| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
(Stroke. 2001;32:1684.)
© 2001 American Heart Association, Inc.
Original Contributions |
A Comparison of the Costs and Survival of Hospital-Admitted Stroke Patients Across Europe
From the Department of Public Health Sciences, GKT School of Medicine, London, UK (R.G., J.H., A.B., R.D., C.D.A.W.); Institute of Cardiology, Medical Academy, Kaunas, Lithuania (D. Rastenytë); Institute of Psychiatry and Neurology, Warsaw, Poland (D. Ryglewicz); National Institute of Public Health, Helsinki, Finland (C.S.); Department of Neurology and Psychiatry, Careggi University, Florence, Italy (M.L.); and Department of Neurology, University Hospital, Dijon, France (M.G.).
Correspondence to Richard Grieve, Health Services Research Unit, London School of Hygiene and Tropical Medicine, Keppel Street, London WC1E 7HT, UK. E-mail richard.grieve{at}lshtm.ac.uk
 |
Abstract |
|---|
 Top Abstract IntroductionMethodsResultsDiscussionReferences |
|---|
Background and Purpose—Policy makers require evidence on the costs and outcomes of different ways of organizing stroke care. This study compared the costs and survival of different ways of providing stroke care.
Methods—Hospitalized stroke patients from 13 European centers were included, with demographic, case-mix, and resource use variables measured for each patient. Unit costs were collected and converted into US dollars using the purchasing power parity (PPP) index. Cox and linear regression analyses were used to compare survival and costs between the centers adjusting for case mix.
Results—A total of 1847 patients were included in the study. After case-mix adjustment, the mean predicted costs ranged from $466 [95% CI 181 to 751] in Riga (Latvia) to $8512 [7696 to 9328] in Copenhagen (Denmark), which reflected differences in unit costs, and resource use. The mean length of hospitalization ranged from 8.3 days in Menorca (Spain) to 36.8 days in Turku B (Finland). In the 3 Finnish centers at least 80% of patients were admitted to wards providing organized stroke care, which was not provided at the centers in Almada (Portugal), Menorca, or Riga. Patients in Turku A and Turku B were less likely to die than those in Riga, Warsaw (Poland), or Menorca. The adjusted hazard ratios were 0.18 [0.10 to 0.32] for Turku A, 0.18 [0.10 to 0.32] for Turku B, 0.68 [0.48 to 0.96] for Warsaw, and 0.56 [0.33 to 0.96] for Menorca, all compared with Riga.
Conclusions—The cost of stroke care varies across Europe because of differences in unit costs, and resource use. Further research is needed to assess which ways of organizing stroke care are the most cost-effective.
Key Words: costs and cost analysis • stroke management • survival
|
Introduction |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
To allocate healthcare resources in an efficient way, policy makers require information on the costs and outcomes of competing healthcare interventions.1 However, published information on the costs of health services is often not available or is of poor quality.2 To address this problem, some Western European countries (Finland, the Netherlands, Portugal, Denmark, and France) have moved toward the statutory use of economic evaluation in decision-making which requires information on costs and outcomes to be presented to assess whether particular interventions should be publicly funded.3
This issue is pertinent for stroke care where, apart from aspirin for ischemic stroke, no acute pharmaceutical intervention has been shown to be effective for the majority of patients,4 and yet high costs are incurred in managing the disease.5 There is considerable variation in the process of stroke care and associated outcomes across Europe,6 7 8 but little is known about the costs associated with the different ways of delivering care. A recent review9 found that previous studies have generally assessed the costs of care using routine data sets, which were not sufficiently detailed to present information on the resource use, total costs, and survival associated with different models of stroke care.
This study uses a detailed costing methodology to compare the resource use, total costs, and survival associated with providing stroke care in different European settings.
|
Methods |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The settings and methods used in this study have been previously described.6 10 Briefly, 13 centers in 10 different European countries were included in the study (Almada, Portugal; Menorca, Spain; Florence, Italy; Dijon, France; Copenhagen, Denmark; Kuopio, Finland; Turku [A and B], Finland; London, United Kingdom; Warsaw, Poland; Kaunas [A and B], Lithuania; and Riga, Latvia). The participating hospitals were not intended to be representative of the stroke care in the particular country but to give examples of the different ways in which stroke care may be provided across Europe. All the centers took patients as direct admissions, did not use a triage system for selecting cases, and provided routine acute care to the local community. These particular centers were selected because they had the necessary research resources to collect the data required by this study.
