QJM Advance Access originally published online on May 6, 2005
QJM 2005 98(6):415-425; doi:10.1093/qjmed/hci065
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||
Cost-effectiveness of integrated stroke services
From the 1Institute for Medical Technology Assessment (iMTA), 2Department of Health Policy and Management (iBMG), and 3Department of Cardiology, Clinical Epidemiology & Thoraxcenter, Erasmus MC, Rotterdam, The Netherlands
Address correspondence to Dr N.J.A. van Exel, Erasmus MC, Institute for Medical Technology Assessment (iMTA), Office WL4-121, PO Box 1738, 3000 DR Rotterdam, The Netherlands. e-mail: n.vanexel{at}erasmusmc.nl
Received 24 August 2004 and in revised form 12 January 2005
 |
Summary |
|---|
 Top Summary IntroductionMethodsResultsDiscussionReferences |
|---|
Background: Randomized trials have shown that integrating services for acute stroke care may lead to organizational improvements, higher efficiency and better patient outcomes in the acute phase.
Aim: To compare the costs and effects of stroke services in an experimental group of patients compared to a group of patients receiving conventional care.
Design: Prospective non-randomized controlled trial.
Methods: We compared all consecutively hospitalized stroke patients in three experimental stroke service settings (Delft, Haarlem and Nijmegen, n = 411) with concurrent patients receiving conventional stroke care (n = 187) over 6 months follow-up. Main end-points were total costs per patient and total health-adjusted days per 100 patients as measured by the EuroQol-5D score during follow-up.
Results: Mean total costs per patient were 
16 000 (95%CI 14 670–16 930): 13 160 in Delft, 16 790 in Haarlem, 20 230 in Nijmegen, and 13 810 in the control regions. Early discharge in Delft saved about 2500 hospital costs per patient. General patient health in Delft was significantly better than in the control regions; Haarlem and Nijmegen showed no difference in health.
Discussion: Our study confirms the potential to improve stroke outcomes in a cost-effective way in Dutch settings. This was seen in the group of patients in Delft, a complete and relatively simple stroke service, but not in two other regions with more complex stroke services. Important factors are reduction of hospital days and, most likely, adequate multidisciplinary rehabilitation.
|
Introduction |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Stroke is a leading and growing cause of death, long-term disability and health care costs in Western societies.1–5 Awareness is rising of the importance of the provision of integrated long-term care to promote rehabilitation and to reduce costs.6–8 This is true in the Netherlands, where about 27 000 people (
0.2% of the population) suffer a stroke each year, and ageing of the population will cause this number to increase by 30% in 2015.9 Presently, one third of Dutch patients with a first-ever stroke die within 36 months, making stroke the third leading cause of death in the Netherlands.10 About 60% survive with moderate or severe handicaps.11 In 1999, stroke was responsible for 2.9% of total health care costs, and for 6.0% in the population aged 75 and over. Thus stroke ranked second on the list of most costly diseases for the elderly, after dementia.12 Stroke survivors may suffer physical health effects, as well as many psychological and social problems.2–4 Consequently, many different disciplines and types of organization are needed to provide appropriate health care to these patients, with complex coordination. Many patients do not receive the care they require, from the appropriate professional, at the time and place they need it. In particular, many patients stay in hospital without medical need, waiting for discharge to a nursing home or rehabilitation centre, or waiting for professional home support or home adaptations.13–15 The number of such ‘waiting days’ may be quite substantial. For the Netherlands, about 10 of the average 28 days of stay in hospital after stroke are without medical justification, leading to increased costs, and possibly affecting patients’ recovery.16 Given these human and organizational problems, it is clear why considerable attention is currently being paid in the literature as to how stroke patients may receive more effective and efficient care, especially within a better integrated care continuum.1,13,17–27 Some RCTs have shown that organizational integration of stroke care may lead to higher efficiency and better health outcomes in the acute phase of total stroke care (up to 4–6 weeks), but that differences tend to be negligible or not evident thereafter;15,20,21,26,28 in other studies effects were not measurable, due to weaknesses in study design or low sample size.13,15,24
The Evaluation of Dutch Integrated Stroke Service Experiments (EDISSE) study investigated the health and economic impact of integrated stroke services in depth with respect to costs, health effects, quality and organization of care.29 The present paper presents our findings on the cost-effectiveness of integrated stroke services in patients in an experimental setting, compared to a group of stroke patients receiving current standard care in the Netherlands.
|
Methods |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Our study design is a prospective non-randomized controlled trial, comparing all consecutively hospitalized stroke patients in experimental stroke settings versus concurrent patients receiving conventional stroke care in control settings. Recruitment of experimental patients was from September 1999 to May 2000, and one patient in a control setting was recruited for every four in the experimental setting. Follow-up was at 6 months after stroke onset.
