- Email: [email protected]
Bleeding Risk During Oral Anticoagulation in Atrial Fibrillation Patients Older Than 80 Years
Journal of the American College of Cardiology © 2009 by the American College of Cardiology Foundation Published by Elsevier Inc.
Vol. 54, No. 11, 2009 ISSN 0735-1097/09/$36.00 doi:10.1016/j.jacc.2009.05.046
Heart Rhythm Disorders
Bleeding Risk During Oral Anticoagulation in Atrial Fibrillation Patients Older Than 80 Years Daniela Poli, MD,*† Emilia Antonucci, MS,*† Elisa Grifoni, MD,*† Rosanna Abbate, MD,*† Gian Franco Gensini, MD,*†‡ Domenico Prisco, MD*† Florence, Italy Objectives
We sought to evaluate the rate of bleeding in relation to age (⬍80 and ⱖ80 years), the quality of anticoagulation (expressed as time spent in international normalized ratio therapeutic range), and factors associated with bleeding events.
Background
Stroke prevention in patients with atrial fibrillation (AF) is an increasingly crucial public health target, particularly in patients ages ⱖ80 years.
Methods
We conducted a prospective observational study on 783 patients with AF on oral anticoagulant treatment (OAT).
Results
Patients spent a median 14%, 71%, and 15% of time below, within, and above the intended therapeutic range, respectively. No difference in OAT quality was found between patients age ⬍80 and ⱖ80 years. During followup, 94 patients experienced bleeding complications (rate 3.7 ⫻ 100 patient/years), 37 major (rate 1.4 ⫻ 100 patient/years), and 57 minor (rate 2.2 ⫻ 100 patient/years). Different rates of major hemorrhage were observed between patients age ⬍80 and ⱖ80 years (0.9 vs. 1.9 ⫻ 100 patient/years; p ⫽ 0.004). Bleeding risk also was greater in patients with a history of previous cerebral ischemic event (odds ratio [OR]: 2.5; 95% confidence interval: 1.3 to 4.8; p ⫽ 0.007). A Cox regression analysis confirmed age ⱖ80 years associated with bleeding risk (OR: 2.0).
Conclusions
These results indicate that the rate of major bleeding complications may be kept acceptably low also in very elderly AF patients on OAT, provided a careful management of anticoagulation is obtained. (J Am Coll Cardiol 2009;54:999–1002) © 2009 by the American College of Cardiology Foundation
The prevalence of atrial fibrillation (AF) increases with age, and increasing numbers of elderly patients are candidates for and could benefit from the use of anticoagulants. This trend is expected to increase in the future (1); therefore, there is a need for reliable information on the bleeding risk of elderly patients. The quantification of bleeding risk and the identification of modifiable factors are crucial for management decisions. The bleeding risk of elderly patients administered oral anticoagulant treatment (OAT) was studied by several authors (1– 4). However, data remain scarce on very old patients (i.e., those age ⱖ80 years). We conducted a prospective observational study on patients with AF receiving OAT that were followed by our anticoagulation clinic to evaluate the following: 1) the rate of bleeding in relation to age ⬍80 and ⱖ80 years; 2) the quality of anticoagulation; and 3) factors associated with bleeding events.
From the *Department of Medical and Surgical Critical Care, University of Florence, Florence, Italy; †Department of Heart and Vessels, Thrombosis Centre, Azienda Ospedaliera Universitaria Careggi, Florence, Italy; and the ‡Centro S. Maria agli Ulivi, Fondazione Don Carlo Gnocchi Onlus IRCCS, Impruneta, Florence, Italy. Manuscript received March 13, 2009; revised manuscript received May 6, 2009, accepted May 25, 2009.
Methods Patients. From June 1998 to December 2007, we prospectively investigated 783 consecutive nonvalvular AF patients referred for OAT management to our anticoagulation clinic. All were outpatients and were enrolled at the beginning of OAT, mainly after hospital discharge and after giving informed consent. All patients were treated with warfarin, and the international normalized ratio (INR) was maintained at the intended therapeutic range of 2 to 3. At each follow-up visit, OAT was monitored by pro-thrombin time expressed as INR, which was determined by the use of a capillary blood test (Thrombotest, Nycomed Pharma AS, Oslo, Norway). Dose adjustment was performed by a physician by the use of a computer-assisted prescription system (PARMA Instrumentation Laboratory SPA, Milan, Italy) (5). Date of starting OAT and dates of all visits were recorded. We calculated patient/years by our patient database. Time of observation was calculated on consecutive periods of observation. When a patient left the center for ⬎2 months, this interval was excluded from the analysis, and adverse events that eventually occurred during this period were not included in the analysis.
