean standard deviation (SD) for continuous variables and percentages for categorical variables.
trials) and apixaban, although these differences did not reach statistical significance. The number of patients treated with apixaban in our study was rather small, and results might be difficult to interpret. Discussion In very old patients with NVAF, this study shows that the higher risk of bleeding associated with DOACs vs dicumarol could be mostly due to the worse clinical profile of patients receiving DOACs. Moreover, the risk of bleeding was rather high, and warrants close clinical monitoring. Older old and frail patients with AF receive oral anticoagulants less frequently than their younger and nonfrail counterparts. Among 207 patients with AF admitted to an acute care hospital and followed during 6 months, Perera et al35 showed that frail patients, as identified by the modified Edmonton frailty scale, received warfarin less frequently, and that frailty was associated with an increased risk of embolic stroke but
4
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
Table 2 Hemorrhagic and Thrombotic Complications According to Treatment with Direct Oral Anticoagulants (DOAC) or Dicumarol All Patients (n ¼ 554)
Any bleeding Life-threatening bleedingx Major bleedingk Nonmajor bleeding{ GI bleeding Intracranial bleeding TIA or stroke DVT or PE ACS
DOACs (n ¼ 361)
Dicumarol (n ¼ 193)
All
Dabigatran (n ¼ 203)*
Rivaroxaban (n ¼ 112)y
Apixaban (n ¼ 46)z
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
Cases/100 p-y
17 1.37 4.4 13.6 7.2 0.6 2 1.5 1.5
19.2 1.5 5.2 15.6 9.1 0.7 2.6 1.5 1.3
21.0 1.26 5.4 17.2 12.6 0.4 1.2 1.2 1.2
17.2 1.8 5.4 12.6 4.5 1.8 5.4 2.7 0.9
13.6 2.7 2.7 13.6 0.0 0.0 2.7 0.0 2.7
13.7 1.1 3.3 10.8 4.4 0.3 0.9 1.2 1.5
ACS, acute coronary syndrome; DVT, deep venous thrombosis; GI, gastrointestinal; p-y, patient-years; PE, pulmonary embolism TIA, transient ischemic attack. *Prescribed dose was 110 mg twice daily in 182 (89.6%) cases, and 150 mg twice daily in 21 (10.4%) cases. y Prescribed dose was 15 mg once daily in 88 (78.6%) cases and 20 mg once daily in 24 (21.4%) cases. z Prescribed dose was 2.5 mg twice daily in 34 (73.9%) cases and 5 mg twice daily in 12 (26.1%) cases. x Drop of >4 g/dL in serum hemoglobin, requiring transfusion of >2 bags of red blood cells, intracranial bleeding, or bleeding-related death. k Drop of 2-4 g/dL in serum hemoglobin, or requiring transfusion of 1-2 bags of red blood cells. { Epistaxis, hematuria, and other drop in serum hemoglobin or bleeding not requiring transfusion or hospital admission.
Fig. 1. Survival curves for any bleeding (left) and major bleeding (right) among those treated with direct oral anticoagulants (DOACs) or dicumarol. (A) Nonadjusted; (B) adjusted for age, sex, Charlson comorbidity index, tobacco smoking, chronic renal failure, chronic obstructive pulmonary disease, anemia at inclusion, stroke, heart failure, previous falls, albumin, disability in ADL, limitations in mobility, HASBLED score, and CHA2DS2VASC score. Please note that in adjusted analyses (B), the curves of survival without major bleeding overlap completely for DOACs and dicumarol.