Inclusion and Exclusion Criteria
All patients admitted to hospital with a first-ever
stroke, defined by the WHO
criteria,11 were
prospectively recruited to the study. Patients who had a
subarachnoid hemorrhage were excluded, as were 31
patients who refused consent to participate in the study. The
characteristics of the participants were similar to those of the
nonparticipants.12 Patients
were excluded from the main analysis if they were missing the
case-mix data needed for the regression
analyses.
Data Collection
Hospital-based stroke registers were established for
1 year during 1996 to 1997 (9 centers) and 1998 (3 Finnish centers).
For the Copenhagen center, patients were recruited for 2 years from the
Hvidorve Hospital Stroke Database (1994 to 1995). Designated
investigators at each center completed standard forms to record
baseline information.8 Data
were collected on patient characteristics (sex, age, prestroke living
conditions), stroke severity measures (level of consciousness at the
time of maximum impairment [subsequently dichotomized into coma or
noncoma], continence during the first week after stroke, dysphasia,
paralysis at hospital admission), stroke subtype (cerebral infarction,
intracerebral hemorrhage, or unspecified
stroke), and risk factors (prestroke hypertension, previous myocardial
infarction).
Resource Use
The use of hospital and community services was
recorded for 3 months after stroke. The use of
diagnostic investigations was recorded for each patient
from medical records. The length of hospital stay, by ward type
(intensive care unit, neurology unit, dedicated stroke-specific unit,
or general medical ward), was also recorded for each patient. The
total length of stay was calculated by summing the length of stay in
each hospital. Each ward type was categorized according to whether care
provided met the definition for organized stroke care. In particular,
the level of training and education for different members of staff, the
extent to which multidisciplinary teamwork was practiced, and whether
care was led by a clinician with a particular interest in stroke were
recorded during interviews with staff at each
center.13
One investigator (R.G.) visited the centers and used a semistructured questionnaire to record the available staff time (doctors, nurses, therapists) per occupied bed–day on the wards where stroke patients were treated. The total staff time for each patient’s hospital stay was calculated by multiplying the length of total hospital stay by the staff time per occupied bed–day.
The use of outpatient and community services (hospital clinics, therapy, primary care physician visits, home care, nursing and residential homes), was recorded from interviews with the patients and their carers at 3 months after stroke.
Cost Measurement
A standard
method,12 which incorporated
general guidelines on measuring costs in economic
evaluation,13 was used to
collect unit costs at hospital and community service providers in all
the centers. Each patient’s use of resources was multiplied by the
local unit costs to give a total cost per patient.
To measure the cost of hospital staff, the costs per hour of employing each grade of staff working on the relevant wards were required. These costs were calculated by dividing the total annual cost of employing the relevant midgrade health professional (which included the salary, overtime, and employers’ National Insurance contributions) by the total number of hours worked. The costs per hour were then multiplied by the average available time per occupied bed–day for each category of staff on each ward. Costs were then summed across the relevant grades and categories of staff (doctors, nurses, and therapists) to give a total staffing cost per occupied bed–day for each hospital ward. This method allowed the costs per occupied bed–day to reflect the different intensities of staffing input available on the various wards (eg, the staffing costs of intensive care units were higher than those of general medical wards). The costs per occupied bed–day of consumables, drugs, and overheads were calculated by dividing the total annual expenditure in each department by the total number of occupied bed–days. These costs were added to the staffing costs to give a total cost per occupied bed–day (or per diem cost) for the relevant hospital wards at each center.
The unit cost of investigations was taken from the price charged by the relevant department to another healthcare provider. The investigation costs were calculated by multiplying each patient’s use of investigations by the unit cost. The total hospitalization costs were calculated by multiplying the length of stay on each ward by the appropriate cost per occupied bed–day and then adding to this the total cost of investigations.