1000
Poli et al. Bleeding Risk in Elderly AF Patients on Warfarin
For the calculation of time in range, the Rosendaal method was used (6). Patients’ demographic and clinAF ⴝ atrial fibrillation ical data were collected. The presCHADS2 ⴝ congestive ence of traditional cardiovascular heart failure, age, diabetes, risk factors and other characterisprevious stroke tics associated with ischemic comCI ⴝ confidence interval plications in AF were assessed on INR ⴝ international the basis of patients’ interview, normalized ratio echocardiography results, and hosOAT ⴝ oral anticoagulant pital records. Patients were classitreatment fied as hypertensive if they had a OR ⴝ odds ratio history of hypertension and were TIA ⴝ transient ischemic taking antihypertensive treatment. attack Diabetes was defined according to American Diabetes Association criteria (7). Coronary artery disease was defined on the basis of a history of myocardial infarction or stable and unstable angina. Left ventricular dysfunction was defined as a recent diagnosis of congestive heart failure or a fractional shortening ⬍25% by transthoracic echocardiography (8). History of previous gastrointestinal or cerebral bleeding events, thrombocytopenia, and malignancy was recorded. Follow-up visits were scheduled every 2 to 4 weeks for INR monitoring. Hospital admissions, concomitant therapies, intercurrent illnesses, bleeding, and thrombotic events during follow-up were recorded. Patients who missed check-ups for ⬎2 months were contacted (personally or through their family or general practitioner) and the reason for interrupting treatment monitoring was recorded. In the case of death, further information about its cause was requested. When this information was lacking, national register of causes of death and autopsy results (if available) were consulted. The use of aspirin in association to OAT was limited to patients with a recent episode of acute coronary syndrome or when they had recurrent episodes. Patients were instructed to avoid the use of aspirin and were routinely asked for occasional or chronic use of aspirin and recorded on electronic files. Outcomes. Data were censored after the occurrence of transient ischemic attack (TIA) or stroke, major bleeding, after the cessation of OAT or when the patient stopped being monitored by our center. Stroke, TIA, and bleeding events were defined as previously described (9). When an ischemic or bleeding event occurred, the INR related to the event was recorded. The INR was defined as temporally related to the adverse event when it was obtained at the time of the event or during the preceding 8 days. No patient was receiving hormone-replacement therapy. We classified our patients for stroke risk by CHADS2 (congestive heart failure, age, diabetes, previous stroke score) (10). Statistical analysis. Incidence rates for ischemic and bleeding events were calculated as the number of events per 100 patient/years of observation. The SPSS statistical software package (software for Windows, version 11.5, Statistical Package for Social Sciences, Chicago, Illinois) was used for Abbreviations and Acronyms
JACC Vol. 54, No. 11, 2009 September 8, 2009:999–1002
data processing. Data are expressed as median and range as the result of to their skewed distribution. Preliminary statistics was performed by the use of the Wilcoxon signed rank test. Statistical analysis was performed by use of the Fisher exact test (categorical data), and a p value ⬍0.05 was chosen for statistical significance. A univariate analysis and a Cox regression analysis were used to ascertain which factors were significantly associated with the risk of bleeding during follow-up. All odds ratios (ORs) are given with their 95% confidence intervals (CIs), and a 2-sided value of p ⬍ 0.05 was chosen for statistical significance. Results We performed a prospective study on 783 AF patients (507 men) for a total follow-up period of 2,567 patient/years. The characteristics of patients are listed in Table 1. At enrollment the patients’ median age was 75 years (range 37 to 94 years) and 180 patients were age ⱖ80 years. During follow-up, 147 patients became age ⬎80 years. The overall exposure to warfarin for each patient was calculated in relation to aging, before and after his/her 80th birthday. During follow-up, 83 patients died (3.2 ⫻ 100 patient/ years), 9 of them for hemorrhagic complications (0.35 ⫻ 100 patient/years). Among these, 45 were age ⱖ80 years (13.8%), and 38 were age ⬍80 years (8.3%). Ninety-four patients had bleeding complications (3.7 ⫻ 100 patient/ years): 37 were major (1.4 ⫻ 100 patient/years) and 57 were minor (2.2 ⫻ 100 patient/years). The INR value related to major bleeding events was ⬍3 in 25 (67.5%) of 37 patients, between 3 and 4 in 7 (19%) of 37 patients, and ⬎4 in 5 (13.5%) of 37 patients. Major bleeding occurred at a median time of 36 months (range 3 to 124 months) after the beginning of OAT; in particular, 9 (24%) of 37 events occurred in the first year. The rates of hemorrhage in patients ⬍80 and ⱖ80 years are reported in Table 2. Overall, patients spent a median of 14%, 71%, and 15% of time below, within, and above the intended therapeutic range, respectively. No difference was found between patients age ⬍80 and ⱖ80 years (data not shown). Mean dosage of warfarin was 27.2 mg/week for patients age ⬍80 years and 21.1 mg/week for patients age ⱖ80 years (p ⬍ 0.001). No difference was found in relation to time spent below, within, and above the intended therapeutic range between patients with bleeding complications and those without, both in patients age ⬍80 and ⱖ80 years. Twentyseven patients (3.5%) also were treated with aspirin; no difference in the rate of bleeding events was found between patients treated or not treated with aspirin. Bleeding risk was similar between sexes. Univariate analysis showed that age ⱖ80 years, a history of previous ischemic event, and CHADS2 score were associated with bleeding risk. Instead, bleeding was not associated with hypertension and history of previous hemorrhage (Table 3). However, among the 8 patients who had
Poli et al. Bleeding Risk in Elderly AF Patients on Warfarin
JACC Vol. 54, No. 11, 2009 September 8, 2009:999–1002
1001
Clinical of Patients of Patients Table 1Characteristics Clinical Characteristics Age All n
<80 Yrs
783
Patient/yrs Follow-up (yrs) Sex (M/F)
456
>80 Yrs
2,567
1,738
829
2.7 (0.1⫺13.0)
2.2 (0.1⫺13.0)
3.6 (0.1⫺12.9)
507/276
Age (yrs)
p Value
327
313/143
194/133
73 (34⫺79)
83 (80⫺99)
Chronic AF
633 (81)
362 (80)
271 (83)
Left ventricular dysfunction
166 (21)
97 (21)
69 (21)
NS NS
Hypertension
495 (64)
292 (64)
203 (62.2)
NS
Diabetes
157 (20)
98 (21)
59 (18.)
