Prognostic Implications of Warfarin Cessation After Major Trauma
http://www.100md.com
《循环学杂志》
the Division of Clinical Pharmacology (D.G.H.) and the Institute for Clinical Evaluative Sciences (D.G.H., A.K., D.A.R.)University of Toronto
the Department of Health Policy, Management and Evaluation and Department of Medicine (D.G.H., D.A.R.), Sunnybrook & Women’s College Health Sciences Centre, Toronto, Canada.
Abstract
Background— Warfarin therapy is often withheld from elderly patients who fall or otherwise experience injury because of concerns regarding the long-term risk of hemorrhage in these individuals. We studied whether stopping warfarin after trauma is associated with a higher risk of subsequent adverse cardiovascular events.
Methods and Results— We conducted a retrospective, population-based, cohort study using linked administrative databases in the province of Ontario, Canada for the years 1992 to 2001. A total of 8450 elderly patients (age >65 years) who survived an incident of major trauma and were receiving warfarin before injury were followed up for a mean of 3.3 years. During the 6-month interval after trauma, 1827 (22%) patients discontinued warfarin, whereas 6623 (78%) patients continued warfarin. Warfarin cessation was not associated with an increased risk of subsequent stroke (hazard ratio [HR] 0.99, 95% CI 0.82 to 1.21) or myocardial infarction (HR 0.94, 95% CI 0.74 to 1.20) but was associated with a lower risk of major hemorrhage (HR 0.69, 95% CI 0.54 to 0.88) and a higher risk of venous thromboembolism (HR 1.59, 95% CI 1.07 to 2.36). Adjustment for baseline demographics, stroke risk factors, other comorbidities, and characteristics of the trauma did not materially change these findings. On-treatment analyses yielded similar results.
Conclusions— Cessation of warfarin in elderly patients after major trauma was not associated with an increased risk of arterial thrombotic events but was associated with a significantly increased risk of venous thromboembolism.
Key Words: anticoagulants ; epidemiology ; stroke ; follow-up studies
Introduction
Warfarin is an oral anticoagulant widely used for preventing thromboembolism in patients with atrial fibrillation, prosthetic heart valves, and venous thrombosis.1 Despite its proven efficacy for preventing strokes, warfarin is frequently withheld from elderly patients because of concerns about major hemorrhage.2–9 Factors that increase the risk of bleeding are often prevalent in the elderly, including prior gastrointestinal hemorrhage, erratic compliance, and a tendency to fall.10,11 The tradeoff between benefits and risks of warfarin leads to considerable uncertainty about the decision to prescribe this therapy to older patients in the community.12,13 In one study of physician attitudes regarding warfarin for stroke prevention, for example, the most commonly cited contraindication was an increased risk of falls (reported by 71% of respondents).14
Patients who sustain major trauma while receiving warfarin may bleed more seriously from their injuries, particularly when craniocerebral trauma occurs.15–18 However, the medical literature offers remarkably little evidence to guide physicians about resumption of long-term oral anticoagulation in patients who survive major trauma. The potential protection from stroke and other thrombotic events favors the reinstitution of long-term anticoagulation. Conversely, the occurrence of trauma may be a marker of poor functional status, comorbidity, or unreliable patient behavior that may seriously compromise the benefits of long-term warfarin therapy.19
We hypothesized that elderly patients who discontinue warfarin after major trauma face an increased risk of adverse cardiovascular events. Specifically, our theory was that major trauma does not reverse the rationale for resuming warfarin. By monitoring a population-based cohort of elderly patients who sustained major trauma while coincidentally receiving warfarin, we assessed the long-term effects associated with discontinuation of warfarin after trauma. We did not examine the short-term (<6 months) benefits and risks of oral anticoagulation after trauma for prophylaxis or other indications.
Methods
Setting and Design
We conducted a population-based, retrospective, cohort study by linking multiple administrative healthcare databases over a 9-year period (January 1, 1992, to January 1, 2001) in the province of Ontario. Throughout the study, Ontario was Canada’s most populous province, with 12 million residents, of whom 1.5 million were aged 65 years or older. Elderly patients in Ontario had universal access to hospital care, physicians’ services, and prescription drug coverage. In addition, healthcare records could be analyzed with encrypted health card numbers to track individual patients over time.20 Ethical approval for the study was obtained from the Sunnybrook and Women’s College Health Sciences Centre Research Ethics Board.
Data Sources
We used 4 large databases: the Ontario Drug Benefit database, which records prescription medications dispensed to elderly patients in Ontario; the Canadian Institute for Health Information (CIHI) discharge abstract database, which records hospital admissions, including detailed diagnostic and procedural information; the Ontario Health Insurance Plan database, which provides information on physician claims for inpatient and outpatient services; and the Ontario Registered Persons database, which contains vital statistics on all residents in the province. These 4 databases have been used extensively to study population-based health outcomes.21,22
Patients
Patients were included if they were 66 years of age or older at the time of major trauma, defined as a hospitalization for which trauma was the most responsible diagnosis. We used International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) E codes 800 through 999 to define and categorize all subtypes of trauma. We excluded patients whose admission was primarily related to late effects of trauma (E code 929), complications of medical procedures (E codes 930 to 949), or poisonings (E codes 849 to 879). We focused on patients who had recently received warfarin, defined as at least 1 prescription for warfarin within 6 months before the trauma. We excluded those patients who died in the hospital or who did not survive for at least 6 months after discharge to remove deaths that might be linked to the incident trauma and recovery.
Comorbidities
The frequency of major stroke risk factors and other comorbid diseases was determined by analysis of inpatient and outpatient databases before trauma. In accord with past research, the 7 prespecified stroke risk factors were atrial fibrillation, a prosthetic heart valve, diabetes mellitus, hypertension, heart failure, previous stroke or transient ischemic attack, and coronary artery disease.23 Additional prespecified comorbidities were a history of falls, alcoholism, dementia, previous major hemorrhage, malignancy, and renal insufficiency. The Charlson Comorbidity Index score was used to quantify overall comorbidity.24 Two additional generic indices of comorbidity were provided by the number of hospital admissions in the preceding 3 years and the number of distinct medications prescribed in the past year.25
Outcomes
The primary end point was the first occurrence of ischemic stroke, defined as admission to hospital with ischemic stroke as the most responsible diagnosis (ICD-9-CM codes 433, 434, and 436). The accuracy rate of coding for stroke in the CIHI database exceeds 90%.23,26 The 4 secondary end points examined were major hemorrhage, acute myocardial infarction, venous thromboembolism, and death. Major hemorrhage was defined according to an admission with the most responsible diagnosis of hemorrhage or for a procedure to repair hemorrhage (appendix A for ICD-9-CM diagnosis and procedure codes for hemorrhage). Myocardial infarction was defined as an admission with the most responsible diagnosis of myocardial infarction (ICD-9-CM 410). These diagnosis codes have a positive predictive value of 90% to 96%.27,28 Venous thromboembolism was defined as an admission with a most responsible diagnosis of pulmonary embolism (ICD-9-CM 415.1) or deep vein thrombosis (ICD-9-CM 325, 437.6, 451 to 453). These diagnosis codes have a sensitivity of 97% and a specificity of 90% for true thromboembolic disease.29 All-cause mortality was obtained from the Registered Persons database.