For outpatient, community, and institutional care, interviews were undertaken with a range of providers at each center to derive the average cost of care. The median cost of the item concerned was used for the unit cost, which was then multiplied by the level of service use to give the total cost of the services provided.
Total costs per patient were the sum of the hospitalization, outpatient, community, and institutional care costs. All costs were adjusted to a 1998 price base by using price indices.14 15 To adjust for price differences between the respective economies, costs were converted from local currencies into US dollars by using the purchasing power parity (PPP) index.14 15
Outcome Measurement
The primary outcome measure was survival following
stroke. Information on whether the patient died and the date of death
were obtained from death certificates, medical records, and
national registers. Survival time was calculated by subtracting the
date of stroke from the date of death.
Functional status was measured using the Barthel Index16 (recorded on a scale from 0 to 20), assessed by face-to-face or telephone interview at 3 months after stroke. The scale was dichotomized into functionally independent (Barthel Index score of 20) or functionally dependent (Barthel score <20).
Statistical and Sensitivity
Analysis
Differences between the centers in case mix were
tested using the 
2 test for categorical
and the Kruskal-Wallis test for continuous variables. Multiple
linear regression was used to compare the total costs of care across
the centers adjusting for case mix, and the coefficients were used to
predict the mean adjusted total costs in each center. In the
sensitivity analysis, the impact of between-center differences
in unit costs was examined by using the unit costs from the center with
the median per diem costs (Dijon) to calculate total costs (measured in
US$/PPP).
Cox regression analysis compared survival between the centers adjusting for case-mix differences (age, sex, stroke subtype, level of consciousness, incontinence, dysphasia, and paralysis). Interaction effects were not significant and were excluded from the final model. In the sensitivity analysis, a Cox regression model compared survival between the centers, adjusting for age, sex, and stroke subtype. This model was run on the analysis sample and the full patient sample to investigate the effect of excluding the cases with missing data. Finally, various Cox regression models were formulated that included risk factors (previous myocardial infarction, prestroke hypertension) and additional patient characteristics (prestroke living conditions) to test the sensitivity of the results to the choice of variables used in the case-mix adjustment.
|
Results |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
A total of 1981 patients were registered, of whom 134 were missing key case-mix variables and were excluded from the main analysis. The patient and case-mix characteristics of the 1847 patients included are shown in Table 1
. There were significant differences
between the centers in patient characteristics, stroke severity, stroke
subtype, and risk factors
(P<0.001 for all
variables).
|
The different models of care that existed across the centers
were exemplified by the within-hospital resource use variation
(Table 2). The mean total duration of hospitalization
ranged from 8.3 days in Menorca to 36.8 days in Turku B.
|
There were also differences in the organization of care. The centers in Almada and Menorca provided care mainly on general medical wards, which did not meet the criteria for organized stroke care. In the other Western European centers, organized stroke care was provided on neurology wards (Dijon, Kuopio, and Turku A), dedicated acute stroke units (Florence and Copenhagen), general medical wards (Turku B), elderly care wards, and rehabilitation stroke units (London), and in rehabilitation hospitals (Florence, Dijon, Kuopio, and Turku A and B).
The majority of patients in the Eastern European centers did not receive organized stroke care, apart from those admitted to the stroke unit in Warsaw (51%) or those transferred to the rehabilitation hospitals in Kaunas (33% in Kaunas A, 23% in Kaunas B).
Differences in staff input
(Table 2
) illustrated the variation in the intensity of care
provided. For example, the input from doctors at the Copenhagen center
was higher than in London over a similar length of stay. In addition,
the type of staff input varied between the centers. For example,
nursing input at the stroke unit in Florence was provided entirely by
fully qualified nurses whereas on the stroke unit in London, 40% of
the nurses had only received a basic level of training. The use of
investigations varied between the centers; the mean number of CT scans
was highest in the centers in Dijon and Turku A, where care was led by
neurologists.
