点击显示 收起
AstraZeneca (P.S.J.M.), Alderley Park, UK
AstraZeneca R&D Lund (F.L.A.), Sweden
Guy’s, King’s & St Thomas’ School of Medicine (L.K), London UK.
Abstract
Background and Purpose— Stroke outcomes in patients with atrial fibrillation (AF) tend to be worse than those in patients without AF. The objective of this study was to evaluate whether the cost benefits of anticoagulation for stroke prevention in AF may currently be underestimated by existing economic models that do not distinguish between different stroke outcomes.
Methods— A literature review was conducted in 3 areas: (1) studies comparing stroke outcomes in AF and non-AF patients; (2) studies providing long-term cost of stroke estimates; and (3) studies modeling the cost-effectiveness of anticoagulation with a vitamin K antagonist (eg, warfarin) in AF patients.
Results— There is considerable evidence that stroke in AF patients has a worse outcome than in patients without AF, including higher mortality, severity, and recurrence rates, and greater functional impairment and dependency. Estimates of the long-term cost of stroke of different severities were between US $24 991 for a mild stroke over 5 years and US $142 251 for a major ischemic stroke over a lifetime (2004 prices). The cost of a severe ischemic stroke may typically be 3-times that of mild stroke. However, cost-effectiveness models for anticoagulation in patients with AF have used average (not AF-specific) cost-of-stroke data, and most have used stroke severity distributions derived from clinical trials, which may differ from those in clinical practice.
Conclusions— Existing economic models underestimate the cost benefits of anticoagulation for stroke prevention because they do not adjust for poorer outcomes associated with cardioembolic strokes.
Key Words: anticoagulants arrhythmia cost-benefit analysis primary prevention stroke
Introduction
Recent years have seen an increasing emphasis on the use of health economic data, in conjunction with evidence of clinical effectiveness, for the development of clinical guidelines.1,2 This is particularly important for stroke, a disease characterized by a high prevalence, as well as a high burden of illness and economic cost, but in which preventive strategies such as blood pressure control, smoking cessation, diet and exercise have been shown to be effective. Of the preventive measures for stroke in patients with atrial fibrillation (AF), anticoagulation has been associated with the greatest benefits, resulting in a 67% to 80% reduction in the risk of stroke.3 Atrial fibrillation affects 5% of all elderly people and increases the risk of stroke 5-fold compared with patients in normal sinus rhythm.4,5 Because both the incidence of AF and the stroke risk attributable to AF increase with age (the latter being 1.5% at age 50 to 59 years; 9.9% at 70 to 79 years; and 23.5% at 80 to 89 years), AF and stroke will pose major societal costs over the coming decades.4,6–9
Cost-effectiveness models of anticoagulation for the prevention of stroke in patients with AF are based on the incremental change in costs and outcomes (eg, quality-adjusted life years) associated with anticoagulant therapy, treatment-related bleeding events, and ischemic strokes. The objective of this review was to evaluate whether the cost benefits of anticoagulation for stroke prevention in AF may currently be underestimated by existing economic models that do not account for the different outcomes of cardioembolic stroke in people with AF compared with strokes in patients without AF. A literature review was conducted in 3 areas: (1) studies comparing stroke outcomes in patients with and without AF; (2) studies providing long-term cost-of-stroke estimates; and (3) studies modeling the cost-effectiveness of anticoagulation.
Materials and Methods
We conducted a literature review of studies comparing stroke outcomes in patients with and without AF, studies estimating the cost of stroke, and studies modeling the cost-effectiveness of anticoagulation. Searches were conducted on MEDLINE in December 2003, including results from 1966 up to that date, with the following search criteria: "stroke outcome/severity/cost" AND "atrial fibrillation;" "cost/cost-effectiveness/economic evaluation" AND "anticoagulation" AND "atrial fibrillation." The initial search identified 114 articles. Studies reporting primary data analysis, economic models, or cost-of-stroke data with at least 1 year of follow-up were selected for inclusion. The final analysis included 14 studies comparing stroke outcomes in AF and non-AF patients, 15 cost-of-stroke studies, and 9 cost-effectiveness modeling studies.