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6 6%
5.40%
5.40%
5.19% 5%
4.37%
4.43% *
4.63%
4.40%
4%
3.33%
3.30% 3%
2.70% 2%
1%
0%
Warfarin * For 110 mg. dose
Dabigatran Current
RELY
Rivaroxaban ROCKET
Apixaban
ARISTOTLE
Fig. 2. Incidence of major bleeding (cases/100 patient-years) in the current study versus in clinical trials of patients aged 75 years and older.
not of bleeding. Fortunately, development of DOACs have made anticoagulant therapy easier for very old and frail patients; however, despite the fact that about 90%, 80%, and 75% of patients in our study were on low doses of dabigatran, rivaroxaban, and apixaban, respectively, the risk of bleeding was rather high, with 1 in 5 patients requiring medical attention, mainly because of nonmajor bleeding. This indicates that these very old patients deserve close clinical monitoring regardless of whether they are treated with low-dose DOACs or VKAs. Several randomized clinical trials and observational studies have examined the efficacy and bleeding risks of DOACs vs VKAs in older individuals. In a meta-analysis of the main phase-3 trials of DOACs vs warfarin in patients with AF, Ruff et al36 found that DOACs were more effective than warfarin in stroke prevention among patients aged 75 years, mainly because of reduction in hemorrhagic stroke, with no significant difference in major bleeding rates. Sardar et al,37 in a meta-analyses of 10 trials with patients aged 75 years with or without AF, reported that DOACs did not cause excess bleeding and were associated with equal or greater efficacy than conventional therapy. In a network meta-analysis of randomized and nonrandomized studies, Lin et al38 found a nonsignificant trend toward lower risk of stroke and of major bleeding with DOACs vs warfarin in patients aged 65-74 years; however, the efficacy of DOACs was attenuated, and there was an increased risk of major bleeding with dabigatran 110 mg, dabigatran 150 mg, and rivaroxaban, and also of gastrointestinal bleeding with dabigatran 150 mg, among patients aged 75 years. In another recent meta-regression analysis of studies in patients aged 65 years, Bai et al39 observed that DOACs were associated with a lower risk of major bleeding in 4 randomized trials (risk ratio 0.82, 95% CI 0.70e0.95), but not in 5 real-life studies (relative ratio 0.93, 95% CI 0.75e1.11), and that the benefit in major bleeding rates with DOACs in trials was attenuated by increasing age. Subgroup analyses in clinical trials and observational studies provide relevant information on variations in anticoagulation effectiveness and safety according to age. Nevertheless, exclusion criteria in clinical trials, such as reduced life expectancy in the investigator’s opinion, as well as the heterogeneity of older patients, limit generalization of results to very old and frail patients. Moreover, retrospective analyses of series of older patients from clinical practice have included
5
patients younger than ours and only provided limited data on relevant variables, such as physical function and cognition.38 Thus, our study is particularly relevant because it included very old patients with high prevalence of comorbidities and disability, who were attended in a routine clinical practice setting. Furthermore, we have adjusted the analyses for relevant variables in very old patients, such as ADL disability, mobility limitations, falls, previous stroke, renal dysfunction, or anemia. The only significant difference in bleeding episodes between individual DOACs in our patients was an increased risk of gastrointestinal bleeding with dabigatran, even though 90% of patients were on a low dose, that is, 110 mg twice daily. Although in the RELY trial the rate of gastrointestinal bleeding in patients >80 years under low-dose dabigatran was similar to that for warfarin,21 other studies have reported results in line with ours. Using administrative data, AvgilTsadok et al40 found a reduced risk of intracranial hemorrhage and higher rate of gastrointestinal bleeding with dabigatran vs warfarin in patients aged 75 years. Other analyses have also shown increased gastrointestinal bleeding risk of dabigatran and