In the Eastern European centers, patients were usually admitted to a neurology ward, where the inputs available from nurses and therapists were generally lower than in the neurological units in Western European.
A total of 90 cases were lost to follow-up at 3 months
(Table 3), a dropout rate of <8% in all the centers
except Riga. The remaining results are shown for the 1757 cases with
complete follow-up data. The mean use of residential and nursing homes
was very low in Eastern Europe and highest in the Turku B center, where
there was also a higher proportion of disabled stroke survivors.
Patients from the centers in Almada and Menorca generally had a level
of community service use similar to those in other Western European
centers, where patients stayed in hospital longer. The low
availability of community and outpatient services was noticeable in
Eastern Europe, and generally few patients had access to occupational
and speech therapists after discharge.
|
The costs of a hospital day
(Table 4) varied widely between Eastern and Western
Europe, partly reflecting differences in labor costs. For example, the
cost per hour of a specialist doctor was $10.01 in Warsaw compared with
$29.36 in Copenhagen after PPP adjustment. Cost differences within
Western Europe were more likely to reflect the relative intensity of
care provided: for example, the dedicated stroke unit in Florence had a
higher input from doctors and qualified nurses than the one in
London.
|
The differences in the mean total costs
(Figure) reflected the variation in the lengths of
hospital stay, with those centers with an average stay of >30 days
also having the highest costs. While the costs of hospitalization were
the largest cost component, in certain centers the costs of outpatient
care (Turku A, Almada, and Kuopio) and community care (Dijon and
Copenhagen) were also important.
|
In the baseline analysis, the centers with the
lowest mean predicted costs were Riga ($466), Kaunas A ($1517), Kaunas
B ($2241), and Menorca ($2262); the highest mean costs were in London
($7344), Turku A ($7808), and Copenhagen ($8512)
(Table 5). The extent to which costs varied between the
centers was sensitive to the measure of unit costs used. When the unit
costs of the Dijon center were applied to all the centers, the range of
mean predicted costs was much narrower. However, the centers were
generally placed in a similar order: Copenhagen, London, and Turku A
were still in the highest quartile of total costs; the only change was
that Kaunas B was replaced in the lowest quartile by
Kuopio.
|
Cox regression analysis showed that patients in Riga
were the most likely to die, followed by those in Warsaw, Menorca, and
London
(Table 6). The relative probability of patient survival
was highest at the centers in Turku, Florence, and Copenhagen. Because
Riga was the center with the highest loss to follow-up, the
analysis was repeated, with the assumption that all the
patients who were lost to follow-up in Riga were alive at 3 months.
This did not alter the relative position of the centers nor did
changing the reference case. The sensitivity analysis suggested
that the results were reasonably robust to the case-mix adjustment made
and the inclusion criteria chosen. Running the model on the same number
of cases but using only demographic and stroke subtype variables
for the case-mix adjustment meant that the center in Kuopio had the
highest relative probability of patient survival; the ordering of the
remaining centers was largely unchanged. Using the above variables
but including all the registered cases did not affect the ordering of
the centers in the results of the survival analysis. Adjusting
for risk factors and prestroke residence meant that the Copenhagen
center had the highest survival rate, while the ordering of the other
centers was largely unchanged.
|
|
Discussion |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
The aim of this study was to compare the costs and survival of different ways of providing stroke care across Europe. A detailed approach was taken to costing that enabled differences in the processes of care to be described. The study found that the mean total costs of care varied between the centers because of differences in the unit costs and the resources used. The rates of survival also varied between the centers, and our study suggests that to achieve the World Health Organization target of a 1-month case fatality rate of <20% by 2005,17 some centers need to deploy resources more effectively.