Before the start of the study, an independent expert committee from the government funding agency selected three study regions out of 53 proposals from groups of interested Dutch health institutions, using selection criteria determined a priori.21 These criteria were formulated to select the most likely successful regions to enhance quality and integration of care, patient outcomes and efficiency in the whole care continuum. The proposal from Delft was selected, as it had a complete care continuum, although with a relatively simple structure (one hospital, one nursing home and one home care organization), and with case management by specialized stroke nurses as specific intervention on top of that. Haarlem had three participating hospitals that were not fully dedicated to the project. It was selected because of its specific interventions in the rehabilitation phase. Nijmegen was selected as another complete stroke service, but had a more complex setting: two hospitals, various nursing homes and with specific interventions in the chronic phase (specialized home care and out clinic consultation by the hospitals). The selected proposals received additional funding to enhance the development of experimental integrated services. All patients in these experimental settings were compared to similar concurrent patients from general hospitals in three other Dutch regions. The latter were selected from a group of 23 hospitals participating in a previous national study.22 Based on the data collected in this already completed study, three hospitals were selected as representing the average Dutch patient, receiving the average current level of Dutch stroke care.30 The criteria for this selection were: the average age of patients, duration of hospital stay, case-fatality, Barthel Index at discharge and destination after discharge. Next, an independent group of researchers measured health outcomes, care satisfaction, direct and indirect medical costs, and quality of care in the experimental group and in the controls. Internationally published questionnaires were applied.
Simultaneously, a parallel qualitative assessment was made. This component examined the less tangible enhancing characteristics and constraints in both the experimental and control setting. Given the potential for bias in a non-randomized study design, this qualitative study also provided information on the contextual variations between the different regions at the provider side, i.e. within the service availability, service structure and service delivery models applied.
Defining integrated stroke services
Methodologically, the assessment of integrated stroke services is the evaluation of a complex mixture of interventions at the organizational, professional and patient levels. It needs careful definition, or at least a full description to allow for a transparent (cost-) effectiveness analysis.13,15 Formally, Dutch stroke services are defined as a network of service providers working together in an organized way to proved adequate services in all stages of the follow-up of stroke patients. It includes a hospital stroke unit.21 An expert group made this broad definition more explicit, defining a core set and an optimal set of criteria.29 It emphasizes a setting integrating all relevant institutions: hospitals, nursing homes, rehabilitation centers, general practitioners and home care providers working together to provide multidisciplinary, coordinated care through organized patient transfers and protocols. This definition is in accordance with international views.28
The local implementation of the stroke services may vary considerably, however.25 The Delft integrated stroke service has a relatively straight-forward structure, and consisted of one hospital with a stroke unit and a major nursing home, specialized nurses in home care and a transmural stroke nurse responsible for patient transfers. The Nijmegen integrated services had a focus on post-acute care at home; this was more extensive and district nurses received specialized stroke care training. The Haarlem stroke services had a focus on nursing home and home rehabilitation. Thus, in spite of the formal selection procedure and the definition of a common core of services, we observed differences in the organization of integrated services between the three experimental regions. The general objective of the stroke services was however the same: to ensure that carers, both specialists and generalists, from all the different care organizations with the stroke service combine to provide the right care for the right patient, at the right time, in the right manner and in the most suitable location. However, the organizational solutions to achieve this aim varied greatly and covered complex areas such as patient logistics (regulating the flow of patients through the system, knowing when and how blockages occur and resolving them, using various instruments such as case management, discharge protocols, capacity constraints), information logistics (discharge, multidisciplinairy communication, patient files) and co-ordination mechanisms (guidelines, protocols, training, project management), but also the history of regional collaboration, organisational leadership and cultures, budget restraints, etc. Because of these variations, we decided to examine the cost-effectiveness of each regional set-up separately.
Cost-effectiveness analysis
A cost-effectiveness analysis addresses the value for money of an intervention in health care by examining the difference in total costs and health effects between the intervention (here experimental stroke services) and the current alternatives (here conventional stroke care). Our study adopts the health care perspective: effects are expressed in terms of health-related quality of life (HRQoL), and costs of conventional and stroke service care are taken into account, insofar as these fall within the health care sector. Moreover, we conducted an economic analysis; we aimed at using actual costs of resource use, rather than tariffs.