Results
Stroke Outcomes in AF
Earlier studies of stroke outcomes focused on mortality and stroke recurrence rates, whereas later studies have considered stroke severity in addition to functional and neurological impairment among stroke survivors. Collectively, studies analyzing 10 000 strokes clearly demonstrate that stroke in patients with AF has worse outcomes than strokes in patients in sinus rhythm (non-AF stroke) (Table 1).10–23
Overall, ischemic stroke associated with AF is approximately twice as likely to be fatal as non-AF stroke. Adjusting for age and comorbidities, short-term mortality rates from stroke (in-hospital or 1 month after stroke) were significantly increased in the presence of AF10,12–17,19–21 (19% to 35%) compared with no AF (5% to 14%; Table 1).10,12–14,16,19 Atrial fibrillation also independently predicted higher 3-month stroke mortality after controlling for age, gender, and other baseline variables,10 higher mortality 1 year after stroke in multivariate analyses,18,19,22 and higher relative risk of death at 6 months.13
At 6 months11 and 1 year19,22 after an initial stroke, recurrence rates are 2-times higher in patients with AF compared with those without AF (Table 1). Various studies have demonstrated more severe (larger) brain lesions,12,17,20 a greater incidence of abnormal computed tomograms,13 and more patients with cortical involvement (68% versus 43%)20 or an anterior circulation infarct (34% versus 25%)10 among stroke patients with AF. Stroke was more likely to be severe (evaluated on the basis of neurological impairment during the initial hospitalization) or fatal in subjects with AF,19 and it was associated with worse scores on the Canadian Neurological Scale,17 Scandinavian Neurological Stroke Scale,20 and National Institutes of Health Stroke Scale.21 The frequency of bedridden state after stroke was also higher among AF patients (41% versus 24%), was correlated with other clinical severity data, and independent of age and other stroke risk factors.23
The Copenhagen study20 found significantly longer hospital stays from the time of acute admission to the end of rehabilitation among AF stroke patients compared with non-AF stroke patients in a stroke unit (50 versus 40 days). However, in the European BIOMED study,10 mean length of stay in general acute care hospitals was not significantly different between groups, but inter-unit variability in routine length of stay across 22 stroke units in 7 countries may have contributed to masking any differences between AF stroke and non-AF stroke patients.
Stroke in patients with AF is associated with more impairment measured by the modified Rankin Scale at 30 days,17 3 months,10 and 1 year,22 as well as with greater dependency measured on the Barthel Index in the acute phase,19 at 3 months,10,19 and at 6 months (Table 1).19 For example, 33% of non-AF patients had acute severe dependency, and this decreased to 16% at 3 and 6 months, whereas among AF patients 73% had severe dependency in the acute phase, 58% at 3 months, and 36% at 6 months.19 Two studies10,20 have shown that a significantly lower proportion of AF stroke patients are discharged to their own home (Table 1), although these analyses did not adjust for age. In addition, 30% more patients with than without AF were institutionalized (9% versus 6%, respectively).10
Cost of Stroke
Long-term cost estimates show that the cost of stroke is substantial in all countries.24–40 Lifetime estimates of the cost of a stroke per patient range between US $25 558 (all stroke types) and US $142 251 (major ischemic stroke) after adjustment to dollars and 2004 prices (base costs, Swedish krona 180 000 in 199126 and ￡75 985 in 1996,34 respectively). Three publications reported costs for strokes of different severities (Table 2).26,34,35 The cost increments were between 1.5- and 3-fold. The lowest cost was for all mild strokes over 5 years (US $24 991)35 and the highest was for a major ischemic stroke over a lifetime (US $142 251, as noted).34 Importantly, no publications were found reporting costs of stroke specifically for a population with AF. It has been shown that hospitalized AF patients incur higher costs than matched patients without AF,41 although interventions for the arrhythmia, such as cardioversion, would have contributed to the cost differential in this analysis.