Subgroups
Predefined subgroups were identified to compare event rates in 9 settings: a history of atrial fibrillation; history of stroke or transient ischemic attack; history of diabetes mellitus; history of hypertension; history of congestive heart failure; history of coronary artery disease; history of malignancy; absence of head injury on trauma; and trauma resulting from a fall. The intent of these subgroup analyses was to examine warfarin therapy in distinct patient populations, particularly in those settings established by the literature as representing elevated risk of major vascular events.
Statistical Analysis
We defined the period of observation as beginning 6 months after discharge. Follow-up ended at the time of a primary event, death, or the end of the study if the patient course was uneventful (March 31, 2002). The primary analysis compared patients who discontinued warfarin by 6 months after discharge with patients who continued taking warfarin. Crude differences between the 2 groups in the risk of outcomes were calculated as hazard ratios (HRs) with Cox regression models and subsequently adjusted for demographic factors, stroke risk factors, and other covariates (comorbidities and characteristics of the trauma).
We also conducted an on-treatment analysis by checking for warfarin claims in the prescription database within 6 months immediately before the occurrence of an outcome. In the treated group, we regarded on-treatment outcomes as those for which at least 1 claim for warfarin was made within 6 months before the event. Similarly, for those who discontinued warfarin, we retained only outcomes for which no claim for warfarin was made within the 6-month preevent interval. The intent of the on-treatment analysis was to examine the consistency of findings with a more precise measure of patient compliance.
Results
During the 9-year study period, 11 139 patients receiving warfarin sustained trauma and were treated at 252 hospitals in Ontario. Of these, 1284 (11%) died in the hospital, and an additional 1405 (13%) died within 6 months of discharge; 61 of these deaths were due to stroke. Of the 8450 long-term survivors, 6623 (78%) resumed the use of warfarin during the 6-month period after discharge, whereas 1827 (22%) did not resume warfarin (Table). Collectively, this cohort provided 27 530 person-years of follow-up after major trauma (mean duration of follow-up, 3.3 years). The majority (82%) of trauma episodes were due to falls.
Characteristics of Patients Prescribed Warfarin Before Major Trauma
Patients who discontinued using long-term warfarin after trauma were similar to those who continued its use in basic demographics, stroke risk factors, and other comorbidities. The major exception was that atrial fibrillation was somewhat less common in those who discontinued the use of warfarin. The subtypes of trauma were fairly evenly distributed between the 2 groups. A past history of major hemorrhage was equally common in both groups.
A total of 592 patients sustained an ischemic stroke during follow-up, equivalent to a rate of 2.2% per year. We observed no difference in the risk of stroke in the patients who discontinued warfarin compared with those who continued its use, even after adjustment for demographic factors, stroke risk factors, and other covariates (Figures 1 and 2). Although the rate of stroke varied substantially across the different subgroups, in no subgroup was warfarin cessation associated with an increased risk of stroke (including patients with atrial fibrillation or those with prior cerebrovascular disease). The on-treatment analysis continued to show no evidence of an association between warfarin cessation and an increased risk of ischemic stroke.
A total of 399 patients sustained an acute myocardial infarction during follow-up, equivalent to an annual incidence of 1.5%. As with stroke, we observed no apparent relation between warfarin discontinuation and the subsequent long-term risk of myocardial infarction (Figure 3). This lack of association was apparent in both the crude and adjusted analyses, all subgroups, and the on-treatment analysis. There were 4203 deaths in the trauma cohort, with an overall mortality rate of 16% per year. We found a slight but statistically significant increase in mortality in those who discontinued using warfarin (HR 1.09, 95% CI 1.02 to 1.18), a finding which was no longer significant after adjustment for covariates (HR 1.05, 95% CI 0.98 to 1.13).
The long-term risk of major hemorrhage for trauma patients who discontinued warfarin use was 1.3% annually, which was significantly lower than the rate for trauma patients who resumed using warfarin (1.9%; P=0.0025 for difference). In the unadjusted model, the risk of major hemorrhage was increased by 45% in those who continued using warfarin, an association that persisted after adjustment for age, sex, stroke risk factors, comorbidities, and trauma characteristics (Figure 4). The increased risk of hemorrhage for trauma patients who continued to use warfarin was consistent across most major subgroups, accentuated in patients with a previous history of stroke or transient ischemic attack, and further strengthened in the on-treatment analysis.
There were 115 episodes of venous thromboembolism over the course of follow-up, equivalent to a rate of 0.4% annually. Discontinuation of warfarin use was associated with a significant increase in the risk of venous thromboembolism (Figure 5). This finding was strengthened in the on-treatment analysis and was consistent across most subgroups. Overall, the increase in venous thromboembolism primarily reflected a substantial increase in hospitalizations for deep vein thrombosis (HR 1.94, 95% CI 1.18 to 3.18), with no significant increase in pulmonary embolism (HR 1.01, 95% CI 0.52 to 1.97).
A number of risk factors for the combined outcome of death or stroke were corroborated in the anticipated manner in the full cohort. As expected, age was the strongest predictor and was associated with a 44% increase in risk for each decade increase in age (95% CI 38% to 51%). Similarly, male gender (HR 1.21, 95% CI 1.15 to 1.25), past stroke or transient ischemic attack (HR 1.12, 95% CI 1.06 to 1.19), renal insufficiency (HR 1.11, 95% CI 1.03 to 1.19), and dementia (HR 1.19, 95% CI 1.10 to 1.29) were further independent risk factors for the combined end point. Similar results were noted in the atrial fibrillation subgroup.
Discussion
Multiple studies support the efficacy and safety of warfarin for preventing stroke, myocardial infarction, and venous thromboembolism.30–32 In particular, warfarin reduces the risk of ischemic stroke by approximately two thirds in patients with atrial fibrillation.33 In many settings, the moderately raised incidence of bleeding related to warfarin is greatly outweighed by the benefits in preventing stroke, disability, and death34,35; however, controversy surrounds the risk-benefit ratio of warfarin use after trauma, and we found that 1 in 5 patients discontinued their use of warfarin after hospitalization for major trauma. Moreover, those patients who discontinued long-term warfarin experienced fewer adverse events than were observed in the placebo groups of prior randomized trials.