Health care costs of stroke were measured at the patient level, and were based on individual health care utilization and unit costs for health resources.31 Health care utilization in hospitals, nursing homes and rehabilitation centres was registered by inspecting medical patient files on in-patient days, assessment, medication and consultations of physicians and paramedical personnel. Information on out-patient care, such as GP visits, medication, home (paramedical) care, home adaptations and assistive devices was collected during patient interviews at 2 and 6 months after hospital admission. Full costs (including capital costs and overhead) of in-patient days, rehabilitation treatment, stroke units and stroke service coordination costs (both start-up and structural) were estimated in each facility in all six regions, using the direct allocation method.32 For paramedical care (i.e. physical, speech and occupational therapy, and social work) the costs per consultation were based on time registration and salary costs of the relevant paramedical disciplines. The costs of diagnostic procedures, medication, surgery and home care were based on current tariffs, as these services only represent a small proportion of total costs. The costs of home adaptations and assistive devices were based on common market prices, as published by providers of these services.
Health effects were measured in quality-adjusted life-years (QALYs), in which remaining life years (after stroke onset) are weighted for health-related quality of life (HRQoL). The use of QALYs in cost-effectiveness analyses is recommended by the US Panel on Cost-effectiveness in Health and Medicine.33 The net number of days adjusted for health-related quality of life in the stroke patients during the 6-month follow-up (i.e. the number of HRQoL days after stroke) was measured using the EuroQol-5D.34 This generic health status instrument describes patients’ general state of health. It is a valid and efficient instrument for measuring HRQoL after stroke.35,36 It distinguishes five dimensions of health: mobility, self-care, daily activities, pain/complaints, and anxiety/depression, each with three levels (corresponding to no problems, mild problems and severe problems in the relevant dimensions). Combining these dimensions and levels, the EuroQol-5D distinguishes 243 possible health states. We used available valuations of these health states from the general public.37 One day lived in the best imaginable state of health is equivalent to one HRQoL day; death is equivalent to zero HRQoL days. Some states are regarded as worse than death by the general public; accordingly, these have a negative value. The EuroQol-5D scale was administered during patient interviews at 2 and 6 months after stroke. Missing values for patients unable to report their HRQoL and HRQoL values at hospital admission and discharge were based on a direct relationship between the Barthel Index (BI)38 and the EuroQoL-5D.39 BI was administered at hospital admission and discharge, and 2 or 6 months after stroke. HRQoL values at admission to hospital up to the end of the follow-up (or otherwise until death) describe a HRQoL curve for a patient in the 6 months following a stroke. The area under this curve is the equivalent of the number of HRQoL days a patient has lived during the follow-up period.
Our cost-effectiveness analysis computes health effects (aggregate HRQoL days after stroke onset) and costs between the groups of patients in the experimental regions and current standard care for stroke patients (i.e. the combined control regions). The comparison of the difference in costs and health effects is presented in a cost-effectiveness plane, with difference in costs on the vertical axis and difference in health effects on the horizontal axis.40 Where an intervention appears at the top left quadrant in the plane, higher costs are paid for lower effectiveness; the intervention is then unacceptable from a cost-effectiveness perspective and conventional care remains the treatment of choice. If the intervention appears in the lower right quadrant, lower costs are associated with positive health effects; the intervention is then acceptable. In the other two quadrants, higher (or lower) cost levels have to be weighted against greater (or lesser) effectiveness. To answer the question whether the health effects are worth the money, a decision needs to be made on the cost-effectiveness threshold (CE threshold) the intervention has to meet. There are however no general guidelines for CE thresholds. Different countries tend to accept different cut-off values: 13 000–65 000 per QALY for Canada; 25 000 per QALY for Australia, £30 000–40 000 per QALY for the UK.41 Therefore, as an example, we used a CE threshold of 35 000 per QALY. In other words, the stroke service experiment is acceptable only if it costs less (saves more) than 35 000 per quality-adjusted life-year saved (lost).
Finally, in a prospective non-randomized controlled studies, it is necessary to account for measurable differences between the (sub-)groups of patients.1,42,43 Given sufficient sample size in each region, we expected no differences between patient groups in the experimental and control settings. We used the Barthel Index as a standard to assess differences in case-mix between subgroups. We measured no statistically significant differences (p<0.01) at baseline upon hospital admission, either with regard to mortality and stroke severity, or in age, gender, education level and pre-stroke residential status (single, non-single).29 Hence, it was not necessary to standardize HRQoL results between regions.
|
Results |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Study population
The study included 598 stroke patients followed-up after hospital admission: 411 receiving care through integrated stroke services and 187 receiving conventional care in the control areas. Slightly more (54%) were female, mean age was 73.5 years (56% were aged
75 years), and one third of the patients were single before their stroke. On average, women were somewhat older and more often single. As there are substantial regional differences in outcomes, Table 1 presents the experimental patients by region.