Economics of Anticoagulation
The cost-effectiveness of anticoagulation among AF patients classified as high-risk of stroke is clearly established, whereas cost-effectiveness estimates among patients at lower risk may be more sensitive to the input assumptions of economic models. The findings of the economic models shown in Table 3 represent the current body of cost-effectiveness evidence.33,42–48 The results of all of these studies show that in medium- to high-risk patients with AF or older patients with AF, whether male or female, anticoagulation is dominant (increases quality-adjusted life-years and is cost-saving) or has a low marginal cost-effectiveness ratio (a low additional cost per quality-adjusted life-year gained) in comparison with no anticoagulant therapy33,42,46–48 and in comparison with aspirin,44,45 whereas low-risk AF populations are associated with a higher marginal cost-effectiveness ratio.45
The cost of AF strokes in these models will be characterized by the severity distribution used and the average "unit" costs attached to each stroke severity category. Table 4 shows the distributions of stroke event severity that were used in 5 published economic models. Four use the severity distribution observed in randomized controlled trials of anticoagulation in AF patients rather than from unselected patients in routine clinical practice.33,44,45,47 One of these is based on a single, small study (n=420) in which only 15 strokes occurred. Thomson et al46 state that no valid published studies reported the risk of different outcomes after stroke associated with AF. Only Thomson et al use epidemiological data for severity—based on a personal communication with the Oxfordshire Community Stroke Project. Perhaps as expected, the proportion of fatal strokes in this study using epidemiological data are higher than in those using data from clinical trials (Table 4).
None of the 5 cost-effectiveness models for anticoagulation in patients with AF uses AF-specific stroke cost data. Instead, more generic stroke cost and price estimates are used. As with the long-term estimates illustrated in Table 2, the costs applied to stroke events vary with the sources of data, assumptions, and the extrapolated time periods used. The acute (one-time) cost of severe stroke has been considered to be 2.8-fold46 or 3.5-fold44,45 that of mild stroke. These costs are then weighted based on the assumed severity distribution of strokes. Estimates of the cost of long-term care and the proportion of patients requiring nursing home care used in cost-effectiveness models also vary.44,45,47 Again, these are not based on resource use evaluated from an AF population.
Discussion
Economic models to assess the cost-effectiveness of anticoagulation in AF for the primary prevention of ischemic stroke are based on 3 input parameter estimates, namely the change (compared with most relevant alternative therapy) in cost and quality of life associated with: (1) the anticoagulant therapy itself; (2) adverse (bleeding) events; and (3) ischemic stroke events. These parameters are a function of factors such as the differential risk of events, health care, patient resource use associated with management processes, and treatment process utilities (ie, quality-of-life estimates).
There is considerable evidence that stroke in patients with AF has worse outcomes than non-AF–related strokes, including higher mortality, severity, and recurrence rates and greater functional impairment and dependency. These factors suggest that AF strokes have a higher cost compared with strokes not related to AF. Therefore, the case for cost-effectiveness of anticoagulation therapy, which has been demonstrated in existing economic models, is likely to be even stronger when using higher cost-of-stroke estimates based on events in patients with AF.
The monetary valuation placed on stroke events is a key component of the economic assessment of events that could be avoided by anticoagulation. Large variations arise because of the different ways that stroke services are organized and the different service needs of different subpopulations. In particular, the length of acute stay, type of acute care ward, the existence of hospital-based rehabilitation, and staff skills mix vary considerably and are key drivers of acute care costs.49 Variation in postacute care services is even greater. Impairment and dependency increase the need for health and social services and family support. Discharge from hospital to a residential or nursing home represents the largest share of total costs. However, rehabilitation services and institutional care are not formally provided in many health care systems; therefore, a great deal of the unpaid informal care is not included in the cost estimates.49
The different service needs of different subpopulations are relevant to the potential underestimate of the average costs of stroke typical in AF patients compared with non-AF–related or total stroke events. Data sources for resource use and survival are important to the reliability of cost estimates. For resource use, studies rely on prospective data obtained from stroke registries, clinical or economic trials, and published or other expert opinion. For survival, longer-term epidemiological data are available, but there is a paucity of longer-term observational studies of stroke patients reporting cost data. Most economic studies of stroke follow-up patients for 1 or 2 years and use modeling techniques to extrapolate to lifetime costs (10 to 15 years for this population) based on mortality, progression through disability levels over time, and correlations between disability levels and cost. Generally, cost models based on studies with a longer time period of actual patient-specific resource use data will be more reliable.