Methodological Critique
It should be borne in mind that this study is not a
cost-effectiveness analysis and that certain methodological
issues inherent in international comparisons had to be tackled as part
of the analysis. The levels of initial stroke severity and
stroke subtype varied between the centers, which may partly reflect the
differences in admission rates that exist across
Europe.18 19 20
In addition, the proportion of hospital patients with undetected
strokes may vary between hospitals, partly according to the
availability of routine CT or MRI
scans.21 These variations
made adjustment for case mix necessary when costs and outcomes were
compared between the centers. These case-mix measures were recommended
by Davenport et al22 and
have been used previously to compare outcomes after
stroke.8 The sensitivity
analysis examined the effect of changing the variables
included in the case-mix adjustment and found that that the results
were largely unchanged. Despite this, the observational nature of the
study means that unmeasured case-mix differences between the centers
may explain some of the residual differences in cost and outcome.
Although stroke subtype was specified as either cerebral infarction or
intracerebral hemorrhage, the Oxford Stroke
Project Classification has not been used to classify cerebral
infarction.23 If the
proportion of lacunar strokes (low case fatality rate) or total
anterior circulatory strokes (high case fatality rate) varied between
the centers, this may explain some of the observed differences in
mortality.23
In addition, no standard measure of social class was available across the countries. While the association of social class and poststroke outcome or cost remains unclear,24 the development of a pan-European measure would assist future studies in this area.
To compare costs across the centers, unit costs were converted from local currencies to US dollars. Rather than converting costs by using exchange rates, which are unlikely to reflect differences in the relative price of resources between countries, the conversion was made using the general PPP index. This index is recommended for comparing costs across countries, as it adjusts for differences in relative prices between economies.25 However, our analysis showed that even after adjusting for differences in relative prices, there were still large disparities in unit costs between countries, which might explain differences in total costs. The sensitivity analysis tested whether differences in unit costs explained total cost differences by applying unit costs from one center to all centers. The result was that using the same unit costs reduced the magnitude of the total cost differences between countries, but the order of the centers was largely unchanged and reflected differences in physical resource use.
The aim of the study was to provide specific information on the costs and outcomes of different ways of providing stroke care. To address this issue, centers were chosen across 10 European countries. The aim of the study was not to provide information on the national average cost of stroke care, which would have required many centers to be recruited in each country. The analysis illustrated that costs and survival time vary widely between patients within centers, which explains the wide confidence intervals around the regression coefficients. These confidence intervals overlapped between the individual centers, so the results of the regression analyses were presented with centers grouped in quartiles rather than according to their specific positions.
The centers in Riga, Menorca, and Kaunas had low total costs. Lengths of stay in the acute hospital were relatively short, with care being provided on general medical or neurology wards. Stroke care was characterized by low inputs from therapists and nurses, and apart from the rehabilitation hospitals in Kaunas, care provision did not meet the criteria for organized stroke care. The centers in Riga and Menorca were in the quartile of centers with the lowest survival rate, which may reflect the short hospitalizations and the lack of organized stroke care in these settings.
The centers in Warsaw, Almada, and Kuopio generally made more intensive use of resources, although the duration of hospitalization was still only moderate. The Kuopio center had organized stroke care available in both the acute and rehabilitation centers and had higher rates of survival than the other 2 centers, which had similar costs.
The centers in Turku, Copenhagen, Florence, and Dijon generally had higher total costs. This reflected higher unit costs but also longer total lengths of stay and a generally moderate to high use of health professionals’ time and investigations. In each of these settings, organized stroke care was available in the acute setting on a neurology unit (Dijon, Copenhagen, and Turku A), a dedicated stroke unit (Copenhagen and Florence) or a general medical ward (Turku B). During the acute phase, specialist doctors, mainly neurologists, provided a high input, which may enable more intensive patient monitoring and assist in the detection of poststroke complications.26 Organized stroke care was also generally available in rehabilitation hospitals. Each of these centers had moderate to high levels of nursing and therapy available in the acute and rehabilitation hospitals and were characterized by generally high rates of poststroke survival.
Any general conclusion that devoting more resources to stroke care improves outcomes should be tempered by the results from the London center, where relatively high cost stroke care was accompanied by high levels of poststroke mortality and dependence. While this study included only 1 UK center, a previous study8 reported higher mortality rates in 5 UK centers compared with other Western European centers. National surveys27 28 have found that stroke services in the UK are poorly organized. This suggests that spending more on stroke services may not necessarily improve outcomes. Instead, the issue to address is how existing resources may be deployed in a more cost-effective way.