|
The study followed up all patients through the health care system and at home. We observed substantial differences in length of hospital stay. Medical and non-medical reasons of hospital stay were recorded for all patients and all days of stay in hospital.16 The Delft region showed the lowest average length of stay, with 82% of hospital days for medical reasons. This percentage was only 43% in Haarlem, 46% in Nijmegen and 49% in the control regions. The main reasons for patients occupying a hospital bed without medical reason were waiting lists for nursing homes and professional home care.29 Early discharge saved Delft about
2500 hospital costs per patient.31 On average, 18% of the stroke patients died in hospital, 40% was discharged to home, 9% to a rehabilitation center, and 31% to a nursing home. At the end of the follow-up period, 177 (30%) patients had died, 10 (2%) patients were still in hospital, and over 50% had returned home (Table 2).
|
Effectiveness
Differences in health status after 2 and after 6 months were calculated. The integrated stroke service in the Delft region yielded health effects that were significantly better than in the control regions. The experiments in the Haarlem and Nijmegen regions showed no differences with the control group (Table 3). Health changes in HRQoL days were small in the period between 2 and 6 months after stroke. This confirms other studies that the acute phase is the most important period for recovery and that failure to provide appropriate rehabilitating care from the appropriate professional, during the first period after stroke onset, may affect patient recovery.15,23,28
|
The number of HRQoL days during the 6-month follow-up were calculated for all patients included in the study and aggregated per region. Because the number of patients differed between regions, health effects are presented for 100 patients (Table 3). Patients in Delft showed the highest number of HRQoL days, and integrated stroke services in Delft were significantly better compared to conventional stroke care in the control group; in Haarlem patients were significantly worse off; in Nijmegen patients showed no significant differences. Stroke survivors had more HRQoL days than patients who died; the number of HRQoL days decreased according to the severity of initial post-stroke handicap.
Costs
The average total costs of care per patient for the 6-month follow-up were estimated at 16 000 (95%CI 14 670–16 930).31 Costs were dominated by institutional and accommodation costs, and consisted of 43% hospital costs, 32% nursing home costs, 13% rehabilitation center costs, and 13% extramural costs. Patients who died after stroke incurred lower costs. For patients who survived the acute phase, the most important determinants of cost were age, disability as measured by BI, and residential status. These factors influenced patients’ institutional stroke careers (i.e. the likelihood of being and remaining institutionalized), and interacted.
Substantial differences in costs were observed between the three experimental settings and between the experimental and control settings. Mean total costs per patient were 13 160 in Delft, 16 790 in Haarlem, 20 230 Nijmegen, and 13 810 in the control regions. The costs of stroke service care in Delft and Haarlem did not differ significantly from conventional stroke care. Nijmegen was significantly more expensive than other settings (p<0.001).
Cost-effectiveness
Our cost-effectiveness analysis includes the empirical data and their uncertainty ranges. We present the outcomes by experimental region, compared to the outcomes from the control group receiving usual care. We computed the uncertainty ranges of differences in medical costs and QALYs, and determined the range of reliability for the cost-effectiveness ratio. In Figure 1, these reliability ranges are presented as ellipses in the cost-effectiveness plane for the significance levels of 5%, 50% and 95%. The ranges indicate the cost-effectiveness of integrated stroke services for each region as compared to conventional stroke care in the Netherlands.
|
Integrated stroke services in Delft showed higher health gain at comparable costs per patient (ICER was
19 350 less per QALY gained). The reliability range in Figure 1 shows a likelihood of ±80% that integrated stroke service in Delft was cost-effective. Here, the 95% reliability range indicates that the experiment has led to better health effects. The uncertainty ellipse lies mainly in the two right-hand quadrants (those representing positive health effects). The ellipse lies largely in the two lower quadrants, indicating a high probability that this health gain was achieved at lower cost. The ellipse also lies mainly below the CE threshold indicated in the figure. The 5% and 50% reliability intervals are in the lower right quadrant, which represents lower costs and greater health effects.
The integrated stroke service in Haarlem generated significantly less health effects at comparable costs (ICER was 78 480 per QALY lost). Figure 1 clearly shows that the stroke service experiment in Haarlem was not cost-effective: the 95% reliability range largely lies in the upper left quadrant, which represents higher costs and reduced health effects. The stroke service experiment in Nijmegen generated similar health effects at significantly higher costs (ICER was 3 041 550 per QALY gained). The reliability range shows clearly that the stroke service experiment in Nijmegen was not cost-effective: the ellipses lie entirely in the two top quadrants and above the CE threshold. It indicates that costs were consistently higher, while the chance of positive health effects was slightly greater than the chance of reduced health effects. The range for the Haarlem patients lies consistently above the CE threshold.