rivaroxaban vs VKAs in patients aged >75 years.41 Also using an administrative database, Abraham et al42 found that, compared with rivaroxaban and dabigatran, apixaban was the DOAC with the best gastrointestinal safety profile in patients aged 75 years. Most of our patients were on low doses of anticoagulants. This is relevant because the safety and efficacy of low doses of some DOACs have not been sufficiently assessed in very old patients. Indeed, only the lower dose of dabigatran (110 mg) was studied in the RELY trial. Also, doses of rivaroxaban and apixaban were reduced under predetermined criteria in the ROCKET and ARISTOTLE trials, respectively; although the safety was similar in patients 75 years and in the general trial population, only a small proportion of patients in the ARISTOTLE and ROCKET trials was on low-dose apixaban and rivaroxaban, respectively. A real world study in UK found that bleeding risk in patients 80 years was lower for apixaban 2.5 mg than for warfarin (HR: 0.78, 95% CI: 0.63-0.96) and not significantly different for dabigatran 110 mg (HR 0.89, 0.76-1.04) and rivaroxaban 15 mg (HR 1.13, 0.91-1.40).43 In this study, apixaban was associated with a nonsignificant higher risk of stroke or systemic embolism than warfarin (HR 1.15, 0.94-1.41) whereas the risk was lower for rivaroxaban 15 mg (HR 0.71, 0.52-0.95) and similar for dabigatran 110 mg (HR 0.98, 0.82-1.17). However, although patients on DOACs were older than those on warfarin, which was properly adjusted for in the data analysis, there was no information about the degree of frailty or disability in these patients. In our study, patients showed a nonstatistically significant higher risk of major bleeding than patients 75 years in major clinical trials.18e20 In fact, the rates of bleeding were high. Consequently, the introduction of anticoagulation in older patients with comorbidities and disabilities should be carefully considered. Our main study results did not vary with age or decreased renal function, heart failure, dementia, antiplatelet treatment, ADL disability, or severe mobility limitations. This indicates that a broad spectrum of patients could benefit from DOACs. The risk for bleeding seems to be similar with all types of oral anticoagulants and supports the need for individualized treatment.44 Some methodological issues in this study warrant comment. First, the study was conducted among outpatients from a Geriatric Medicine Department in a single hospital; hence, the results should be confirmed in other settings and hospitals. Second, we did not obtain information about time in therapeutic range among patients on dicumarol; thus, we could not adjust bleeding risk for time in therapeutic range. Third, we did not include patients without antithrombotic treatment or those on aspirin or low-weight heparins, who might have the highest risk of bleeding. In fact, the study included only patients whose physician in charge believed they could benefit
6
C. Rodríguez-Pascual et al. / JAMDA xxx (2018) 1e6
from oral anticoagulants. Notwithstanding this, the clinical characteristics of these patients corresponded to those routinely attended in geriatric departments. Finally, an administrative permit is required in Spain to prescribe DOACs for each individual patient, so DOAC treatment is restricted to patients with limitations for VKAs or with high risk of intracranial bleeding; thus, a subset of patients with a severe clinical profile may have been selected for DOACs. Future research in very old patients should focus on better defining risk factors for anticoagulant-associated bleeding. Also, research in daily clinical practice should establish the risk-benefit ratio of anticoagulation and define the limit where risks outweigh benefits. Finally, the type of DOAC that is most appropriate according to the clinical characteristics of each patient should be defined. References 1. Go AS, Hylek EM, Phillips KA, et al. 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 2001;285:2370e2375. 2. Lloyd-Jones DM, Wang TJ, Leip EP, et al. Lifetime risk for development of atrial fibrillation: The Framingham Heart Study. Circulation 2004;110: 1042e1046. 