The largest limitation of the present study is that the lack of random allocation to warfarin therapy raises the possibility of selection bias; furthermore, poor compliance might slant the analysis against warfarin. However, 4 additional findings suggest that hidden confounding is unlikely to explain the results. First, we made adjustments for many factors that are known to be prognostically important. These adjustments did not affect the results of our analysis. Second, analyses of specific high-risk subgroups of patients were consistent with the overall cohort. Third, bias due to poor compliance would not explain why all the main findings were replicated in the on-treatment analyses. Finally, the significantly increased risk of hemorrhage in those who continued the use of warfarin suggests at least some degree of anticoagulation in these patients, thus arguing against major noncompliance.
One possible explanation for our findings is behavioral. Elderly patients who are prone to trauma are more likely to have comorbidities (such as dementia) and neurobehavioral attributes (such as poor attention) that may predispose to reduced efficacy with warfarin.36–39 Such characteristics could result in poor compliance with warfarin therapy and potential lapses in follow-up, monitoring, and dose adjustment.40–42 These patients would thus be predisposed to subtherapeutic (stroke-promoting) and supratherapeutic (hemorrhage-promoting) levels of anticoagulation. Consistent with this explanation, many patients in the present study cohort had a history of dementia, polypharmacy, malignancy, and other comorbid conditions that may compromise the efficacy and safety of warfarin.
A second possible explanation is that patients with severe cerebrovascular disease might not survive major trauma or its sequelae.43–45 This might also explain why the annual incidence of ischemic stroke observed in the present study cohort was lower than expected from population studies. Hence, major trauma may have depleted the patients who were most prone to strokes (the same subset of patients most likely to benefit from warfarin).46,47 Because the observation period of the present study began at 6 months after trauma, our results do not address whether warfarin is beneficial in the early posttrauma period. Intriguingly, however, we found a significant increase in the long-term risk of venous thromboembolism in trauma patients who discontinued warfarin. Although randomized controlled trials have shown significant reductions in the risk of venous thrombosis with early anticoagulation after trauma, this is the first study, to the best of our knowledge, that suggests long-term warfarin use may also be protective.
The literature on the safety of resuming warfarin after trauma is remarkably limited. A recent decision analysis concluded that a patient would need to fall 300 times per year to outweigh the benefits of long-term warfarin use for stroke prevention35; however, that decision analysis was based on randomized controlled trials that generally excluded trauma patients. Moreover, the presence of comorbidities that increase the risk of hemorrhage and cluster with the risk of trauma were not considered in the analysis. A recent cohort study of unselected trauma patients found substantially lower rates of discharge to the community in patients using warfarin at the time of trauma.48 Moreover, warfarin users had significantly worse functional outcomes than nonusers of warfarin, particularly those who sustained head injury at the time of trauma.
We found that discontinuation of long-term warfarin use after major trauma was not associated with an increased risk of stroke or myocardial infarction but was associated with a significant increase in the risk of venous thromboembolism. Further research is necessary to replicate our work. In the meantime, physicians who prescribe long-term warfarin to patients with a history of trauma should carefully weigh the competing risks of hemorrhage and venous thromboembolism.
Acknowledgments
Dr Hackam was supported by a Canadian Institutes for Health Research Fellowship Award. Dr Redelmeier was supported by a Canada Research Chair in Medical Decision Sciences and the Error Management Unit of Sunnybrook and Women’s College. We thank Mark Crowther, John Blakely, Andreas Laupacis, Bill Geerts, Steven Shumak, David Juurlink, Chaim Bell, and Damon Scales for their helpful comments on previous drafts of the manuscript and Tracy Willson for secretarial support.
References
Hirsh J, Fuster V, Ansell J, Halperin JL. American Heart Association/American College of Cardiology Foundation guide to warfarin therapy. Circulation. 2003; 107: 1692–1711.
McCormick D, Gurwitz JH, Goldberg RJ, Becker R, Tate JP, Elwell A, Radford MJ. Prevalence and quality of warfarin use for patients with atrial fibrillation in the long-term care setting. Arch Intern Med. 2001; 161: 2458–2463.
Cohen N, Almoznino-Sarafian D, Alon I, Gorelik O, Koopfer M, Chachashvily S, Shteinshnaider M, Litvinjuk V, Modai D. Warfarin for stroke prevention still underused in atrial fibrillation: patterns of omission. Stroke. 2000; 31: 1217–1222.
Mendelson G, Aronow WS. Underutilization of warfarin in older persons with chronic nonvalvular atrial fibrillation at high risk for developing stroke. J Am Geriatr Soc. 1998; 46: 1423–1424.
Brass LM, Krumholz HM, Scinto JD, Mathur D, Radford M. Warfarin use following ischemic stroke among Medicare patients with atrial fibrillation. Arch Intern Med. 1998; 158: 2093–2100.
Brass LM, Krumholz HM, Scinto JM, Radford M. Warfarin use among patients with atrial fibrillation. Stroke. 1997; 28: 2382–2389.
Whittle J, Wickenheiser L, Venditti LN. Is warfarin underused in the treatment of elderly persons with atrial fibrillation; Arch Intern Med. 1997; 157: 441–445.
Stafford RS, Singer DE. National patterns of warfarin use in atrial fibrillation. Arch Intern Med. 1996; 156: 2537–2541.
Antani MR, Beyth RJ, Covinsky KE, Anderson PA, Miller DG, Cebul RD, Quinn LM, Landefeld CS. Failure to prescribe warfarin to patients with nonrheumatic atrial fibrillation. J Gen Intern Med. 1996; 11: 713–720.
Beyth RJ, Quinn LM, Landefeld CS. Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin. Am J Med. 1998; 105: 91–99.
McMahan DA, Smith DM, Carey MA, Zhou XH. Risk of major hemorrhage for outpatients treated with warfarin. J Gen Intern Med. 1998; 13: 311–316.
Bungard TJ, Ghali WA, McAlister FA, Buchan AM, Cave AJ, Hamilton PG, Mitchell LB, Shuaib A, Teo KK, Tsuyuki RT. Physicians’ perceptions of the benefits and risks of warfarin for patients with nonvalvular atrial fibrillation. CMAJ. 2001; 165: 301–302.
McCrory DC, Matchar DB, Samsa G, Sanders LL, Pritchett EL. Physician attitudes about anticoagulation for nonvalvular atrial fibrillation in the elderly. Arch Intern Med. 1995; 155: 277–281.
Monette J, Gurwitz JH, Rochon PA, Avorn J. Physician attitudes concerning warfarin for stroke prevention in atrial fibrillation: results of a survey of long-term care practitioners. J Am Geriatr Soc. 1997; 45: 1060–1065.
Mina AA, Bair HA, Howells GA, Bendick PJ. Complications of preinjury warfarin use in the trauma patient. J Trauma. 2003; 54: 842–847.
Ferrera PC, Bartfield JM. Outcomes of anticoagulated trauma patients. Am J Emerg Med. 1999; 17: 154–156.