Previous stroke/TIA
229 (29)
112 (25)
117 (36)
Peripheral artery disease
111 (14)
60 (13)
51 (16)
NS
Dyslipidemia
187 (26)
124 (27)
63 (21)
0.01
NS 0.001
Smoking habits
128 (17)
90 (21)
38 (12)
0.003
Coronary artery disease
166 (21)
99 (22)
67 (21)
NS
Previous bleeding
23 (2.9)
11 (2.4)
12 (3.7)
NS
Active cancer
35 (4.5)
16 (3.5)
19 (5.8)
NS
CHADS2 score 0
54 (7)
54 (12)
—
NS
1
170 (21)
132 (29)
38 (12)
NS
2
229 (30)
124 (27)
105 (32)
NS
3
174 (22)
81 (18)
93 (28)
NS
4
110 (14)
40 (9)
70 (21)
NS
5
34 (4.5)
16 (3.5)
18 (6)
NS
6
9 (1.5)
6 (1.5)
3 (1)
NS
Values are n, median (range), or n (%). AF ⫽ atrial fibrillation; CHADS2 ⫽ congestive heart failure, age, diabetes, previous stroke; TIA ⫽ transient ischemic attack.
a history of cerebral hemorrhage before beginning warfarin, one had a recurrent cerebral hemorrhage. These patients showed a greater risk for developing a recurrent cerebral hemorrhage than those without a previous cerebral bleeding (4.2 ⫻ 100 patient/years vs. 0.75 ⫻ 100 patient/years), even if not statistically significant (relative risk: 5.4; 95% CI: 0.1 to 34.4; p ⫽ 0.2). Cox regression analysis confirmed that age ⱖ80 years is associated with bleeding (OR: 2.0; 95% CI: 1.1 to 4.0; p ⫽ 0.05). Discussion Rate of bleeding events by age. In this large prospective observational study on patients with AF administered warfarin we found that, even if greater than in younger patients,
the absolute rate of major bleeding of patients ages ⱖ80 years is acceptably low (rate 2.5 ⫻ 100 patient/years). Conversely, Hylek et al. (11), when studying elderly AF patients, found a rate of major bleeding more than 5 times greater than that observed in our study. In the Hylek et al. study (11), 35% of patients experienced coronary artery disease instead of 20% of our cohort, and the concomitant use of aspirin was 40% in those patients and limited to 3.5% in ours. This point could help to explain the lower rate of bleeding recorded in our experience. In our cohort, all patients were followed from the beginning of OAT; therefore, all events, even the early ones, were recorded. No patient was lost at follow-up, and we actively looked for all bleeding events. We cannot exclude
Bleeding in Relation <80 to or Age >80<80 Yearsor >80 Years Table 2 Events Bleeding Eventsto in Age Relation Age <80 Yrs (1,738 Patient/Yrs)
Age >80 Yrs (829 Patient/Yrs)
n
Rate
n
Rate
RR (95% Confidence Interval)
Major bleeding
16
0.9
21
2.5
1.9 (1.2⫺2.8)
0.004
Cerebral
10
0.5
10
1.2
2.1 (0.8⫺5.5)
0.1 0.03
Gastrointestinal
p Value
6
0.3
9
1.1
3.1 (1.0⫺10.7)
Other
—
—
2
0.2
—
—
Fatal
3
0.2
6
0.7
4.2 (0.9⫺26.0)
0.04
31
1.8
26
3.1
1.7 (1.0⫺3.0)
0.03
Minor bleeding RR ⫽ relative risk.
1002
Poli et al. Bleeding Risk in Elderly AF Patients on Warfarin
JACC Vol. 54, No. 11, 2009 September 8, 2009:999–1002
Odds Ratios Associated Bleeding Risk Table 3 Odds Ratios With Associated With Bleeding Risk Odds Ratio
95% Confidence Interval
Age ⱖ80 yrs
3.1
1.5–6.2
0.002
Stroke/TIA
2.5
1.3–4.8
0.007
Hypertension
1.3
0.6–2.7
0.5
Previous bleeding
1.0
0.1–6.9
0.9
CHADS2 score
1.3
1.0–1.7
0.03
p Value
Abbreviations as in Table 1.