3. Miyasaka Y, Barnes ME, Gersh BJ, et al. Secular trends in incidence of atrial fibrillation in Olmsted County, Minnesota, 1980 to 2000, and implications on the projections for future prevalence. Circulation 2006;114:119e125. 4. Zoni-Berisso M, Lercari F, Carazza T, Domenicucci S. Epidemiology of atrial fibrillation: European perspective. Clin Epidemiol 2014;6:213e220. 5. Wolf PA, Abbott RD, Kannel WB. Atrial fibrillation as an independent risk factor for stroke: The Framingham study. Stroke 1991;22:983e988. 6. Go AS, Mozaffarian D, Roger VL, et al. American Heart Association Statistics Committee and Stroke Statistics Subcommittee. Heart disease and stroke statisticsd2014 update: A report from the American Heart Association. Circulation 2014;129:e28ee292. 7. Coppens M, Eikelboom JW, Hart RG, et al. The CHA2DS2-VASc score identifies those patients with atrial fibrillation and a CHADS2 score of 1 who are unlikely to benefit from oral anticoagulant therapy. Eur Heart J 2013;34(3): 170e176. 8. Lip GY, Clementy N, Pericart L, et al. Stroke and major bleeding risk in elderly patients aged >75 years with atrial fibrillation: The Loire Valley atrial fibrillation project. Stroke 2015;46:143e150. 9. Singer DE, Chang Y, Fang MC, et al. The net clinical benefit of warfarin anticoagulation in atrial fibrillation. Ann Intern Med 2009;151:297e305. 10. Friberg L, Rosenqvist M, Lip GY. Net clinical benefit of warfarin in patients with atrial fibrillation: A report from the Swedish Atrial Fibrillation Cohort Study. Circulation 2012;125:2298e2307. 11. Hart RG, Pearce LA, Aguilar MI. Meta-analysis: antithrombotic therapy to prevent stroke in patients who have nonvalvular atrial fibrillation. Ann Intern Med 2007;146:857e867. 12. Van der Meer FJ, Rosendaal FR, Vandenbroucke JP, Briët E. Bleeding complications in oral anticoagulant therapy. An analysis of risk factors. Arch Intern Med 1993;153:1557e1562. 13. Hutten BA, Lensing AW, Kraaijenhagen RA, Prins MH. Safety of treatment with oral anticoagulants in the elderly. A systematic review. Drugs Aging 1999;14: 303e312. 14. Dodson JA, Petrone A, Gagnon DR, et al. Incidence and determinants of traumatic intracranial bleeding among older veterans receiving warfarin for atrial fibrillation. JAMA Cardiol 2016;1:65e72. 15. Go AS, Hylek EM, Borowsky LH, et al. Warfarin use among ambulatory patients with nonvalvular atrial fibrillation: The AnTicoagulation and Risk factors In Atrial fibrillation (ATRIA) study. Ann Intern Med 1999;131: 927e934. 16. Hylek EM, D’Antonio J, Evans-Molina C, et al. Translating the results of randomized trials into clinical practice: the challenge of warfarin candidacy among hospitalized elderly patients with atrial fibrillation. Stroke 2006;37: 1075e1080. 17. Nieuwlaat R, Capucci A, Lip GY, et al. Euro Heart Survey Investigators. Antithrombotic treatment in real-life atrial fibrillation patients: a report from the Euro Heart Survey on Atrial Fibrillation. Eur Heart J 2006;27: 3018e3026. 18. Eikelboom JW, Wallentin L, Connolly SJ, et al. Risk of bleeding with 2 doses of dabigatran compared with warfarin in older and younger patients with atrial fibrillation: an analysis of the Randomized Evaluation of Long-term anticoagulant therapY (RE-LY) trial. Circulation 2011;123:2363e2372.
19. Halperin JL, Hankey GJ, Wojdyla DM, et al. ROCKET AF Steering Committee and Investigators. Efficacy and safety of rivaroxaban compared with warfarin among elderly patients with nonvalvular atrial fibrillation in the Rivaroxaban Once Daily, Oral, Direct Factor Xa Inhibition Compared With Vitamin K Antagonism for Prevention of Stroke and Embolism Trial in Atrial Fibrillation (ROCKET AF). Circulation 2014;130:138e146. 20. Halvorsen S, Atar D, Yang H, et al. Efficacy and safety of apixaban compared with warfarin according to age for stroke prevention in atrial fibrillation: observations from the ARISTOTLE trial. Eur Heart J 2014;35:1864e1872. 21. Lauw MN, Eikelboom JW, Coppens M, et al. Effects of dabigatran according to age in atrial fibrillation. Heart 2017;103:1015e1023. 22. Lip GY, Lane DA. Stroke prevention with oral anticoagulation therapy in patients with atrial fibrillationdfocus on the elderly. Circ J 2013;77:1380e1388. 