Figures 2 and 3 present the cost-effectiveness acceptability curves for Delft and Nijmegen, generated by non-parametric bootstrapping. These acceptability curves represent the probability that the regional stroke service experiment was cost-effective.44,45 For Delft, the low asymptote (representing the area in the low right quadrant in Figure 1) intersects the y-axis at 0.54 and the top asymptote (representing the area in the top left quadrant in Figure 1) at 0.92; this means that there is a 0.54 probability that the stroke service in Delft was less costly and at the same time more effective than conventional stroke care, and a 0.08 (1–0.92) probability that it was more costly as well as less effective. In between the asymptotes, the acceptability curve intersects the y-axis at 0.61, and with increasing threshold it increases to 0.85; this means that there is a 0.61 probability that the stroke service in Delft was cost-saving and a 0.85 probability that it was more effective. The probability that the experiment is acceptable at an ICER limit of 35 000 (Figure 1) was 0.75. For Nijmegen, there is a 0.00 probability that the stroke service was less costly and more effective than conventional stroke care, and a 0.45 (1–0.55) probability that it was more costly and less effective. The stroke service in Nijmegen was always more costly than conventional care, while there is a 0.54 probability that it was more effective. The probability that the experiment is acceptable at an ICER limit of 35 000 (Figure 1) was 0.01. Haarlem was omitted from Figure 2, as Figure 1 already made the odds clear; the probability that the stroke service in Haarlem is less costly and more effective than conventional stroke care was 0.01, while the probability that it is more costly and less effective was 0.87.
|
|
|
Discussion |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
Our study shows that integrated stroke services may lead to significant improvement in general health status after stroke onset in a cost-efficient way. This was seen in the group of stroke patients in Delft, a complete though relatively simple stroke service, but not in the two groups in regions with more complex stroke services. Delft, which chose the most straightforward set-up (i.e. integration between one hospital with a stroke unit, one nursing home with a stroke unit, one rehabilitation center and one home care provider, throughout supported by a stroke nurse), was the only region that, in retrospect, managed to meet the minimum criteria for stroke service for the majority of patients during most of the follow up period. Thus the most structured and complete care continuum (Delft) demonstrated the best results, and was a cost-effective alternative for traditional care for stroke patients. Comparing the outcomes of all three experiments, it appears that the cost-effectiveness of integrated stroke services may depend on the organization and focus of the integrated stroke setting. Only another similar but more elaborated study would be able to confirm this observation. This study would need more experimental clusters and would need to pre-define more precisely the structure and nature of the integration experiment.
This study confirms results from other effectiveness and efficiency studies on management strategies to promote a continuum of stroke care, both in randomized13,21 and non-randomized designs.13,20,46 A prospective non-randomized controlled design of studies of complex intervention is often seen as an inferior form of evidence, for example in Cochrane reviews. Others plead for considering non-randomized studies in a balanced way in a comparison with randomized studies, to assess whether RCT results materialize in clinical practice.28,47 In the Dutch case, one may argue, however, that the evaluation of a care continuum organized within a strictly geographically-defined area may make a non-randomized design necessary. The objective of the Dutch program is to provide appropriate care for all stroke patients in the region, and in this case, the practicalities of a real-life setting make randomization virtually impossible.17 Often non-randomized settings in health program evaluation may lead to false positive findings due to selection bias (both among providers and patients) and leave other local trends in epidemiology and care unaccounted. Our quasi-experimental study, however, shows mixed results among the stroke services settings in a concurrent comparison. We did observe some non-experimental changes in the control group settings, possibly reducing the differences in health outcomes and costs.
Given the non-randomized design, we considered case-mix differences in our quantitative analysis and, in a qualitative analysis, both case-mix and provider bias.29 We observed no significant differences in case-mix at baseline, given group sample sizes. We did however observe differences in the organisation of integrated services between the three experimental regions. These variations in integrated stroke services and the differences in group outcomes made us decide to examine the cost-effectiveness of each regional set-up separately.