underreporting for minor bleeding because we recorded clinically relevant events. However, we enrolled consecutive outpatients with a good life expectancy to justify warfarin prophylaxis. More frail patients are unlikely to be referred to an outpatient service, which could in part explain the low bleeding rate recorded. Quality of anticoagulation control. Our study was conducted on patients managed by a devoted agency, providing the best quality of OAT (12). In our patients, the time spent in range was 71%, similar to that recorded in patients analyzed by the European ISCOAT (Italian Study on Complications of Oral Anticoagulant Therapy) and EAA (European Action on Anticoagulation) studies (5,13) but greater than that reported in the cohort followed up by Hylek et al. (11). The use of computer-assisted systems for dose adjustment has been demonstrated to be useful in OAT management (5), which could explain the low bleeding rate in our study. Factors associated with bleeding. It is increasingly recognized that in patients with AF many risk factors for warfarin related bleeding are also indications for warfarin use (14). In particular a history of TIA/stroke is strongly associated with bleeding risk (15–17). Accordingly, in our study a history of previous ischemic event was associated with a high OR for bleeding during OAT. Instead, hypertension seemed not to be associated with bleeding, probably because all hypertensive patients were receiving therapy with a good blood pressure control. In our series history of previous bleeding event was not correlated with an increased bleeding risk. However, considering only patients who had had a cerebral bleeding before starting OAT, the rate of recurrent cerebral bleeding was particularly high. The small number of these patients limits the statistical significance of these data; nevertheless, it suggests the need for a careful evaluation of the risk/benefit ratio of OAT in patients with a previous cerebral bleeding. Conclusions Our study shows that in elderly patients with AF administered warfarin, the risk of bleeding is acceptably low. This finding suggests that patients ⱖ80 years of age could receive benefit from warfarin prophylaxis when a good quality of anticoagulation is obtained.
Reprint requests and correspondence: Dr. Daniela Poli, Centro di Riferimento Regionale per la Trombosi, Azienda Ospedaliera Universitaria Careggi, V.le Morgagni 85-50134 Firenze, Italy. E-mail: [email protected].
REFERENCES
1. Mant JW. Pro: “warfarin should be the drug of choice for thromboprophylaxis in elderly patients with atrial fibrillation.” Why warfarin should really be the drug of choice for stroke prevention in elderly patients with atrial fibrillation. Thromb Haemost 2008;100:14 –5. 2. Beyth RJ, Quinn L, Landefeld CS. A multicomponent intervention to prevent major bleeding complications in older patients receiving warfarin. A randomized, controlled trial. Ann Intern Med 2000;133: 687–95. 3. Palareti G, Hirsh J, Legnani C, et al. Oral anticoagulation treatment in the elderly: a nested, prospective, case-control study. Arch Intern Med 2000;160:470 – 8. 4. Torn M, Bollen WL, van der Meer FJ, van der Wall EE, Rosendaal FR. Risks of oral anticoagulant therapy with increasing age. Arch Intern Med 2005;165:1527–32. 5. Poller L, Keown M, Ibrahim S, et al. An international multicenter randomized study of computer-assisted oral anticoagulant dosage vs. medical staff dosage. J Thromb Haemost 2008;6:935– 43. 6. Rosendaal FR, Cannegieter SC, van der Meer FJM, Briet E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993;69:236 –9. 7. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 2002;26:S5–20. 8. Stroke Prevention in Atrial Fibrillation Investigators. Risk factors for thromboembolism during aspirin therapy in patients with atrial fibrillation: the Stroke Prevention in Atrial Fibrillation study. J Stroke Cerebrovasc Dis 1995;5:147–57. 9. Poli D, Antonucci E, Grifoni E, Abbate R, Gensini GF, Prisco D. Stroke risk in atrial fibrillation patients on warfarin. Predictive ability of risk stratification schemes for primary and secondary prevention. Thromb Haemost 2009;101:367–72. 10. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA 2001;285:28647–70. 11. Hylek EM, Evans-Molina C, Shea C, Henault LE, Regan S. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation 2007;115: 2689 –96. 12. Nichol MB, Knight TK, Dow T, et al. Quality of anticoagulation monitoring in nonvalvular atrial fibrillation patients: comparison of anticoagulation clinic versus usual care. Ann Pharmacother 2008;4: 62–70. 13. Palareti G, Leali N, Coccheri S, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet 1996;348:423– 8. 14. Lip GY. The balance between stroke prevention and bleeding risk in atrial fibrillation: a delicate balance revisited. Stroke 2008;39:1406 – 8. 15. Saloheimo P, Juvela S, Hillbom M. Use of aspirin, epistaxis, and untreated hypertension as risk factors for primary intracerebral hemorrhage in middle-aged and elderly people. Stroke 2001;32:399 – 4. 16. Hart RG, Tonarelli SB, Pearce LA. Avoiding central nervous system bleeding during antithrombotic therapy: recent data and ideas. Stroke 2005;36:1588 –93. 17. Poli D, Antonucci E, Marcucci R, et al. Risk of bleeding in very old atrial fibrillation patients on warfarin: relationship with ageing and CHADS2 score. Thromb Res 2007;121:347–52 Key Words: atrial fibrillation y bleeding y elderly y quality of anticoagulation y warfarin.