Karni A, Holtzman R, Bass T, Zorman G, Carter L, Rodriguez L, Bennett-Shipman VJ, Lottenberg L. Traumatic head injury in the anticoagulated elderly patient: a lethal combination. Am Surg. 2001; 67: 1098–1100.
Mathiesen T, Benediktsdottir K, Johnsson H, Lindqvist M, von Holst H. Intracranial traumatic and non-traumatic haemorrhagic complications of warfarin treatment. Acta Neurol Scand. 1995; 91: 208–214.
Lawlor DA, Patel R, Ebrahim S. Association between falls in elderly women and chronic diseases and drug use: cross sectional study. BMJ. 2003; 327: 712–717.
Stelfox HT, Redelmeier DA. An analysis of one potential form of health care fraud in Canada. CMAJ. 2003; 169: 118–119.
Juurlink DN, Mamdani M, Kopp A, Laupacis A, Redelmeier DA. Drug-drug interactions among elderly patients hospitalized for drug toxicity. JAMA. 2003; 289: 1652–1658.
Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med. 1999; 341: 1359–1367.
Tu JV, Gong Y. Trends in treatment and outcomes for acute stroke patients in Ontario, 1992–1998. Arch Intern Med. 2003; 163: 293–297.
Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992; 45: 613–619.
Schneeweiss S, Seeger JD, Maclure M, Wang PS, Avorn J, Glynn RJ. Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data. Am J Epidemiol. 2001; 154: 854–864.
Mayo NE, Chockalingam A, Reeder BA, Phillips S. Surveillance for stroke in Canada. Health Rep. 1994; 6: 62–72.
Raiford DS, Perez GS, Garcia Rodriguez LA. Positive predictive value of ICD-9 codes in the identification of cases of complicated peptic ulcer disease in the Saskatchewan hospital automated database. Epidemiology. 1996; 7: 101–104.
Levy AR, Tamblyn RM, Fitchett D, McLeod PJ, Hanley JA. Coding accuracy of hospital discharge data for elderly survivors of myocardial infarction. Can J Cardiol. 1999; 15: 1277–1282.
van Walraven C, Mamdani M, Cohn A, Katib Y, Walker M, Rodger MA. Risk of subsequent thromboembolism for patients with pre-eclampsia. BMJ. 2003; 326: 791–792.
Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med. 1999; 131: 492–501.
Hackam D, Anand SS, Yusuf S. Oral anticoagulant therapy in patients with coronary artery disease. Semin Vasc Med. 2003; 3: 323–332.
Ridker PM, Goldhaber SZ, Danielson E, Rosenberg Y, Eby CS, Deitcher SR, Cushman M, Moll S, Kessler CM, Elliott CG, Paulson R, Wong T, Bauer KA, Schwartz BA, Miletich JP, Bounameaux H, Glynn RJ, for the PREVENT Investigators. Long-term, low-intensity warfarin therapy for the prevention of recurrent venous thromboembolism. N Engl J Med. 2003; 348: 1425–1434.
Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med. 1999; 131: 492–501.
Man-Son-Hing M, Laupacis A. Anticoagulant-related bleeding in older persons with atrial fibrillation: physicians’ fears often unfounded. Arch Intern Med. 2003; 163: 1580–1586.
Man-Son-Hing M, Nichol G, Lau A, Laupacis A. Choosing antithrombotic therapy for elderly patients with atrial fibrillation who are at risk for falls. Arch Intern Med. 1999; 159: 677–685.
van Doorn C, Gruber-Baldini AL, Zimmerman S, Richard Hebel J, Port CL, Baumgarten M, Quinn CC, Taler G, May C, Magaziner J. Dementia as a risk factor for falls and fall injuries among nursing home residents. J Am Geriatr Soc. 2003; 51: 1213–1218.
Guo Z, Wills P, Viitanen M, Fastbom J, Winblad B. Cognitive impairment, drug use, and the risk of hip fracture in persons over 75 years old: a community-based prospective study. Am J Epidemiol. 1998; 148: 887–892.
Campbell AJ, Borrie MJ, Spears GF. Risk factors for falls in a community-based prospective study of people 70 years and older. J Gerontol. 1989; 44: M112–M117.
Rapport LJ, Hanks RA, Millis SR, Deshpande SA. Executive functioning and predictors of falls in the rehabilitation setting. Arch Phys Med Rehabil. 1998; 79: 629–633.
Sudlow M, Thomson R, Kenny RA, Rodgers H. A community survey of patients with atrial fibrillation: associated disabilities and treatment preferences. Br J Gen Pract. 1998; 48: 1775–1778.
van der Meer FJ, Briet E, Vandenbroucke JP, Sramek DI, Versluijs MH, Rosendaal FR. The role of compliance as a cause of instability in oral anticoagulant therapy. Br J Haematol. 1997; 98: 893–900.
Kumar S, Haigh JR, Rhodes LE, Peaker S, Davies JA, Roberts BE, Feely MP. Poor compliance is a major factor in unstable outpatient control of anticoagulant therapy. Thromb Haemost. 1989; 62: 729–732.
Milzman DP, Boulanger BR, Rodriguez A, Soderstrom CA, Mitchell KA, Magnant CM. Pre-existing disease in trauma patients: a predictor of fate independent of age and injury severity score. J Trauma. 1992; 32: 236–243.
Zietlow SP, Capizzi PJ, Bannon MP, Farnell MB. Multisystem geriatric trauma. J Trauma. 1994; 37: 985–988.
Sacco WJ, Copes WS, Bain LW Jr, MacKenzie EJ, Frey CF, Hoyt DB, Weigelt JA, Champion HR. Effect of preinjury illness on trauma patient survival outcome. J Trauma. 1993; 35: 538–542.
Yamaguchi T, for the Japanese Nonvalvular Atrial Fibrillation-Embolism Secondary Prevention Cooperative Study Group. Optimal intensity of warfarin therapy for secondary prevention of stroke in patients with nonvalvular atrial fibrillation : a multicenter, prospective, randomized trial. Stroke. 2000; 31: 817–821.
Morocutti C, Amabile G, Fattapposta F, Nicolosi A, Matteoli S, Trappolini M, Cataldo G, Milanesi G, Lavezzari M, Pamparana F, Coccheri S, for the SIFA (Studio Italiano Fibrillazione Atriale) Investigators. Indobufen versus warfarin in the secondary prevention of major vascular events in nonrheumatic atrial fibrillation. Stroke. 1997; 28: 1015–1021.
Wojcik R, Cipolle MD, Seislove E, Wasser TE, Pasquale MD. Preinjury warfarin does not impact outcome in trauma patients. J Trauma. 2001; 51: 1147–1151.(Daniel G. Hackam, BSc, MD)
the Department of Health Policy, Management and Evaluation and Department of Medicine (D.G.H., D.A.R.), Sunnybrook & Women’s College Health Sciences Centre, Toronto, Canada.