23. Barco S, Cheung YW, Eikelboom JW, Coppens M. New oral anticoagulants in elderly patients. Best Pract Res Clin Haematol 2013;26:215e224. 24. Sharma M, Cornelius VR, Patel JP, et al. Efficacy and harms of direct oral anticoagulants in the elderly for stroke prevention in atrial fibrillation and secondary prevention of venous thromboembolism: Systematic review and meta-analysis. Circulation 2015;132:194e204. 25. Romanelli RJ, Nolting L, Dolginsky M, et al. Dabigatran versus warfarin for atrial fibrillation in real-world clinical practice: A systematic review and metaanalysis. Circ Cardiovasc Qual Outcomes 2016;9:126e134. 26. Ho P, Brooy BL, Hayes L, et al. Direct oral anticoagulants in frail older adults: A geriatric perspective. Semin Thromb Hemost 2015;41:389e394. 27. Charlson ME, Pompei P, Ales KL, MacKenzie CR. A new method of classifying prognostic comorbidity in longitudinal studies: Development and validation. J Chronic Dis 1987;40:373e383. 28. Katz S, Down TD, Cash HR, Grotz RC. Progress in the development of the index of ADL. Gerontologist 1970;10:20e30. 29. Lawton MP, Brody EM. Assessment of older people: Self-maintaining and instrumental activities of daily living. Gerontologist 1969;9:179e186. 30. Reisberg B, Ferris SH, De Leon MJ, Crook T. The global deterioration scale for assessment of primary degenerative dementia. Am J Psychiatry 1982;139: 1136e1139. 31. Levey AS, Coresh J, Greene T, et al. Chronic kidney disease epidemiology collaboration. Using standardized serum creatinine values in the Modification of Diet in Renal Disease study equation for estimating glomerular filtration rate. Ann Intern Med 2006;145:247e254. 32. Gage BF, Waterman AD, Shannon W, et al. Validation of clinical classification schemes for predicting stroke: Results from the National Registry of Atrial Fibrillation. JAMA 2001;285:2864e2870. 33. Lip GYH, Nieuwlaat R, Pisters R, et al. 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;137:263e272. 34. Pisters R, Lane DA, Nieuwlaat R, et al. A novel user-friendly score (HAS-BLED) to assess 1-year risk of major bleeding in patients with atrial fibrillation: The Euro Heart Survey. Chest 2010;138:1093e1100. 35. Perera V, Bajorek BV, Matthews S, Hilmer SN. The impact of frailty on the utilization of antithrombotic therapy in older patients with atrial fibrillation. Age Ageing 2009;38:156e162. 36. Ruff CT, Giugliano RP, Braunwald E, et al. Comparison of the efficacy and safety of new oral anticoagulants with warfarin in patients with atrial fibrillation: A meta-analysis of randomised trials. Lancet 2014;383(9921):955e962. 37. Sardar P, Chatterjee S, Chaudhari S, Lip GY. New oral anticoagulants in elderly adults: Evidence from a meta-analysis of randomized trials. J Am Geriatr Soc 2014;62:857e864. 38. Lin L, Lim WS, Zhou HJ, et al. Clinical and safety outcomes of oral antithrombotics for stroke prevention in atrial fibrillation: A systematic review and network meta-analysis. J Am Med Dir Assoc 2015;16:1103. e1e1103.e19. 39. Bai Y, Guo SD, Deng H, et al. Effectiveness and safety of oral anticoagulants in older patients with atrial fibrillation: A systematic review and meta-regression analysis. Age Ageing 2017;5:767e773. 40. Avgil-Tsadok M, Jackevicius CA, Essebag V, et al. Dabigatran use in elderly patients with atrial fibrillation. Thromb Haemost 2016;115:152e160. 41. Abraham NS, Singh S, Alexander GC, et al. Comparative risk of gastrointestinal bleeding with dabigatran, rivaroxaban, and warfarin: Population based cohort study. BMJ 2015;350:h1857. 42. Abraham NS, Noseworthy PA, Yao X, et al. Gastrointestinal safety of direct oral anticoagulants: A large population-based study. Gastroenterology 2017;152: 1014e1022.e1. 43. Nielsen PB, Skjøth F, Søgaard M, et al. Effectiveness and safety of reduced dose non-vitamin K antagonist oral anticoagulants and warfarin in patients with atrial fibrillation: propensity weighted nationwide cohort study. BMJ 2017; 356:j510. 44. Voukalis C, Lip GYH. Oral anticoagulation for elderly patients with non-valvular atrial fibrillation: Recent insights from randomised trials and the “real world”. Heart 2017;103:977e978.