Vol. 54, No. 11, 2009 ISSN 0735-1097/09/$36.00 doi:10.1016/j.jacc.2009.05.046
Heart Rhythm Disorders
Bleeding Risk During Oral Anticoagulation in Atrial Fibrillation Patients Older Than 80 Years Daniela Poli, MD,*† Emilia Antonucci, MS,*† Elisa Grifoni, MD,*† Rosanna Abbate, MD,*† Gian Franco Gensini, MD,*†‡ Domenico Prisco, MD*† Florence, Italy Objectives
We sought to evaluate the rate of bleeding in relation to age (⬍80 and ⱖ80 years), the quality of anticoagulation (expressed as time spent in international normalized ratio therapeutic range), and factors associated with bleeding events.
Background
Stroke prevention in patients with atrial fibrillation (AF) is an increasingly crucial public health target, particularly in patients ages ⱖ80 years.
Methods
We conducted a prospective observational study on 783 patients with AF on oral anticoagulant treatment (OAT).
Results
Patients spent a median 14%, 71%, and 15% of time below, within, and above the intended therapeutic range, respectively. No difference in OAT quality was found between patients age ⬍80 and ⱖ80 years. During followup, 94 patients experienced bleeding complications (rate 3.7 ⫻ 100 patient/years), 37 major (rate 1.4 ⫻ 100 patient/years), and 57 minor (rate 2.2 ⫻ 100 patient/years). Different rates of major hemorrhage were observed between patients age ⬍80 and ⱖ80 years (0.9 vs. 1.9 ⫻ 100 patient/years; p ⫽ 0.004). Bleeding risk also was greater in patients with a history of previous cerebral ischemic event (odds ratio [OR]: 2.5; 95% confidence interval: 1.3 to 4.8; p ⫽ 0.007). A Cox regression analysis confirmed age ⱖ80 years associated with bleeding risk (OR: 2.0).
Conclusions
These results indicate that the rate of major bleeding complications may be kept acceptably low also in very elderly AF patients on OAT, provided a careful management of anticoagulation is obtained. (J Am Coll Cardiol 2009;54:999–1002) © 2009 by the American College of Cardiology Foundation
The prevalence of atrial fibrillation (AF) increases with age, and increasing numbers of elderly patients are candidates for and could benefit from the use of anticoagulants. This trend is expected to increase in the future (1); therefore, there is a need for reliable information on the bleeding risk of elderly patients. The quantification of bleeding risk and the identification of modifiable factors are crucial for management decisions. The bleeding risk of elderly patients administered oral anticoagulant treatment (OAT) was studied by several authors (1– 4). However, data remain scarce on very old patients (i.e., those age ⱖ80 years). We conducted a prospective observational study on patients with AF receiving OAT that were followed by our anticoagulation clinic to evaluate the following: 1) the rate of bleeding in relation to age ⬍80 and ⱖ80 years; 2) the quality of anticoagulation; and 3) factors associated with bleeding events.
From the *Department of Medical and Surgical Critical Care, University of Florence, Florence, Italy; †Department of Heart and Vessels, Thrombosis Centre, Azienda Ospedaliera Universitaria Careggi, Florence, Italy; and the ‡Centro S. Maria agli Ulivi, Fondazione Don Carlo Gnocchi Onlus IRCCS, Impruneta, Florence, Italy. Manuscript received March 13, 2009; revised manuscript received May 6, 2009, accepted May 25, 2009.
Methods Patients. From June 1998 to December 2007, we prospectively investigated 783 consecutive nonvalvular AF patients referred for OAT management to our anticoagulation clinic. All were outpatients and were enrolled at the beginning of OAT, mainly after hospital discharge and after giving informed consent. All patients were treated with warfarin, and the international normalized ratio (INR) was maintained at the intended therapeutic range of 2 to 3. At each follow-up visit, OAT was monitored by pro-thrombin time expressed as INR, which was determined by the use of a capillary blood test (Thrombotest, Nycomed Pharma AS, Oslo, Norway). Dose adjustment was performed by a physician by the use of a computer-assisted prescription system (PARMA Instrumentation Laboratory SPA, Milan, Italy) (5). Date of starting OAT and dates of all visits were recorded. We calculated patient/years by our patient database. Time of observation was calculated on consecutive periods of observation. When a patient left the center for ⬎2 months, this interval was excluded from the analysis, and adverse events that eventually occurred during this period were not included in the analysis.