In this study, patients in London were admitted to general medical wards, an elderly care unit, or a dedicated rehabilitation stroke unit. Care in the general medical wards would not meet the criteria for organized stroke care, and the literature suggests that providing this nonspecialized stroke care may lead to worse outcomes.29 General physicians or geriatricians rather than neurologists led care in all the settings. No consensus exists on the optimal specialty to manage stroke patients,30 and it may be of greater concern that the level of input available from all doctors was much lower in the United Kingdom than most other European centers. A UK survey of hospital consultants recently reported that they felt they had inadequate time available for stroke patients.27 An additional hypothesis posed by this study is that apart from the level of input, the timing of the input provided by different health professionals is important. In the UK center, for example, while a high average level of input from therapists over the patients’ hospitalization was provided, this was not necessarily available during the first 7 days after stroke. In particular, patients admitted to general medical wards had a low rate of access to therapists during the first few days after stroke. Further work is required to assess the optimal level of input from different health professionals at various time points after stroke.
To conclude, this study highlighted variations in costs and outcomes after stroke care between different European settings. The method used illustrated that these variations may partly reflect differences in the overall level of resources available and the way in which those resources are organized. Further research is now required to establish measures of social class and PPP, which can be used in international studies assessing the cost-effectiveness of different ways to organize services.
Appendix: Study Participants
Prof G. Boysen, Dr V.
Porsdal, Department of Neurology, Hvidorve Hospital,
Hvidorve, Denmark; Prof A. Czlonkowska, Dr
D. Ryglewicz, Institute of Psychiatry and Neurology, Warsaw,
Poland; Dr R. Beech, Center for
Health Planning and Management, Keele, UK;
Dr C. Mckevitt, Department of
Public Health Medicine, GKT School of Medicine, London, UK;
Dr A. Rudd, Department of Care
of the Elderly, Guy’s & St Thomas’ Hospital, London, UK;
Prof D. Enina, Dr I. Purina,
Latvian Neuroangiological Center, Riga, Latvia;
Dr D. Rastenytë, Institute of
Cardiology, Medical Academy, Kaunas, Lithuania;
Prof M. Giroud, Dr M. Menassa, Dr M.
Lemesle, Service de Neurologie, Center Hospitalier Regional
et Universitaire de Dijon, France; Prof K.
Kunze, Neurologischen Universitatsklinik, Hamburg-Eppendorf,
Germany; Dr J. Berger,
Institute of Mathematics and Computer Science in Medicine, University
Hospital Eppendorf, Hamburg, Germany; Prof
Z. Nagy, Dr C. Ovary, Dr Z. Voko, National Stroke Center,
Budapest, Hungary; Prof D. Inzitari, Dr A.
Di Carlo, Dr M. Lamassa, Dr P. Vanni, Dipartimento di
Scienze Neurologiche & Psichiatriche, Ospedale Careggi, Florence,
Italy; Dr S. Spolveri, Dr M.C.
Baruffi, Florence, Italy; Dr I.
Remédios, Dr J. Coisinha, Hospital Garcia de Orta, Almada,
Portugal; Dr J. Dias, Dr. J.
Catarino, Divisão de Epidemiologia, Direcção Geral de
Saude, Lisboa, Portugal; Dr M. Torrent, Dr
M.A. Llach, Dr. M.P. Juan, Area de Salud de Menorca-INSALUD,
Menorca, Spain; Prof C. Vutuc, Dr G.
Haidinger, Department of
Epidemiology, University of Vienna, Austria;
Prof G. Schnaberth, Dr C. Alf,
Neurological Hospital Rosenhuegel, Vienna, Austria;
Prof J. Tuomilehto, Dr C. Sarti, Dr V.
Moltchanov, National Institute of Public Health, Helsinki,
Finland; Dr P. Immonen-Räihä, A.
Mononen, Heart and Stroke Register, Turku Town Hospital,
Turku, Finland; and Prof J. Sivenius, M.