The lack of evidence of health outcome improvement may be caused by a ceiling effect in the BI, leading to a underestimation of the effectiveness of treatment.48 The relation between BI and EuroQol-5D we used should not be stretched beyond the rehabilitation phase, about 6 months after stroke. A generic measure such as EuroQol-5D does not take fully into account stroke-specific cognitive and speech impairments, over and above the extent to which it affects self-care.49,50
The cost estimates in our study might be biased in three ways. First, we measured the start-up costs of co-ordination, but did not consider these costs in the cost-effectiveness ratio calculations, as this is not relevant for the long-run. Start up costs were estimated at 96 000 per region (range 66 000–134 000) during an implementation period of between 2 and 3 years. Structural costs per patient ranged between 175 and 260 (i.e. about 3% [2.7–3.4%] of total patient costs).31 Inclusion of these costs would shift the ellipses in Figure 1 only fractionally upward. We expect our comparative conclusions to remain unchanged, especially the favorable conclusion on the cost-effectiveness of integrated stroke services in Delft. Second, we did not take into account the impact of the experiments on informal care-givers. The mean subjective burden of giving care to a stroke patient was moderate, although about 30% of informal caregivers experienced substantial burden.51The level of both objective and subjective burden was associated with patient HRQoL and the number of care-giving tasks that needed to be performed. Due to substantially shorter institutionalization, the stroke service care in Delft was associated with longer stay and rehabilitation care at home.31 We expect earlier discharge to home to increase care-giver burden, and considering the age group, in particular spousal care-givers. Third, we did not consider the impact of shifting scarce health care funds to the coordination and intensification of care for stroke patients for other patient care. Our results demonstrate that in all regions, except for Delft, most of the waiting days are generated by patients being discharged to a nursing home, particularly by those going to a long-term residential unit.29 All three stroke services extended nursing home capacity and developed stroke units in nursing homes. The unit in Delft, however, was the only one that admitted all nursing home-dependent patients, including patients with suspected poor rehabilitative possibilities and long-term nursing home dependency. A considerable proportion of these patients in Delft returned home within 6 months. This finding may indicate the benefit of stroke rehabilitation units for patients who in other regions were excluded from rehabilitation programs because of severe disabilities. Although rehabilitation tends to be directed toward stroke patients with moderately severe disabilities, other research also demonstrated that stroke rehabilitation units may also improve outcome in severely disabled stroke patients.52 At the same time, patients that return home earlier in the rehabilitation phase probably need more (in)formal support. We expect that during the follow-up period the burden on informal care-givers was higher, and will report this later.
In conclusion, our study confirms the potential to improve health outcomes in an efficient way in the provision of stroke care in The Netherlands. The main impeding factor in the implementation of integrated care may be a focus on just a few essential elements in the integrated care continuum. Two main contributing factors are the reduction of hospital length of stay and, simultaneously, the adequate provision of multidisciplinary acute and long-term (rehabilitation) care during the different stages of disease after stroke.28
|
Acknowledgments |
|---|
We wish to acknowledge our colleagues in the project: Niek Klazinga, Jeroen van Wijngaarden, and Cornelie van Putte. We wish to thank Maiwenn Al for her assistance with the confidence surfaces and acceptability curves. The funding of this research by the Netherlands Organization for Health Research and Development (ZonMw, www.zonmw.nl) is also gratefully acknowledged.
|
References |
|---|
|
TopSummaryIntroductionMethodsResultsDiscussionReferences |
|---|
1. American Heart Association. Heart disease and stroke statistics—2003 update. Dallas, American Heart Association, 2002.
2. Thorvaldsen P, Asplund K, Kuulasmaa K, et al. for the WHO MONICA Project. Stroke incidence, case fatality, and mortality in the WHO MONICA Project. Stroke 1995; 26:361–7.
3. Elkins JS, Johnston SC. Thirty-Year Projections for Deaths From Ischemic Stroke in the United States. Stroke 2003; 34:2109–12.
4. Cooper R, Cutler J, Desvigne-Nickens P, et al. Trends and disparities in coronary heart disease, stroke, and other cardiovascular diseases in the United States: findings of the National Conference on Cardiovascular Disease Prevention. Circulation 2000; 102:3137–47.
5. Jamrozik K, Broadhurst RJ, Lai N, et al. Trends in the Incidence, Severity, and Short-Term Outcome of Stroke in Perth, Western Australia. Stroke 1999; 30:2105–11.
6. Anderson C, Ni Mhurchu C, Brown PM, Carter K. Stroke rehabilitation services to accelerate hospital discharge and provide home-based care: an overview and cost analysis. Pharmacoeconomics 2002; 20:537–52.[CrossRef][Web of Science][Medline]
7. Payne KA, Huybrechts KF, Caro JJ, Craig Green TJ, Klittich WS. Long term cost-of-illness in stroke: an international review. Pharmacoeconomics 2002; 20:813–25.[CrossRef][Web of Science][Medline]
8. Indredavik B. Stroke Units: the Norwegian Experience. Cerebrovascular Diseases 2003; 15(S1):19–20.
9. RIVM. Public Health Status and Forecasts 1997 [in Dutch]. Utrecht, National Institute of Public Health and the Environment (RIVM), 1997.
