Section 5 Resource Implications in the original guideline document outlines the economic evidence for aspects of stroke recovery and rehabilitation. The section aims to be useful in guiding decisions about the structure of services and may be used by those who plan or organise care.
Stroke Unit Care
One systematic review identified three studies comparing the costs and outcomes of stroke units to that on a general ward. All three studies were based in Europe (United Kingdom, Sweden and Germany) and included costs of community and outpatient care. All three studies found modest cost savings (3-11%) using stroke unit care, however the figures failed to reach significance. The authors concluded that there was "some" evidence for the costs to be at least equivalent to conventional care.
Two subsequent published studies were also identified. One modelling study used the Markov model to predict the medium-term (5 year) impact of setting up stroke units in France. The study estimated a 12% increase in costs involved in setting up and running stroke units compared with conventional care. Furthermore, the study predicted an incremental cost-effectiveness ratio (ICER) for stroke units of 1,359 pounds per year of life gained without disability. The authors suggested this was well within the threshold (53,400 pounds) recognised by the international scientific community. However, only running costs were involved in the evaluation and further costs, such medical imaging, have not been considered.
The other subsequent trial-based study assessed three different models of providing coordinated stroke care compared to routine care in the Netherlands. Organised care was found be offered at similar costs while achieving improved health outcomes. Caution is required to apply the results of this study, as the health system model is significantly different to the Australian health care system.
In Australia, one researcher has demonstrated that when modelled over the lifetime of a cohort of first-ever stroke patients, stroke units when compared to conventional care produced considerable gains in terms of health benefits with these additional benefits entailing additional costs. There was an additional lifetime cost (in Australian dollars) of AUD$1,288 per disability adjusted life years (DALY) recovered, or alternatively AUD$20,172 per stroke averted or AUD$13,487 per premature death averted. It was determined that the stroke unit intervention was cost-effective given the small additional costs per extra unit of benefit gained.
Currently only 19% of public hospitals report providing stroke unit care. 403 Stroke units improve outcomes for people with stroke (see section 1.1.1 in the original guideline document). Furthermore, the cost of providing stroke unit care once set up is only slightly higher, or at least equivalent, to general ward care. Although this literature does not specifically indicate the real costs of setting up a stroke unit, there is evidence that health services should be organised to provide stroke unit care and that considerable gains in terms of health benefits could be achieved.
Early Supported Discharge (ESD)
One systematic review identified eight trials evaluating the economic implications of ESD compared with conventional care. Two studies were conducted in Australia with the remainder from Hong Kong (one), Canada (one), Sweden (two) and the United Kingdom (two). All but one of the studies compared ESD using home-based services compared to conventional services (noted to be either hospital rehabilitation or mix of hospital and community rehabilitation). Of the eight studies included, six studies were noted as having medium or high methodological quality. These studies reported a trend for reduced costs of between 4-30% with ESD, however this cost saving was found to be statistically significant in only one of the six studies. The authors concluded that there was "moderate" evidence that ESD services provided care at modestly lower total costs than conventional care. However the heterogeneity of the ESD care provided was noted along with the uncertain impact of ESD care on hospital readmission and informal carers. The review also concurred with the previous summary (see section 1.2.1 in the original guideline document) that ESD favours stroke survivors with mild or moderate disability.
One subsequent UK trial-based study assessed the outcomes and costs of early domiciliary care compared to hospital based care. A societal perspective for costs was used based on 1997/8 prices. Mean costs for health care and social care costs over 12 months were 6,840 pounds for domiciliary care compared to 11,450 pounds for stroke units. In terms of Quality Adjusted Life Years (QALYs) these were less for domiciliary care when compared to stroke unit care (0.221 versus 0.297). Cost effectiveness was calculated using incremental cost-effectiveness ratios (ICERs) for avoiding an additional 1% of deaths or institutionalisation that ranged from 496 pounds (without informal costs) to 1,033 pounds (with highest estimate of informal costs) for stroke unit care compared with domiciliary care. Based on each additional QALY gained the costs ranged from 64,097 pounds to 136,609 pounds. Hence in this study, health outcomes were lower using this ESD model but ESD was found to be cheaper.
Data specific to the Australian context was included in the previous review and warrants further discussion. The data from a meta analysis of ESD (12 trials, N=1277, search date March 2001) was used to apply costs from the Australian health system. Hospital costs were taken from the Australian National Hospital Cost Data for 1998/1999, domiciliary rehabilitation costs were taken from a single study of domiciliary rehabilitation care (Adelaide stroke study) and costs related to other community services were taken from the Australian Department of Health and family Services Report, 1996/1997. Using a cost minimisation analysis (i.e., health outcomes were found to be equivalent) ESD was found to be 15% lower regarding overall mean costs (in Australian dollars ($A16,016 versus $18,350). Cost estimates were based over a 12-month period and did not include any indication of set up costs. It was highlighted that the included studies were all based in urban centres confirming the view that ESD should only be considered where appropriate resources are available to provide effective domiciliary care. A small shift of costs from the secondary sector to primary section was noted (more general practitioner [GP] visits with ESD care) however no difference was found in the cost of routine community and outpatient services. Overall ESD was found to provide a cost saving alternative to conventional care and the authors concluded that it therefore should be considered for certain subgroups of people with stroke.
The above studies provide limited evidence regarding the cost-effectiveness of ESD in Australia. It can be concluded from these studies that ESD may produce equivalent outcomes at potentially a reduced cost for urban settings.
Community Rehabilitation
Economic evaluations of community rehabilitation are limited to cost-description studies. One systematic review identified four trials comparing different models of community care and found conflicting results. Three studies were undertaken in the United Kingdom and one in Sweden. Two studies comparing home-based rehabilitation to a day hospital or outpatient rehabilitation models reported consistent increases in costs for home-based care between 26-27%; however, this increase was not found to be significant. Another study found physiotherapy services were 38% lower (statistically significant) for home-based care compared to a day hospital. The fourth study found community rehabilitation (home-based) was of similar costs in the first twelve months when compared to hospital rehabilitation. Two included studies noted that the cost burden was shifted from hospital services to home help or social services. The authors of the review, however, stated that no conclusions could be drawn.
From this literature it is not possible to make conclusions regarding the cost effectiveness of any one model of community rehabilitation and whether or not any additional costs that may be incurred result in more health gains than current practice.