1000
Poli et al. Bleeding Risk in Elderly AF Patients on Warfarin
For the calculation of time in range, the Rosendaal method was used (6). Patients’ demographic and clinAF ⴝ atrial fibrillation ical data were collected. The presCHADS2 ⴝ congestive ence of traditional cardiovascular heart failure, age, diabetes, risk factors and other characterisprevious stroke tics associated with ischemic comCI ⴝ confidence interval plications in AF were assessed on INR ⴝ international the basis of patients’ interview, normalized ratio echocardiography results, and hosOAT ⴝ oral anticoagulant pital records. Patients were classitreatment fied as hypertensive if they had a OR ⴝ odds ratio history of hypertension and were TIA ⴝ transient ischemic taking antihypertensive treatment. attack Diabetes was defined according to American Diabetes Association criteria (7). Coronary artery disease was defined on the basis of a history of myocardial infarction or stable and unstable angina. Left ventricular dysfunction was defined as a recent diagnosis of congestive heart failure or a fractional shortening ⬍25% by transthoracic echocardiography (8). History of previous gastrointestinal or cerebral bleeding events, thrombocytopenia, and malignancy was recorded. Follow-up visits were scheduled every 2 to 4 weeks for INR monitoring. Hospital admissions, concomitant therapies, intercurrent illnesses, bleeding, and thrombotic events during follow-up were recorded. Patients who missed check-ups for ⬎2 months were contacted (personally or through their family or general practitioner) and the reason for interrupting treatment monitoring was recorded. In the case of death, further information about its cause was requested. When this information was lacking, national register of causes of death and autopsy results (if available) were consulted. The use of aspirin in association to OAT was limited to patients with a recent episode of acute coronary syndrome or when they had recurrent episodes. Patients were instructed to avoid the use of aspirin and were routinely asked for occasional or chronic use of aspirin and recorded on electronic files. Outcomes. Data were censored after the occurrence of transient ischemic attack (TIA) or stroke, major bleeding, after the cessation of OAT or when the patient stopped being monitored by our center. Stroke, TIA, and bleeding events were defined as previously described (9). When an ischemic or bleeding event occurred, the INR related to the event was recorded. The INR was defined as temporally related to the adverse event when it was obtained at the time of the event or during the preceding 8 days. No patient was receiving hormone-replacement therapy. We classified our patients for stroke risk by CHADS2 (congestive heart failure, age, diabetes, previous stroke score) (10). Statistical analysis. Incidence rates for ischemic and bleeding events were calculated as the number of events per 100 patient/years of observation. The SPSS statistical software package (software for Windows, version 11.5, Statistical Package for Social Sciences, Chicago, Illinois) was used for Abbreviations and Acronyms
JACC Vol. 54, No. 11, 2009 September 8, 2009:999–1002
data processing. Data are expressed as median and range as the result of to their skewed distribution. Preliminary statistics was performed by the use of the Wilcoxon signed rank test. Statistical analysis was performed by use of the Fisher exact test (categorical data), and a p value ⬍0.05 was chosen for statistical significance. A univariate analysis and a Cox regression analysis were used to ascertain which factors were significantly associated with the risk of bleeding during follow-up. All odds ratios (ORs) are given with their 95% confidence intervals (CIs), and a 2-sided value of p ⬍ 0.05 was chosen for statistical significance. Results We performed a prospective study on 783 AF patients (507 men) for a total follow-up period of 2,567 patient/years. The characteristics of patients are listed in Table 1. At enrollment the patients’ median age was 75 years (range 37 to 94 years) and 180 patients were age ⱖ80 years. During follow-up, 147 patients became age ⬎80 years. The overall exposure to warfarin for each patient was calculated in relation to aging, before and after his/her 80th birthday. During follow-up, 83 patients died (3.2 ⫻ 100 patient/ years), 9 of them for hemorrhagic complications (0.35 ⫻ 100 patient/years). Among these, 45 were age ⱖ80 years (13.8%), and 38 were age ⬍80 years (8.3%). Ninety-four patients had bleeding complications (3.7 ⫻ 100 patient/ years): 37 were major (1.4 ⫻ 100 patient/years) and 57 were minor (2.2 ⫻ 100 patient/years). The INR value related to major bleeding events was ⬍3 in 25 (67.5%) of 37 patients, between 3 and 4 in 7 (19%) of 37 patients, and ⬎4 in 5 (13.5%) of 37 patients. Major bleeding occurred at a median time of 36 months (range 3 to 124 months) after the beginning of OAT; in particular, 9 (24%) of 37 events occurred in the first year. The rates of hemorrhage in patients ⬍80 and ⱖ80 years are reported in Table 2. Overall, patients spent a median of 14%, 71%, and 15% of time below, within, and above the intended therapeutic range, respectively. No difference was found between patients age ⬍80 and ⱖ80 years (data not shown). Mean dosage of warfarin was 27.2 mg/week for patients age ⬍80 years and 21.1 mg/week for patients age ⱖ80 years (p ⬍ 0.001). No difference was found in relation to time spent below, within, and above the intended therapeutic range between patients with bleeding complications and those without, both in patients age ⬍80 and ⱖ80 years. Twentyseven patients (3.5%) also were treated with aspirin; no difference in the rate of bleeding events was found between patients treated or not treated with aspirin. Bleeding risk was similar between sexes. Univariate analysis showed that age ⱖ80 years, a history of previous ischemic event, and CHADS2 score were associated with bleeding risk. Instead, bleeding was not associated with hypertension and history of previous hemorrhage (Table 3). However, among the 8 patients who had
Poli et al. Bleeding Risk in Elderly AF Patients on Warfarin
JACC Vol. 54, No. 11, 2009 September 8, 2009:999–1002
1001
Clinical of Patients of Patients Table 1Characteristics Clinical Characteristics Age All n
<80 Yrs
783
Patient/yrs Follow-up (yrs) Sex (M/F)
456
>80 Yrs
2,567
1,738
829
2.7 (0.1⫺13.0)
2.2 (0.1⫺13.0)
3.6 (0.1⫺12.9)
507/276
Age (yrs)
p Value
327
313/143
194/133
73 (34⫺79)
83 (80⫺99)
Chronic AF
633 (81)
362 (80)
271 (83)
Left ventricular dysfunction
166 (21)
97 (21)
69 (21)
NS NS
Hypertension
495 (64)
292 (64)
203 (62.2)
NS
Diabetes
157 (20)
98 (21)
59 (18.)