Abstract
Background— Warfarin therapy is often withheld from elderly patients who fall or otherwise experience injury because of concerns regarding the long-term risk of hemorrhage in these individuals. We studied whether stopping warfarin after trauma is associated with a higher risk of subsequent adverse cardiovascular events.
Methods and Results— We conducted a retrospective, population-based, cohort study using linked administrative databases in the province of Ontario, Canada for the years 1992 to 2001. A total of 8450 elderly patients (age >65 years) who survived an incident of major trauma and were receiving warfarin before injury were followed up for a mean of 3.3 years. During the 6-month interval after trauma, 1827 (22%) patients discontinued warfarin, whereas 6623 (78%) patients continued warfarin. Warfarin cessation was not associated with an increased risk of subsequent stroke (hazard ratio [HR] 0.99, 95% CI 0.82 to 1.21) or myocardial infarction (HR 0.94, 95% CI 0.74 to 1.20) but was associated with a lower risk of major hemorrhage (HR 0.69, 95% CI 0.54 to 0.88) and a higher risk of venous thromboembolism (HR 1.59, 95% CI 1.07 to 2.36). Adjustment for baseline demographics, stroke risk factors, other comorbidities, and characteristics of the trauma did not materially change these findings. On-treatment analyses yielded similar results.
Conclusions— Cessation of warfarin in elderly patients after major trauma was not associated with an increased risk of arterial thrombotic events but was associated with a significantly increased risk of venous thromboembolism.
Key Words: anticoagulants ; epidemiology ; stroke ; follow-up studies
Introduction
Warfarin is an oral anticoagulant widely used for preventing thromboembolism in patients with atrial fibrillation, prosthetic heart valves, and venous thrombosis.1 Despite its proven efficacy for preventing strokes, warfarin is frequently withheld from elderly patients because of concerns about major hemorrhage.2–9 Factors that increase the risk of bleeding are often prevalent in the elderly, including prior gastrointestinal hemorrhage, erratic compliance, and a tendency to fall.10,11 The tradeoff between benefits and risks of warfarin leads to considerable uncertainty about the decision to prescribe this therapy to older patients in the community.12,13 In one study of physician attitudes regarding warfarin for stroke prevention, for example, the most commonly cited contraindication was an increased risk of falls (reported by 71% of respondents).14
Patients who sustain major trauma while receiving warfarin may bleed more seriously from their injuries, particularly when craniocerebral trauma occurs.15–18 However, the medical literature offers remarkably little evidence to guide physicians about resumption of long-term oral anticoagulation in patients who survive major trauma. The potential protection from stroke and other thrombotic events favors the reinstitution of long-term anticoagulation. Conversely, the occurrence of trauma may be a marker of poor functional status, comorbidity, or unreliable patient behavior that may seriously compromise the benefits of long-term warfarin therapy.19
We hypothesized that elderly patients who discontinue warfarin after major trauma face an increased risk of adverse cardiovascular events. Specifically, our theory was that major trauma does not reverse the rationale for resuming warfarin. By monitoring a population-based cohort of elderly patients who sustained major trauma while coincidentally receiving warfarin, we assessed the long-term effects associated with discontinuation of warfarin after trauma. We did not examine the short-term (<6 months) benefits and risks of oral anticoagulation after trauma for prophylaxis or other indications.
Methods
Setting and Design
We conducted a population-based, retrospective, cohort study by linking multiple administrative healthcare databases over a 9-year period (January 1, 1992, to January 1, 2001) in the province of Ontario. Throughout the study, Ontario was Canada’s most populous province, with 12 million residents, of whom 1.5 million were aged 65 years or older. Elderly patients in Ontario had universal access to hospital care, physicians’ services, and prescription drug coverage. In addition, healthcare records could be analyzed with encrypted health card numbers to track individual patients over time.20 Ethical approval for the study was obtained from the Sunnybrook and Women’s College Health Sciences Centre Research Ethics Board.
Data Sources
We used 4 large databases: the Ontario Drug Benefit database, which records prescription medications dispensed to elderly patients in Ontario; the Canadian Institute for Health Information (CIHI) discharge abstract database, which records hospital admissions, including detailed diagnostic and procedural information; the Ontario Health Insurance Plan database, which provides information on physician claims for inpatient and outpatient services; and the Ontario Registered Persons database, which contains vital statistics on all residents in the province. These 4 databases have been used extensively to study population-based health outcomes.21,22
Patients
Patients were included if they were 66 years of age or older at the time of major trauma, defined as a hospitalization for which trauma was the most responsible diagnosis. We used International Classification of Disease, Ninth Revision, Clinical Modification (ICD-9-CM) E codes 800 through 999 to define and categorize all subtypes of trauma. We excluded patients whose admission was primarily related to late effects of trauma (E code 929), complications of medical procedures (E codes 930 to 949), or poisonings (E codes 849 to 879). We focused on patients who had recently received warfarin, defined as at least 1 prescription for warfarin within 6 months before the trauma. We excluded those patients who died in the hospital or who did not survive for at least 6 months after discharge to remove deaths that might be linked to the incident trauma and recovery.
Comorbidities
The frequency of major stroke risk factors and other comorbid diseases was determined by analysis of inpatient and outpatient databases before trauma. In accord with past research, the 7 prespecified stroke risk factors were atrial fibrillation, a prosthetic heart valve, diabetes mellitus, hypertension, heart failure, previous stroke or transient ischemic attack, and coronary artery disease.23 Additional prespecified comorbidities were a history of falls, alcoholism, dementia, previous major hemorrhage, malignancy, and renal insufficiency. The Charlson Comorbidity Index score was used to quantify overall comorbidity.24 Two additional generic indices of comorbidity were provided by the number of hospital admissions in the preceding 3 years and the number of distinct medications prescribed in the past year.25
Outcomes
The primary end point was the first occurrence of ischemic stroke, defined as admission to hospital with ischemic stroke as the most responsible diagnosis (ICD-9-CM codes 433, 434, and 436). The accuracy rate of coding for stroke in the CIHI database exceeds 90%.23,26 The 4 secondary end points examined were major hemorrhage, acute myocardial infarction, venous thromboembolism, and death. Major hemorrhage was defined according to an admission with the most responsible diagnosis of hemorrhage or for a procedure to repair hemorrhage (appendix A for ICD-9-CM diagnosis and procedure codes for hemorrhage). Myocardial infarction was defined as an admission with the most responsible diagnosis of myocardial infarction (ICD-9-CM 410). These diagnosis codes have a positive predictive value of 90% to 96%.27,28 Venous thromboembolism was defined as an admission with a most responsible diagnosis of pulmonary embolism (ICD-9-CM 415.1) or deep vein thrombosis (ICD-9-CM 325, 437.6, 451 to 453). These diagnosis codes have a sensitivity of 97% and a specificity of 90% for true thromboembolic disease.29 All-cause mortality was obtained from the Registered Persons database.