10. CBS. Mortality by cause of death. Voorburg/Heerlen, Central Bureau of Statistics, 2001. [http://www.cbs.nl]
11. DHF. Stappenplan Stroke Service; een zeven-stappen model om te komen tot regionale samenwerking in de CVA-zorg. The Hague, Dutch Heart Foundation, 1999.
12. RIVM. Cost of illness in the Netherlands 1999. Utrecht, National Institute of Public Health and the Environment (RIVM), 2002.
13. Early Supported Discharge Trialists. Services for reducing duration of hospital care for acute stroke patients. Cochrane Database Syst Rev 2002; 1:CD000443.[Medline]
14. Kwan J, Sandercock P. In-hospital care pathways for stroke. Cochrane Database Syst Rev 2002; 2:CD002924.[Medline]
15. Stroke Unit Trialists. Organised inpatient (stroke unit) care for stroke. Cochrane Database Syst Rev 2002; 1:CD000197.[Medline]
16. Van Straten A, Van der Meulen JHP, Van den Bos GAM, Limburg M. Length of hospital stay and discharge delays in stroke patients. Stroke 1997; 28:137–40.
17. Kwan J, Hand P, Sandercock P. Improving the efficiency of delivery of thrombolysis for acute stroke: a systematic review. Q J Med 2004; 97:273–9.
18. Chiu L, Shyu WC, Liu YH. Comparisons of the cost-effectiveness among hospital chronic care, nursing home placement, home nursing care and family care for severe stroke patients. J Adv Nurs 2001; 33:380–6.[CrossRef][Web of Science][Medline]
19. Evans A, Perez I, Harraf F, Melbourn A, Steadman J, Donaldson N, et al. Can differences in management processes explain different outcomes between stroke unit and stroke-team care? Lancet 2001; 358:1586–92.[CrossRef][Web of Science][Medline]
20. Grieve R, Porsdal V, Hutton J, Wolfe C. A comparison of the cost-effectiveness of stroke care provided in London and Copenhagen. Int J Technol Assess Health Care 2000; 16:684–95.[CrossRef][Web of Science][Medline]
21. Indredavik B, Fjaertoft H, Ekeberg G, Loge AD, Morch B. Benefit of an extended stroke unit service with early supported discharge: A randomized, controlled trial. Stroke 2000; 31:2989–94.
22. Keith RA. Rehabilitation after stroke: cost-effectiveness analyses. J R Soc Med 1996; 89:631–3.[Web of Science][Medline]
23. Langhorne P, Duncan P. Does the organization of postacute stroke care really matter? Stroke 2001; 32:268–74.
24. Langhorne P, Cadilhac D, Feigin V, Grieve R, Liu M. How should stroke services be organised? Lancet Neurol 2002; 1:62–8.[CrossRef][Web of Science][Medline]
25. Sulch D, Kalra L. Integrated care pathways in stroke management—a review. Age Ageing 2000; 29:349–52.
26. Sulch D, Melbourn A, Perez I, Kalra L. Integrated care pathways and quality of life on a stroke rehabilitation unit. Stroke 2002; 33:1600–4.
27. Sulch D, Evans A, Melbourn A, Kalra L. Does an integrated care pathway improve processes of care in stroke rehabilitation? A randomized controlled trial. Age Ageing 2002; 31:175–9.
28. Sinha S, Warburton EA. The evolution of stroke units—towards a more intensive approach? Q J Med 2000; 93:633–8.
29. Huijsman R, Klazinga NS, Scholte op Reimer WJM, Van Wijngaarden JDH, Van Exel NJA, Van Putte-Boon C, Prestholt FT, Koopmanschap MA, Niessen LW. Results of the EDISSE study of three regional experiments with stroke services [in Dutch]. Report. The Hague, ZonMW, 2001.
30. Limburg M. Can stroke patients be treated any better? [in Dutch]. Amstelveen, Ziekenfondsraad, 1998.
31. Van Exel J, Koopmanschap MA, Van Wijngaarden JDH, Scholte op Reimer WJM. Costs of stroke and stroke services: Determinants of patient costs and a comparison of costs of regular care and care organised in stroke services. Cost Eff Resour Alloc 2003; 1:2.[CrossRef][Medline]
32. Drummond MF, O’Brien B, Stoddart GL, Torrance GW. Methods for the economic evaluation of health care programs. Oxford University Press, 1997.
33. Gold MR, Siegel JE, Russell LB, Weinstein MC. Cost-effectiveness in health and medicine. New York: Oxford University Press, 1996.