Previous stroke/TIA
229 (29)
112 (25)
117 (36)
Peripheral artery disease
111 (14)
60 (13)
51 (16)
NS
Dyslipidemia
187 (26)
124 (27)
63 (21)
0.01
NS 0.001
Smoking habits
128 (17)
90 (21)
38 (12)
0.003
Coronary artery disease
166 (21)
99 (22)
67 (21)
NS
Previous bleeding
23 (2.9)
11 (2.4)
12 (3.7)
NS
Active cancer
35 (4.5)
16 (3.5)
19 (5.8)
NS
CHADS2 score 0
54 (7)
54 (12)
—
NS
1
170 (21)
132 (29)
38 (12)
NS
2
229 (30)
124 (27)
105 (32)
NS
3
174 (22)
81 (18)
93 (28)
NS
4
110 (14)
40 (9)
70 (21)
NS
5
34 (4.5)
16 (3.5)
18 (6)
NS
6
9 (1.5)
6 (1.5)
3 (1)
NS
Values are n, median (range), or n (%). AF ⫽ atrial fibrillation; CHADS2 ⫽ congestive heart failure, age, diabetes, previous stroke; TIA ⫽ transient ischemic attack.
a history of cerebral hemorrhage before beginning warfarin, one had a recurrent cerebral hemorrhage. These patients showed a greater risk for developing a recurrent cerebral hemorrhage than those without a previous cerebral bleeding (4.2 ⫻ 100 patient/years vs. 0.75 ⫻ 100 patient/years), even if not statistically significant (relative risk: 5.4; 95% CI: 0.1 to 34.4; p ⫽ 0.2). Cox regression analysis confirmed that age ⱖ80 years is associated with bleeding (OR: 2.0; 95% CI: 1.1 to 4.0; p ⫽ 0.05). Discussion Rate of bleeding events by age. In this large prospective observational study on patients with AF administered warfarin we found that, even if greater than in younger patients,
the absolute rate of major bleeding of patients ages ⱖ80 years is acceptably low (rate 2.5 ⫻ 100 patient/years). Conversely, Hylek et al. (11), when studying elderly AF patients, found a rate of major bleeding more than 5 times greater than that observed in our study. In the Hylek et al. study (11), 35% of patients experienced coronary artery disease instead of 20% of our cohort, and the concomitant use of aspirin was 40% in those patients and limited to 3.5% in ours. This point could help to explain the lower rate of bleeding recorded in our experience. In our cohort, all patients were followed from the beginning of OAT; therefore, all events, even the early ones, were recorded. No patient was lost at follow-up, and we actively looked for all bleeding events. We cannot exclude
Bleeding in Relation <80 to or Age >80<80 Yearsor >80 Years Table 2 Events Bleeding Eventsto in Age Relation Age <80 Yrs (1,738 Patient/Yrs)
Age >80 Yrs (829 Patient/Yrs)
n
Rate
n
Rate
RR (95% Confidence Interval)
Major bleeding
16
0.9
21
2.5
1.9 (1.2⫺2.8)
0.004
Cerebral
10
0.5
10
1.2
2.1 (0.8⫺5.5)
0.1 0.03
Gastrointestinal
p Value
6
0.3
9
1.1
3.1 (1.0⫺10.7)
Other
—
—
2
0.2
—
—
Fatal
3
0.2
6
0.7
4.2 (0.9⫺26.0)
0.04
31
1.8
26
3.1
1.7 (1.0⫺3.0)
0.03
Minor bleeding RR ⫽ relative risk.
1002
Poli et al. Bleeding Risk in Elderly AF Patients on Warfarin
JACC Vol. 54, No. 11, 2009 September 8, 2009:999–1002
Odds Ratios Associated Bleeding Risk Table 3 Odds Ratios With Associated With Bleeding Risk Odds Ratio
95% Confidence Interval
Age ⱖ80 yrs
3.1
1.5–6.2
0.002
Stroke/TIA
2.5
1.3–4.8
0.007
Hypertension
1.3
0.6–2.7
0.5
Previous bleeding
1.0
0.1–6.9
0.9
CHADS2 score
1.3
1.0–1.7
0.03
p Value
Abbreviations as in Table 1.