Subgroups
Predefined subgroups were identified to compare event rates in 9 settings: a history of atrial fibrillation; history of stroke or transient ischemic attack; history of diabetes mellitus; history of hypertension; history of congestive heart failure; history of coronary artery disease; history of malignancy; absence of head injury on trauma; and trauma resulting from a fall. The intent of these subgroup analyses was to examine warfarin therapy in distinct patient populations, particularly in those settings established by the literature as representing elevated risk of major vascular events.
Statistical Analysis
We defined the period of observation as beginning 6 months after discharge. Follow-up ended at the time of a primary event, death, or the end of the study if the patient course was uneventful (March 31, 2002). The primary analysis compared patients who discontinued warfarin by 6 months after discharge with patients who continued taking warfarin. Crude differences between the 2 groups in the risk of outcomes were calculated as hazard ratios (HRs) with Cox regression models and subsequently adjusted for demographic factors, stroke risk factors, and other covariates (comorbidities and characteristics of the trauma).
We also conducted an on-treatment analysis by checking for warfarin claims in the prescription database within 6 months immediately before the occurrence of an outcome. In the treated group, we regarded on-treatment outcomes as those for which at least 1 claim for warfarin was made within 6 months before the event. Similarly, for those who discontinued warfarin, we retained only outcomes for which no claim for warfarin was made within the 6-month preevent interval. The intent of the on-treatment analysis was to examine the consistency of findings with a more precise measure of patient compliance.
Results
During the 9-year study period, 11 139 patients receiving warfarin sustained trauma and were treated at 252 hospitals in Ontario. Of these, 1284 (11%) died in the hospital, and an additional 1405 (13%) died within 6 months of discharge; 61 of these deaths were due to stroke. Of the 8450 long-term survivors, 6623 (78%) resumed the use of warfarin during the 6-month period after discharge, whereas 1827 (22%) did not resume warfarin (Table). Collectively, this cohort provided 27 530 person-years of follow-up after major trauma (mean duration of follow-up, 3.3 years). The majority (82%) of trauma episodes were due to falls.
Characteristics of Patients Prescribed Warfarin Before Major Trauma
Patients who discontinued using long-term warfarin after trauma were similar to those who continued its use in basic demographics, stroke risk factors, and other comorbidities. The major exception was that atrial fibrillation was somewhat less common in those who discontinued the use of warfarin. The subtypes of trauma were fairly evenly distributed between the 2 groups. A past history of major hemorrhage was equally common in both groups.
A total of 592 patients sustained an ischemic stroke during follow-up, equivalent to a rate of 2.2% per year. We observed no difference in the risk of stroke in the patients who discontinued warfarin compared with those who continued its use, even after adjustment for demographic factors, stroke risk factors, and other covariates (Figures 1 and 2). Although the rate of stroke varied substantially across the different subgroups, in no subgroup was warfarin cessation associated with an increased risk of stroke (including patients with atrial fibrillation or those with prior cerebrovascular disease). The on-treatment analysis continued to show no evidence of an association between warfarin cessation and an increased risk of ischemic stroke.
A total of 399 patients sustained an acute myocardial infarction during follow-up, equivalent to an annual incidence of 1.5%. As with stroke, we observed no apparent relation between warfarin discontinuation and the subsequent long-term risk of myocardial infarction (Figure 3). This lack of association was apparent in both the crude and adjusted analyses, all subgroups, and the on-treatment analysis. There were 4203 deaths in the trauma cohort, with an overall mortality rate of 16% per year. We found a slight but statistically significant increase in mortality in those who discontinued using warfarin (HR 1.09, 95% CI 1.02 to 1.18), a finding which was no longer significant after adjustment for covariates (HR 1.05, 95% CI 0.98 to 1.13).
The long-term risk of major hemorrhage for trauma patients who discontinued warfarin use was 1.3% annually, which was significantly lower than the rate for trauma patients who resumed using warfarin (1.9%; P=0.0025 for difference). In the unadjusted model, the risk of major hemorrhage was increased by 45% in those who continued using warfarin, an association that persisted after adjustment for age, sex, stroke risk factors, comorbidities, and trauma characteristics (Figure 4). The increased risk of hemorrhage for trauma patients who continued to use warfarin was consistent across most major subgroups, accentuated in patients with a previous history of stroke or transient ischemic attack, and further strengthened in the on-treatment analysis.
There were 115 episodes of venous thromboembolism over the course of follow-up, equivalent to a rate of 0.4% annually. Discontinuation of warfarin use was associated with a significant increase in the risk of venous thromboembolism (Figure 5). This finding was strengthened in the on-treatment analysis and was consistent across most subgroups. Overall, the increase in venous thromboembolism primarily reflected a substantial increase in hospitalizations for deep vein thrombosis (HR 1.94, 95% CI 1.18 to 3.18), with no significant increase in pulmonary embolism (HR 1.01, 95% CI 0.52 to 1.97).
A number of risk factors for the combined outcome of death or stroke were corroborated in the anticipated manner in the full cohort. As expected, age was the strongest predictor and was associated with a 44% increase in risk for each decade increase in age (95% CI 38% to 51%). Similarly, male gender (HR 1.21, 95% CI 1.15 to 1.25), past stroke or transient ischemic attack (HR 1.12, 95% CI 1.06 to 1.19), renal insufficiency (HR 1.11, 95% CI 1.03 to 1.19), and dementia (HR 1.19, 95% CI 1.10 to 1.29) were further independent risk factors for the combined end point. Similar results were noted in the atrial fibrillation subgroup.
Discussion
Multiple studies support the efficacy and safety of warfarin for preventing stroke, myocardial infarction, and venous thromboembolism.30–32 In particular, warfarin reduces the risk of ischemic stroke by approximately two thirds in patients with atrial fibrillation.33 In many settings, the moderately raised incidence of bleeding related to warfarin is greatly outweighed by the benefits in preventing stroke, disability, and death34,35; however, controversy surrounds the risk-benefit ratio of warfarin use after trauma, and we found that 1 in 5 patients discontinued their use of warfarin after hospitalization for major trauma. Moreover, those patients who discontinued long-term warfarin experienced fewer adverse events than were observed in the placebo groups of prior randomized trials.
The largest limitation of the present study is that the lack of random allocation to warfarin therapy raises the possibility of selection bias; furthermore, poor compliance might slant the analysis against warfarin. However, 4 additional findings suggest that hidden confounding is unlikely to explain the results. First, we made adjustments for many factors that are known to be prognostically important. These adjustments did not affect the results of our analysis. Second, analyses of specific high-risk subgroups of patients were consistent with the overall cohort. Third, bias due to poor compliance would not explain why all the main findings were replicated in the on-treatment analyses. Finally, the significantly increased risk of hemorrhage in those who continued the use of warfarin suggests at least some degree of anticoagulation in these patients, thus arguing against major noncompliance.