One implication of greater mortality among patients with AF-related stroke is a reduced population of survivors incurring long-term costs. However, in economic analyses death remains a negative outcome, so the value of more premature deaths avoided (years of life gained) by stroke prevention among AF patients must be realistically incorporated into cost-effectiveness estimates.50 The increased recurrence rate associated with AF stroke also has to be accurately represented in economic analyses in terms of additional costs and years of life lost, because recurrent strokes tend to be more severe and more often fatal than first strokes.37
It has been demonstrated that initial stroke severity (not age or comorbidity) is a significant predictor of length of initial hospital stay.20,51 In addition, stroke characteristics will influence duration of survival and disability (and consequently the costs of long-term care), whether defined by disease subtype,52 Barthel score,53 or Rankin score.37 Admission to a nursing home, which has a large impact on total cost, is more likely in patients with severe stroke. Evidence exists for a positive correlation between severity of stroke and longer-term stroke costs,49,53 which will have implications for AF-related strokes. Limited evidence is available to compare average survival time after stroke, but 5-year mortality after a stroke was shown to be higher (78%) among patients with AF than among those in sinus rhythm (52%).14
Although models assessing the cost-effectiveness of anticoagulation in AF patients use outcome data from AF populations in clinical trials, it is unclear whether the AF-related strokes that occur in these trials are typical (in terms of frequency and severity) of those that occur in real-life clinical practice (eg, anticoagulant intensity may be less well-controlled in real life than in a trial setting; age and comorbidities among the populations may be different). Stroke occurring in patients at subtherapeutic International Normalized Ratio values are more severe than those in patients with a therapeutic intensity of anticoagulation.54 Epidemiological studies may be more informative on these issues than clinical trials. Some studies have used the Framingham55 or other population data for calculation of the absolute risk of stroke,42,46 but Thomson et al46 acknowledge that the risk was derived from data from the entire Framingham cohort, not just those with AF, and that regression equations tend to underestimate risk in those at the higher end of the range. Hence, the absolute risk of stroke in patients with AF may be underestimated in their model.
Among the practical limitations of warfarin is its variable anticoagulant effect. The fear of causing bleeding is one reason why anticoagulation is underused, particularly in the most elderly, despite the fact that they generally have the highest risk of stroke. Some investigators have found a higher incidence of bleeding in elderly patients receiving warfarin,56 whereas others suggest that more studies are needed that better-estimate the risk of intracranial bleeding with and without anticoagulation.47 Most clinical trials of anticoagulation do not include this age group. Therefore, an important part of the research agenda should be to understand the benefit/risk relationship, ideally in a study that includes an economic analysis. If the effectiveness and safety of anticoagulation (whether in usual care or in an anticoagulation clinic setting) can be demonstrated in this age group, then the economic gain could be further emphasized.
The extent of potential underestimation of the unit cost of stroke needs to be explored further, and our key recommendation for new research is to conduct a study to collect long-term costs of stroke in patients with AF (compared with non-AF patients). This would permit new economic analyses on existing raw data from published trials using more specific (presumably higher) cost-of-stroke data. If also applied to new clinical efficacy studies (such as among the elderly or people at lower risk of stroke), it would inform about cost-effectiveness in different groups of patients. This research is urgent because underestimation of the value of anticoagulation will affect clinical practice.
Footnotes
Paul Miller and Fredrik Andersson are employed by AstraZeneca.
References
Hjelmgren J, Berggren F, Andersson F. Health economic guidelines— similarities, differences and some implications. Value Health. 2001; 4: 225–250.