underreporting for minor bleeding because we recorded clinically relevant events. However, we enrolled consecutive outpatients with a good life expectancy to justify warfarin prophylaxis. More frail patients are unlikely to be referred to an outpatient service, which could in part explain the low bleeding rate recorded. Quality of anticoagulation control. Our study was conducted on patients managed by a devoted agency, providing the best quality of OAT (12). In our patients, the time spent in range was 71%, similar to that recorded in patients analyzed by the European ISCOAT (Italian Study on Complications of Oral Anticoagulant Therapy) and EAA (European Action on Anticoagulation) studies (5,13) but greater than that reported in the cohort followed up by Hylek et al. (11). The use of computer-assisted systems for dose adjustment has been demonstrated to be useful in OAT management (5), which could explain the low bleeding rate in our study. Factors associated with bleeding. It is increasingly recognized that in patients with AF many risk factors for warfarin related bleeding are also indications for warfarin use (14). In particular a history of TIA/stroke is strongly associated with bleeding risk (15–17). Accordingly, in our study a history of previous ischemic event was associated with a high OR for bleeding during OAT. Instead, hypertension seemed not to be associated with bleeding, probably because all hypertensive patients were receiving therapy with a good blood pressure control. In our series history of previous bleeding event was not correlated with an increased bleeding risk. However, considering only patients who had had a cerebral bleeding before starting OAT, the rate of recurrent cerebral bleeding was particularly high. The small number of these patients limits the statistical significance of these data; nevertheless, it suggests the need for a careful evaluation of the risk/benefit ratio of OAT in patients with a previous cerebral bleeding. Conclusions Our study shows that in elderly patients with AF administered warfarin, the risk of bleeding is acceptably low. This finding suggests that patients ⱖ80 years of age could receive benefit from warfarin prophylaxis when a good quality of anticoagulation is obtained.
Reprint requests and correspondence: Dr. Daniela Poli, Centro di Riferimento Regionale per la Trombosi, Azienda Ospedaliera Universitaria Careggi, V.le Morgagni 85-50134 Firenze, Italy. E-mail: [email protected].
REFERENCES
1. Mant JW. Pro: “warfarin should be the drug of choice for thromboprophylaxis in elderly patients with atrial fibrillation.” Why warfarin should really be the drug of choice for stroke prevention in elderly patients with atrial fibrillation. Thromb Haemost 2008;100:14 –5. 2. Beyth RJ, Quinn L, Landefeld CS. A multicomponent intervention to prevent major bleeding complications in older patients receiving warfarin. A randomized, controlled trial. Ann Intern Med 2000;133: 687–95. 3. Palareti G, Hirsh J, Legnani C, et al. Oral anticoagulation treatment in the elderly: a nested, prospective, case-control study. Arch Intern Med 2000;160:470 – 8. 4. Torn M, Bollen WL, van der Meer FJ, van der Wall EE, Rosendaal FR. Risks of oral anticoagulant therapy with increasing age. Arch Intern Med 2005;165:1527–32. 5. Poller L, Keown M, Ibrahim S, et al. An international multicenter randomized study of computer-assisted oral anticoagulant dosage vs. medical staff dosage. J Thromb Haemost 2008;6:935– 43. 6. Rosendaal FR, Cannegieter SC, van der Meer FJM, Briet E. A method to determine the optimal intensity of oral anticoagulant therapy. Thromb Haemost 1993;69:236 –9. 7. Report of the expert committee on the diagnosis and classification of diabetes mellitus. Diabetes Care 2002;26:S5–20. 8. Stroke Prevention in Atrial Fibrillation Investigators. Risk factors for thromboembolism during aspirin therapy in patients with atrial fibrillation: the Stroke Prevention in Atrial Fibrillation study. J Stroke Cerebrovasc Dis 1995;5:147–57. 9. Poli D, Antonucci E, Grifoni E, Abbate R, Gensini GF, Prisco D. Stroke risk in atrial fibrillation patients on warfarin. Predictive ability of risk stratification schemes for primary and secondary prevention. Thromb Haemost 2009;101:367–72. 10. Gage BF, Waterman AD, Shannon W, Boechler M, Rich MW, Radford MJ. Validation of clinical classification schemes for predicting stroke: results from the National Registry of Atrial Fibrillation. JAMA 2001;285:28647–70. 11. Hylek EM, Evans-Molina C, Shea C, Henault LE, Regan S. Major hemorrhage and tolerability of warfarin in the first year of therapy among elderly patients with atrial fibrillation. Circulation 2007;115: 2689 –96. 12. Nichol MB, Knight TK, Dow T, et al. Quality of anticoagulation monitoring in nonvalvular atrial fibrillation patients: comparison of anticoagulation clinic versus usual care. Ann Pharmacother 2008;4: 62–70. 13. Palareti G, Leali N, Coccheri S, et al. Bleeding complications of oral anticoagulant treatment: an inception-cohort, prospective collaborative study (ISCOAT). Italian Study on Complications of Oral Anticoagulant Therapy. Lancet 1996;348:423– 8. 14. Lip GY. The balance between stroke prevention and bleeding risk in atrial fibrillation: a delicate balance revisited. Stroke 2008;39:1406 – 8. 15. Saloheimo P, Juvela S, Hillbom M. Use of aspirin, epistaxis, and untreated hypertension as risk factors for primary intracerebral hemorrhage in middle-aged and elderly people. Stroke 2001;32:399 – 4. 16. Hart RG, Tonarelli SB, Pearce LA. Avoiding central nervous system bleeding during antithrombotic therapy: recent data and ideas. Stroke 2005;36:1588 –93. 17. Poli D, Antonucci E, Marcucci R, et al. Risk of bleeding in very old atrial fibrillation patients on warfarin: relationship with ageing and CHADS2 score. Thromb Res 2007;121:347–52 Key Words: atrial fibrillation y bleeding y elderly y quality of anticoagulation y warfarin.