One possible explanation for our findings is behavioral. Elderly patients who are prone to trauma are more likely to have comorbidities (such as dementia) and neurobehavioral attributes (such as poor attention) that may predispose to reduced efficacy with warfarin.36–39 Such characteristics could result in poor compliance with warfarin therapy and potential lapses in follow-up, monitoring, and dose adjustment.40–42 These patients would thus be predisposed to subtherapeutic (stroke-promoting) and supratherapeutic (hemorrhage-promoting) levels of anticoagulation. Consistent with this explanation, many patients in the present study cohort had a history of dementia, polypharmacy, malignancy, and other comorbid conditions that may compromise the efficacy and safety of warfarin.
A second possible explanation is that patients with severe cerebrovascular disease might not survive major trauma or its sequelae.43–45 This might also explain why the annual incidence of ischemic stroke observed in the present study cohort was lower than expected from population studies. Hence, major trauma may have depleted the patients who were most prone to strokes (the same subset of patients most likely to benefit from warfarin).46,47 Because the observation period of the present study began at 6 months after trauma, our results do not address whether warfarin is beneficial in the early posttrauma period. Intriguingly, however, we found a significant increase in the long-term risk of venous thromboembolism in trauma patients who discontinued warfarin. Although randomized controlled trials have shown significant reductions in the risk of venous thrombosis with early anticoagulation after trauma, this is the first study, to the best of our knowledge, that suggests long-term warfarin use may also be protective.
The literature on the safety of resuming warfarin after trauma is remarkably limited. A recent decision analysis concluded that a patient would need to fall 300 times per year to outweigh the benefits of long-term warfarin use for stroke prevention35; however, that decision analysis was based on randomized controlled trials that generally excluded trauma patients. Moreover, the presence of comorbidities that increase the risk of hemorrhage and cluster with the risk of trauma were not considered in the analysis. A recent cohort study of unselected trauma patients found substantially lower rates of discharge to the community in patients using warfarin at the time of trauma.48 Moreover, warfarin users had significantly worse functional outcomes than nonusers of warfarin, particularly those who sustained head injury at the time of trauma.
We found that discontinuation of long-term warfarin use after major trauma was not associated with an increased risk of stroke or myocardial infarction but was associated with a significant increase in the risk of venous thromboembolism. Further research is necessary to replicate our work. In the meantime, physicians who prescribe long-term warfarin to patients with a history of trauma should carefully weigh the competing risks of hemorrhage and venous thromboembolism.
Acknowledgments
Dr Hackam was supported by a Canadian Institutes for Health Research Fellowship Award. Dr Redelmeier was supported by a Canada Research Chair in Medical Decision Sciences and the Error Management Unit of Sunnybrook and Women’s College. We thank Mark Crowther, John Blakely, Andreas Laupacis, Bill Geerts, Steven Shumak, David Juurlink, Chaim Bell, and Damon Scales for their helpful comments on previous drafts of the manuscript and Tracy Willson for secretarial support.
References
Hirsh J, Fuster V, Ansell J, Halperin JL. American Heart Association/American College of Cardiology Foundation guide to warfarin therapy. Circulation. 2003; 107: 1692–1711.
McCormick D, Gurwitz JH, Goldberg RJ, Becker R, Tate JP, Elwell A, Radford MJ. Prevalence and quality of warfarin use for patients with atrial fibrillation in the long-term care setting. Arch Intern Med. 2001; 161: 2458–2463.
Cohen N, Almoznino-Sarafian D, Alon I, Gorelik O, Koopfer M, Chachashvily S, Shteinshnaider M, Litvinjuk V, Modai D. Warfarin for stroke prevention still underused in atrial fibrillation: patterns of omission. Stroke. 2000; 31: 1217–1222.
Mendelson G, Aronow WS. Underutilization of warfarin in older persons with chronic nonvalvular atrial fibrillation at high risk for developing stroke. J Am Geriatr Soc. 1998; 46: 1423–1424.
Brass LM, Krumholz HM, Scinto JD, Mathur D, Radford M. Warfarin use following ischemic stroke among Medicare patients with atrial fibrillation. Arch Intern Med. 1998; 158: 2093–2100.
Brass LM, Krumholz HM, Scinto JM, Radford M. Warfarin use among patients with atrial fibrillation. Stroke. 1997; 28: 2382–2389.
Whittle J, Wickenheiser L, Venditti LN. Is warfarin underused in the treatment of elderly persons with atrial fibrillation; Arch Intern Med. 1997; 157: 441–445.
Stafford RS, Singer DE. National patterns of warfarin use in atrial fibrillation. Arch Intern Med. 1996; 156: 2537–2541.
Antani MR, Beyth RJ, Covinsky KE, Anderson PA, Miller DG, Cebul RD, Quinn LM, Landefeld CS. Failure to prescribe warfarin to patients with nonrheumatic atrial fibrillation. J Gen Intern Med. 1996; 11: 713–720.
Beyth RJ, Quinn LM, Landefeld CS. Prospective evaluation of an index for predicting the risk of major bleeding in outpatients treated with warfarin. Am J Med. 1998; 105: 91–99.
McMahan DA, Smith DM, Carey MA, Zhou XH. Risk of major hemorrhage for outpatients treated with warfarin. J Gen Intern Med. 1998; 13: 311–316.
Bungard TJ, Ghali WA, McAlister FA, Buchan AM, Cave AJ, Hamilton PG, Mitchell LB, Shuaib A, Teo KK, Tsuyuki RT. Physicians’ perceptions of the benefits and risks of warfarin for patients with nonvalvular atrial fibrillation. CMAJ. 2001; 165: 301–302.
McCrory DC, Matchar DB, Samsa G, Sanders LL, Pritchett EL. Physician attitudes about anticoagulation for nonvalvular atrial fibrillation in the elderly. Arch Intern Med. 1995; 155: 277–281.
Monette J, Gurwitz JH, Rochon PA, Avorn J. Physician attitudes concerning warfarin for stroke prevention in atrial fibrillation: results of a survey of long-term care practitioners. J Am Geriatr Soc. 1997; 45: 1060–1065.
Mina AA, Bair HA, Howells GA, Bendick PJ. Complications of preinjury warfarin use in the trauma patient. J Trauma. 2003; 54: 842–847.
Ferrera PC, Bartfield JM. Outcomes of anticoagulated trauma patients. Am J Emerg Med. 1999; 17: 154–156.
Karni A, Holtzman R, Bass T, Zorman G, Carter L, Rodriguez L, Bennett-Shipman VJ, Lottenberg L. Traumatic head injury in the anticoagulated elderly patient: a lethal combination. Am Surg. 2001; 67: 1098–1100.
Mathiesen T, Benediktsdottir K, Johnsson H, Lindqvist M, von Holst H. Intracranial traumatic and non-traumatic haemorrhagic complications of warfarin treatment. Acta Neurol Scand. 1995; 91: 208–214.