Kalinen, Kuopio University Hospital, Kuopio,
Finland.
|
Acknowledgments |
|---|
Supported in part by a grant from the European Commission. The authors would like to thank David Hughes, Kate Tilling, Roger Beech, and Jan van der Meulen for their comments on a previous draft of this paper, and the participants of the Biomed II stroke project (see Appendix) for their help throughout the project.
|
Footnotes |
|---|
A complete list of participants in the Biomed II European Study of Stroke Care appears in the Appendix.
Received November 23, 2000; revision received March 6, 2001; accepted March 7, 2001.
|
References |
|---|
|
TopAbstractIntroductionMethodsResultsDiscussionReferences |
|---|
- Maynard A, Kanavos P. Health Economics: an evolving paradigm. Health Econ. 2000;9:183–190.[Medline] [Order article via Infotrieve]
- Davies L, Coyle D, Drummond M, and the EC Network on the Methodology of Economic Appraisal of Health Technology. Soc Sci Med. 1994;38:1601–1607.
- Hutton J, Maynard A. A NICE challenge for health economics. Health Econ. 2000;9:89–93.[Medline] [Order article via Infotrieve]
- Gubitz G, Sandercock P. Acute ischaemic stroke. BMJ. 320: 692–696.
-
Taylor T, Davies P,
Torner J, Holmes J, Meyer J, Jacobson M. Lifetime costs of stroke in
the United States. Stroke. 1996;27:1459–1466.
[Abstract/Free Full Text] -
McKevitt C, Beech
R, Pound P, Rudd AG, Wolfe CDA. Putting stroke outcomes into context:
assessment of variations in the processes of care.
Eur J Public Health. 2000;10:120–126.
[Abstract/Free Full Text] -
Beech R, Ratcliffe
M, Tilling K, Wolfe CDA. Hospital services for stroke care: a European
perspective. Stroke. 1996;27:1958–1964.
[Abstract/Free Full Text] -
Wolfe CDA, Tilling
K, Beech R, Rudd AG. Variation in case fatality and dependency from
stroke in Western and Central Europe.
Stroke. 1999;30:350–356.
[Abstract/Free Full Text] - Porsdal V, Boysen G. Cost-of-illness studies of stroke. Cerebrovasc Dis. 1997;7:258–263.
-
Grieve R, Dundas
R, Beech R, Wolfe CDA. The development and use of a method to compare
the costs of acute stroke across Europe.
Age Ageing. 2001;30:67-72.
[Abstract/Free Full Text] - Hatano S. Experience from a multicentre stroke register: a preliminary report. Bull World Health Organ. 1976;54:541–553.[Medline] [Order article via Infotrieve]
- Grieve R, Porsdal V, Hutton J, Wolfe CDA. A comparison of the cost-effectiveness of stroke care in London and Copenhagen. Int J Technol Assess Health Care. 2000;16:2:1–12.
- Drummond MF, O’Brien BJ, Stoddart GL, Torrance G. Methods for the Economic Evaluation of Health Care Programmes. 2nd ed. Oxford, UK: Oxford University Press; 1997.
- World Bank. World Bank Indicators 2000. 4th ed. Washington, DC: World Bank; 2000.
- OECD. Main Economic Indicators. Paris, France: OECD; 1999.
- Mahoney F, Barthel D. Functional evaluation: the Barthel Index. Md State Med J. 1965;14:61–65.[Medline] [Order article via Infotrieve]
- World Health Organization. Pan European Consensus Meeting on Stroke Management. Helsingborg, Sweden: World Health Organization; 1995.
- Ricci S, Celani M, La Rosa F, Vitali R, Duca E, Ferraguzzi R, Paolotti M, Seppoloni D, Caputo N, Chiurulla C, et al. SEPIVAC: a community-based study of stroke incidence in Umbria, Italy. J Neurol Neurosurg Psychiatry. 1991;54:695–698.[Abstract]
-
Tuomilehto J,
Sarti C, Narva EV, Salmi K, Sivenius J, Kaarsalo E. The FINMONICA
Stroke Register: community-based stroke registration and
analysis of stroke incidence in Finland, 1983–1985.