Sullivan SD, Lyles A, Luce B, Grigar J. AMCP Guidance for submission of clinical and economic evaluation data to support formulary listing in US health plans and pharmacy benefits management organisations. J Manag Care Pharm. 2001; 7: 272–282.
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.
Narayan SM, Cain ME, Smith JM. Atrial fibrillation. Lancet. 1997; 350: 943–950.
Ryder KM and Benjamin EJ. Epidemiology and significance of atrial fibrillation. Am J Cardiol. 1999; 84: 131R–138R.
Morgan S, and Mant D. Randomised trial of two approaches to screening for atrial fibrillation in UK general practice. Br J Gen Pract. 2002; 52: 373–380.
Stewart S, Murphy N, Walker A, McGuire A, McMurray J. Cost of an emerging epidemic: an economic analysis of atrial fibrillation in the UK. Heart. 2004; 90: 286–292.
Steinberg J. Atrial fibrillation: an emerging epidemic Heart. 2004; 90: 239–240.
Go AS, Hylek EM, Phillips KA, Chang Y, Henault LE, Selby JV, Singer DE. Prevalence of diagnosed atrial fibrillation in adults: national implications for rhythm management and stroke prevention: the AnTicoagulation and Risk Factors in Atrial Fibrillation (ATRIA) Study. JAMA. 2001; 285: 2370–2375.
Lamassa M, Di Carlo A, Pracucci G, Basile AM, Trefoloni G, Vanni P, Spolveri S, Baruffi MC, Landini G, Ghetti A, Wolfe CD, Inzitari D. Characteristics, outcome, and care of stroke associated with atrial fibrillation in Europe: data from a multicenter multinational hospital-based registry (The European Community Stroke Project). Stroke. 2001; 32: 392–398.
Wolf PA, Kannel WB, McGee DL, Meeks SL, Bharucha NE, McNamara PM. Duration of atrial fibrillation and imminence of stroke: the Framingham study. Stroke. 1983; 14: 664–667.
Britton M, Gustafsson C. Non-rheumatic atrial fibrillation as a risk factor for stroke. Stroke. 1985; 16: 182–188.
Candelise L, Pinardi G, Morabito A. Mortality in acute stroke with atrial fibrillation. The Italian acute stroke study group. Stroke. 1991; 22: 169–174.
Gustafsson C, Britton M. Pathogenetic mechanism of stroke in non-valvular atrial fibrillation: follow-up of stroke patients with and without atrial fibrillation. J Intern Med. 1991; 230: 11–16.
Broderick JP, Phillips SJ, Ofallen WM, Frye RL, Whisnant JP. Relationship of cardiac disease to stroke occurrence, recurrence, and mortality. Stroke. 1992; 23: 1250–1256.
Sandercock P, Bamford J, Dennis M, Burn J, Slattery J, Jones L, Booyakarnkul S, Warlow C. Atrial fibrillation and stroke: prevalence in different types of stroke and influence on early and long term prognosis (Oxfordshire community stroke project). BMJ. 1992; 305: 1460–1465.
Censori B, Camerlingo M, Casto L, Ferraro B, Gazzaniga GC, Cesana B, Mamoli A. Prognostic factors in first-ever stroke in the carotid artery territory seen within 6 hours after onset. Stroke. 1993; 24: 532–535.
Anderson CS, Jamrozik KD, Broadhurst RJ, Stewart-Wynne EG. Predicting survival for 1 year among different subtypes of stroke. Results from the Perth community Stroke Study. Stroke. 1994; 25: 1935–1944.
Lin HJ, Wolf PA, Kelly-Hayes M, Beiser AS, Kase CS, Benjamin EJ, D’Agostino RB. Stroke severity in atrial fibrillation. The Framingham Study. Stroke. 1996; 27: 1760–1764.
Jorgensen HS, Nakayama H, Reith J, Raaschou HO, Olsen TS. Acute stroke with atrial fibrillation. The Copenhagen Stroke Study. Stroke. 1996; 27: 1765–1769.