34. The EuroQol Group. EuroQol—a new facility for the measurement of health-related quality of life. Health Policy 1990; 16:199–208.[CrossRef][Web of Science][Medline]
35. Dorman PJ, Waddell F, Slattery J, Dennis M, Sandercock P. Is the EuroQol a valid measure of health-related quality of life after stroke? Stroke 1997; 28:1876–82.
36. Dorman PJ, Slattery J, Farrell B, Dennis M, Sandercock P. A randomised comparison of the EuroQol and Short Dorm-36 after stroke? Br Med J 1997; 315:461.
37. Dolan P. Modelling valuations for EuroQol health states. Medical Care 1999; 35:1095–108.
38. Mahoney FI, Barthel DW. Functional evaluation: the Barthel Index. Md State Med J 1965; 14:61–5.[Medline]
39. Van Exel NJA, Scholte op Reimer WJM, Koopmanschap MA. Assessment of post-stroke quality of life in cost-effectiveness studies: the usefulness of the Barthel Index and the EuroQol EQ-5D. Qual Life Res 2004; 13:427–33.[CrossRef][Web of Science][Medline]
40. Briggs A, Fenn P. Confidence intervals or surfaces? Uncertainty on the cost-effectiveness plane. Health Economics 1998; 7:723–40.[CrossRef][Web of Science][Medline]
41. Harris A, Buxton M, O’Brien B, Rutten F, Drummond M. Using economic evidence in reimbursement decisions for health technologies: experience of 4 countries. Exp Rev Pharmacoeconomics Outcomes Res 2001; 1:7–12.
42. Davenport RJ, Dennis MS, Warlow CP. Effect of correcting outcome data for case mix: an example from stroke medicine. Br Med J 1996; 312:1503–5.
43. Ziguras SJ, Stuart GW, Jackson AC. Assessing the evidence on case management. Br J Psychiatry 2002; 181:17–21.
44. Fenwick E, O’Brien BJ, Briggs A. Cost-effectiveness acceptability curves—facts, fallacies and frequently asked questions. Health Economics 2004; 13:405–15.[CrossRef][Web of Science][Medline]
45. Löthgren M, Zethraeus N. Definition, interpretation and calculation of cost-effectiveness acceptability curves. Health Economics 2000; 9:623–30.[CrossRef][Web of Science][Medline]
46. Stavem K, Rønning OM. Survival of unselected stroke patients in a stroke unit compared with conventional care. Q J Med 2002; 95:143–52.
47. Kunz R, Oxman AD. The unpredictability paradox: review of empirical comparisons of randomised and non-randomised clinical trials. Br Med J 1998; 317:1185–90.
48. Duncan PW, Jorgensen HS, Wade DT. Outcome measures in acute stroke trials: A systematic review and some recommendations to improve practice. Stroke 2000; 21:1429–38.
49. Tatemichi TK, Desmond DW, Stern Y, Paik M, Sano M, Bagiella E. Cognitive impairment after stroke: frequency, patterns, and relationship to functional abilities. J Neurol Neurosurg Psychiatry 1994; 57:202–7.
50. Hochstenbach J, Mulder T. Neuropsychology and the relearning of motor skills following stroke. Int J Rehabil Res 1999; 22:11–19.[Web of Science][Medline]
51. Van Exel NJA, Koopmanschap MA, van den Berg B, Brouwer WBF, van den Bos GAM. Burden of informal caregiving for stroke patients: identification of caregivers at risk of adverse health effects. Cerebrovascular Diseases 2005; 19:11–17.[Web of Science][Medline]
52. Kalra L, Eade J. Role of stroke rehabilitation units in managing severe disability after stroke. Stroke 1995; 26:2031–4.
CiteULike 
Connotea 
Del.icio.us What's this?
This article has been cited by other articles:
|
|
 |
|
  R. Luengo-Fernandez, A. M. Gray, and P. M. Rothwell Costs of Stroke Using Patient-Level Data: A Critical Review of the Literature Stroke, February 1, 2009; 40(2): e18 - e23. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
L. M.T. Schouten, M. E.J.L. Hulscher, R. Akkermans, J. J.E. van Everdingen, R. P.T.M. Grol, and R. Huijsman Factors That Influence the Stroke Care Team's Effectiveness in Reducing the Length of Hospital Stay Stroke, September 1, 2008; 39(9): 2515 - 2521. [Abstract] [Full Text] [PDF]
|
|
|
|
|
|
W. Lo, K. Zamel, K. Ponnappa, A. Allen, D. Chisolm, M. Tang, B. Kerlin, and K. O. Yeates The Cost of Pediatric Stroke Care and Rehabilitation Stroke, January 1, 2008; 39(1): 161 - 165. [Abstract] [Full Text] [PDF]
|
|
| ||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||||