Lawlor DA, Patel R, Ebrahim S. Association between falls in elderly women and chronic diseases and drug use: cross sectional study. BMJ. 2003; 327: 712–717.
Stelfox HT, Redelmeier DA. An analysis of one potential form of health care fraud in Canada. CMAJ. 2003; 169: 118–119.
Juurlink DN, Mamdani M, Kopp A, Laupacis A, Redelmeier DA. Drug-drug interactions among elderly patients hospitalized for drug toxicity. JAMA. 2003; 289: 1652–1658.
Alter DA, Naylor CD, Austin P, Tu JV. Effects of socioeconomic status on access to invasive cardiac procedures and on mortality after acute myocardial infarction. N Engl J Med. 1999; 341: 1359–1367.
Tu JV, Gong Y. Trends in treatment and outcomes for acute stroke patients in Ontario, 1992–1998. Arch Intern Med. 2003; 163: 293–297.
Deyo RA, Cherkin DC, Ciol MA. Adapting a clinical comorbidity index for use with ICD-9-CM administrative databases. J Clin Epidemiol. 1992; 45: 613–619.
Schneeweiss S, Seeger JD, Maclure M, Wang PS, Avorn J, Glynn RJ. Performance of comorbidity scores to control for confounding in epidemiologic studies using claims data. Am J Epidemiol. 2001; 154: 854–864.
Mayo NE, Chockalingam A, Reeder BA, Phillips S. Surveillance for stroke in Canada. Health Rep. 1994; 6: 62–72.
Raiford DS, Perez GS, Garcia Rodriguez LA. Positive predictive value of ICD-9 codes in the identification of cases of complicated peptic ulcer disease in the Saskatchewan hospital automated database. Epidemiology. 1996; 7: 101–104.
Levy AR, Tamblyn RM, Fitchett D, McLeod PJ, Hanley JA. Coding accuracy of hospital discharge data for elderly survivors of myocardial infarction. Can J Cardiol. 1999; 15: 1277–1282.
van Walraven C, Mamdani M, Cohn A, Katib Y, Walker M, Rodger MA. Risk of subsequent thromboembolism for patients with pre-eclampsia. BMJ. 2003; 326: 791–792.
Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med. 1999; 131: 492–501.
Hackam D, Anand SS, Yusuf S. Oral anticoagulant therapy in patients with coronary artery disease. Semin Vasc Med. 2003; 3: 323–332.
Ridker PM, Goldhaber SZ, Danielson E, Rosenberg Y, Eby CS, Deitcher SR, Cushman M, Moll S, Kessler CM, Elliott CG, Paulson R, Wong T, Bauer KA, Schwartz BA, Miletich JP, Bounameaux H, Glynn RJ, for the PREVENT Investigators. Long-term, low-intensity warfarin therapy for the prevention of recurrent venous thromboembolism. N Engl J Med. 2003; 348: 1425–1434.
Hart RG, Benavente O, McBride R, Pearce LA. Antithrombotic therapy to prevent stroke in patients with atrial fibrillation: a meta-analysis. Ann Intern Med. 1999; 131: 492–501.
Man-Son-Hing M, Laupacis A. Anticoagulant-related bleeding in older persons with atrial fibrillation: physicians’ fears often unfounded. Arch Intern Med. 2003; 163: 1580–1586.
Man-Son-Hing M, Nichol G, Lau A, Laupacis A. Choosing antithrombotic therapy for elderly patients with atrial fibrillation who are at risk for falls. Arch Intern Med. 1999; 159: 677–685.
van Doorn C, Gruber-Baldini AL, Zimmerman S, Richard Hebel J, Port CL, Baumgarten M, Quinn CC, Taler G, May C, Magaziner J. Dementia as a risk factor for falls and fall injuries among nursing home residents. J Am Geriatr Soc. 2003; 51: 1213–1218.
Guo Z, Wills P, Viitanen M, Fastbom J, Winblad B. Cognitive impairment, drug use, and the risk of hip fracture in persons over 75 years old: a community-based prospective study. Am J Epidemiol. 1998; 148: 887–892.
Campbell AJ, Borrie MJ, Spears GF. Risk factors for falls in a community-based prospective study of people 70 years and older. J Gerontol. 1989; 44: M112–M117.
Rapport LJ, Hanks RA, Millis SR, Deshpande SA. Executive functioning and predictors of falls in the rehabilitation setting. Arch Phys Med Rehabil. 1998; 79: 629–633.
Sudlow M, Thomson R, Kenny RA, Rodgers H. A community survey of patients with atrial fibrillation: associated disabilities and treatment preferences. Br J Gen Pract. 1998; 48: 1775–1778.
van der Meer FJ, Briet E, Vandenbroucke JP, Sramek DI, Versluijs MH, Rosendaal FR. The role of compliance as a cause of instability in oral anticoagulant therapy. Br J Haematol. 1997; 98: 893–900.
Kumar S, Haigh JR, Rhodes LE, Peaker S, Davies JA, Roberts BE, Feely MP. Poor compliance is a major factor in unstable outpatient control of anticoagulant therapy. Thromb Haemost. 1989; 62: 729–732.
Milzman DP, Boulanger BR, Rodriguez A, Soderstrom CA, Mitchell KA, Magnant CM. Pre-existing disease in trauma patients: a predictor of fate independent of age and injury severity score. J Trauma. 1992; 32: 236–243.
Zietlow SP, Capizzi PJ, Bannon MP, Farnell MB. Multisystem geriatric trauma. J Trauma. 1994; 37: 985–988.
Sacco WJ, Copes WS, Bain LW Jr, MacKenzie EJ, Frey CF, Hoyt DB, Weigelt JA, Champion HR. Effect of preinjury illness on trauma patient survival outcome. J Trauma. 1993; 35: 538–542.
Yamaguchi T, for the Japanese Nonvalvular Atrial Fibrillation-Embolism Secondary Prevention Cooperative Study Group. Optimal intensity of warfarin therapy for secondary prevention of stroke in patients with nonvalvular atrial fibrillation : a multicenter, prospective, randomized trial. Stroke. 2000; 31: 817–821.
Morocutti C, Amabile G, Fattapposta F, Nicolosi A, Matteoli S, Trappolini M, Cataldo G, Milanesi G, Lavezzari M, Pamparana F, Coccheri S, for the SIFA (Studio Italiano Fibrillazione Atriale) Investigators. Indobufen versus warfarin in the secondary prevention of major vascular events in nonrheumatic atrial fibrillation. Stroke. 1997; 28: 1015–1021.
Wojcik R, Cipolle MD, Seislove E, Wasser TE, Pasquale MD. Preinjury warfarin does not impact outcome in trauma patients. J Trauma. 2001; 51: 1147–1151.(Daniel G. Hackam, BSc, MD)