Appelros P, Nydevik I, Seiger A, Terent A. Predictors of severe stroke. Influence of pre-existing dementia and cardiac disorders. Stroke. 2002; 33: 2357–2362.
Appelros P, Nydevik I, Viitanen M. Poor outcome after first-ever stroke. Predictors for death, dependency, and recurrent stroke within the first year. Stroke. 2003; 34: 122–126.
Dulli DA, Stanko H, Levine RL. Atrial fibrillation is associated with severe acute ischemic stroke. Neuroepidemiology. 2003; 22: 118–123.
Taylor TN, Davis PH, Torner JC, Holmes J, Meyer JW, Jacobson MF. Lifetime cost of stroke in the United States. Stroke. 1996; 27: 1459–1466.
Fagan SC, Morgenstern LB, Petitta A, Ward RE, Tilley BC, Marler JR, Levine SR, Broderick JP, Kwiatkowski TG, Frankel M, Brott TG, Walker MD. Cost-effectiveness of tiss ue plasminogen activator for acute ischemic stroke. NINDS rt-PA Stroke Study Group. Neurology. 1998; 50: 883–890.
Dewey HM, Thrift AG, Mihalopoulos C, Carter R, Macdonell RA, McNeil JJ, Donnan GA. Lifetime cost of stroke subtypes in Australia: findings from the North East Melbourne Stroke Incidence Study (NEMESIS). Stroke. 2003; 34: 2502–2507.
Bergman L, van der Meulen JH, Limburg M, Habbema JD. Costs of medical care after first-ever stroke in The Netherlands. Stroke. 1995; 26: 1830–1836.
Persson U, Silverberg R, Lindgren B, Norrving B, Jadback G, Johansson B, Puranen BI. Direct costs of stroke for a Swedish population. Int J Technol Assess Health Care. 1990; 6: 125–137.
Terént A, Marké LA, Asplund K, Norrving B, Jonsson E, Wester PO. Costs of stroke in Sweden: a national perspective. Stroke. 1994; 25: 2363–2369.
Zethraeus N, Molin T, Henriksson P, Jonsson B. Costs of coronary heart disease and stroke: the case of Sweden. J Intern Med. 1999; 246: 151–159.
Levy E, Gabriel S, Dinet J. The comparati ve medical costs of atherothrombotic disease in European countries. Pharmacoeconomics. 2003; 21: 651–659.
Ghatnekar O, Persson U, Glader EL, Terént A. The cost of stroke in Sweden—an incidence estimate. Int J Technol Assessment Health Care. 2004; 20: 375–380.
Lightowlers S, McGuire A. Cost-effectiveness of anticoagulation in non-rheumatic atrial fibrillation in the primary prevention of ischemic stroke. Stroke. 1998; 29: 1827–1832.
Caro JJ, Huybrechts KF for the Stroke Economic Analysis Group. Stroke treatment economic model (STEM): predicting long-term costs from functional status. Stroke. 1999; 30: 2574–2579.
Youman P, Wilson K, Harraf F, Kalra L. The economic burden of stroke in the United Kingdom. Pharmacoeconomics. 2003; 21: 43–50.
Weimar C, Weber C, Wagner M, Busse O, Haberl RL, Lauterbach KW, Diener HC. German Stroke Data Bank Collaborators. Management patterns and health care use after intracerebral hemorrhage. A cost-of-illness study from a societal perspective in Germany. Cerebrovasc Dis. 2003; 15: 29–36.
Spieler JF, De Pouvourville G, Amarenco P. Cost of a recurrent vs cost of first-ever stroke over an 18-month period. Eur J Neurol. 2003; 10: 621–624.
Carod-Artal FJ, Egido-Navarro JA, Gonzalez-Gutierrez JL, Varela de Seijas E. Direct cost of cerebrovascular disease during the first year of follow-up. Rev Neurol. 1999; 29: 1353–1355.
US Department of Labor, Bureau of Labor Statistics, Consumer Price Indexes. Inflation calculator. http://www.bls.gov/cpi/home.htm. Accessed April 2004.