Am J Epidemiol. 1992;135:1259–1270.
[Abstract/Free Full Text] -
Lesmesle M, Milan
C, Faivre J, Moreau T, Giroud M, Dumas R. Incidence trends of
ischemic attacks in a well-defined French population from 1985
through 1994. Stroke. 1999;30:371–377.
[Abstract/Free Full Text] - Ebrahim S, Harwood R. Stroke: Epidemiology, Evidence, and Clinical Practice. 2nd ed. Oxford, UK: Oxford University Press; 1999.
-
Davenport RJ,
Dennis MS, Warlow CP. Effect of correcting outcome data for case mix:
an example from stroke medicine.
BMJ. 1996;312:1503–1505.
[Abstract/Free Full Text] - Bamford J, Sandercock P, Dennis M, Burn J, Warlow C. Classification and natural history of clinically identifiable subtypes of cerebral infarction. Lancet. 1991;337:1521–1526.[Medline] [Order article via Infotrieve]
-
Howard G,
Anderson R, Johnson NJ, Sorlie P, Russell G, Howard VJ. Evaluation of
social status as a contributing factor to the stroke belt region of the
United States. Stroke. 1997;28:936–940.
[Abstract/Free Full Text] - Kavanos P, Mossialos E. International comparisons of health care expenditures: what we know and what we do not know. J Health Serv Res Policy. 1999;4:2 122–126.[Medline] [Order article via Infotrieve]
- Indredavik B, Bakke F, Slordahl SA, Rokseth R, Haheim LL. Treatment in a combined acute and rehabilitation stroke unit: which aspects are important? Stroke. 1999; 30: 917–923.
- Ebrahim S, Redfern J. Stroke Care: A Matter of Chance. A National Survey of Stroke Services. London, UK: Stroke Association; 1999.
- Rudd AG, Irwin P, Rutledge Z, Lowe D, Wade DT, Pearson M. The national sentinel audit: an old tool for raising standards of care. J Roy Coll Phys. 1999;33:460–464.
-
Stroke Unit
Trialists’ Collaboration. Collaborative systematic review of the
randomised trials of organised inpatient (stroke unit) care after
stroke. BMJ. 1997;314:1151–1159.
[Abstract/Free Full Text] - Stroke Unit Trialists’ Collaboration. Organised inpatient (stroke unit) care for stroke (Cochrane review). In: The Cochrane Library, Issue 1, 1999. Oxford, UK: Update Software.
This article has been cited by other articles:
 |
|
 |
|
  H. Markus Improving the outcome of stroke BMJ, August 25, 2007; 335(7616): 359 - 360. [Full Text] [PDF]
|
|
|
|
 |
|
 P. S.J. Miller, F. L. Andersson, and L. Kalra Are Cost Benefits of Anticoagulation for Stroke Prevention in Atrial Fibrillation Underestimated? Stroke, February 1, 2005; 36(2): 360 - 366. [Abstract] [Full Text] [PDF]
|
|
|
|
 |
|
 H Markus Variations in care and outcome in the first year after stroke: a Western and Central European perspective J. Neurol. Neurosurg. Psychiatry, December 1, 2004; 75(12): 1660 - 1661. [Full Text] [PDF]
|
|
|
|
|
|
M. L. Sacchetti Stroke Is Best Managed by Neurologists Stroke, June 1, 2004; 35(6): e149 - e150. [Full Text] [PDF]
|
|
|
|
|
|
Y. Yoneda, T. Uehara, H. Yamasaki, Y. Kita, M. Tabuchi, and E. Mori Hospital-Based Study of the Care and Cost of Acute Ischemic Stroke in Japan Stroke, March 1, 2003; 34(3): 718 - 724. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
J. Teng, N. E. Mayo, E. Latimer, J. Hanley, S. Wood-Dauphinee, R. Cote, and S. Scott Costs and Caregiver Consequences of Early Supported Discharge for Stroke Patients Stroke, February 1, 2003; 34(2): 528 - 536. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||