Organisation for Economic Co-operation and Development. Purchasing Power Parities (PPP), statistics, data and indicators: PPPs and derived indices for all OECD countries, April 2004. Available at http://www.oecd.org/statisticsdata/0,2643,en_2649_34357_1_119656_1_1_1,00.html. Accessed April 2004.
Wolf PA, Mitchell JB, Baker CS, Kannel WB, D’Agostino RB. Impact of atrial fibrillation on mortality, stroke, and medical costs. Arch Intern Med. 1998; 158: 229–234.
Gustafsson C, Asplund K, Britton M, Norrving B, Olsson B, Marke I. Cost effectiveness of primary stroke prevention in atrial fibrillation: Swedish national perspective. BMJ. 1992; 305: 1457–1460.
Eckman MH, Levine HJ, Pauker SG. Decision analytic and cost-effectiveness issues concerning anticoagulant prophylaxis in heart disease. Chest. 1992; 102: 538s–549s.
Gage BF, Cardinalli AB, Albers GW, Owens DK. Cost-effectiveness of warfarin and aspirin for prophylaxis of stroke in patients with nonvalvular atrial fibrillation. JAMA. 1995; 274: 1839–1845.
Gage BF, Cardinalli AB, Owens DK. Cost-effectiveness of preference-based antithrombotic therapy for patients with nonvalvular atrial fibrillation. Stroke. 1998; 29: 1083–1091.
Thomson R, Parkin D, Eccles M, Sudlow M, Robinson A. Decision analysis and guidelines for anticoagulant therapy to prevent stroke in patients with atrial fibrillation. Lancet. 2000; 355: 956–962.
Desbiens N. Deciding on anticoagulating the oldest old with atrial fibrillation: insights from cost-effectiveness analysis. J Am Geriatr Soc. 2002; 50: 863–869.
Samsa GP, Matchar DB, Phillips DL, McGrann J. Which approach to anticoagulation management is best Illustration of an interactive mathematical model to support informed decision making. J Thromb Thrombolys. 2002; 14: 103–111.
Grieve R, Hutton J, Bhalla A, Rastenyte D, Ryglewicz D, Sarti C, Lamassa M, Giroud M, Dundas R, Wolfe CD. A comparison of the costs and survival of hospital-admitted stroke patients across Europe. Stroke. 2001; 32: 1684–1691.
Gold MR, Siegel JE, Russell LB, Weinstein M. Cost-effectiveness in Health and Medicine. New York: Oxford University Press; 1996.
Chang KC, Su IJ, Tsai ST, Shieh DB, Jin YT. Prediction of length of stay of first-ever ischemic stroke. Stroke. 2002; 33: 2670–2674.
Porsdal V, and Boysen G. Direct costs during the first year after intracerebral haemorrhage. Eur J Neurol. 1999; 6: 449–454.
Claesson L, Gosman-Hedstrom G, Johanesson M, Fagerberg B, Blomstrand C. Resource utilization and costs of stroke unit care integrated in a care continuum: a 1-year controlled, prospective, randomized study in elderly patients. The Goteborg 70+ Stroke Study. Stroke. 2000; 31: 2569–2577.
Hylek EM, Go AS, Chang Y, Jensvold NG, Henault LE, Selby JV, Singer DE. Effect of intensity of oral anticoagulation on stroke severity and mortality in atrial fibrillation. N Engl J Med. 2003; 349: 1019–1026.
Wolf PA, D’Agostino RB, Belanger AJ, Kannel WB. Probability of stroke: a risk profile from the Framingham Study. Stroke. 1991; 22: 312–318.
Pengo V, Legnani C, Noventa F, Palareti G. ISCOAT Study Group. (Italian Study on Complications of Oral Anticoagulant Therapy). Oral anticoagulant therapy in patients wi th nonrheumatic atrial fibrillation and risk of bleeding. A Multicenter Inception Cohort Study. Thromb Haemost. 2001; 85: 418–422.