Statistics from Altmetric.com
Large differences in outcome and resource utilisation
Stroke is a major health burden throughout Europe. Despite a reducing incidence in Western Europe,1 the rise in the elderly population and increasing incidence rates in Eastern Europe will result in it becoming a mounting population burden. It consumes a large amount of healthcare resources; in the UK about 5% of the National Health Services’ budget is spent on stroke care. Therefore optimal models for delivery of stroke care, which result in good outcome at reasonable cost, are of great importance. Although robust data has shown that care in a stroke unit (with most trials looking at rehabilitation units) is associated with improved outcome,2 we have limited understanding of which components of this organised care are responsible for this benefit. Comparing practice and outcome across different European countries may give clues that can help us to develop new hypotheses and interventions, which should then be ideally tested in controlled studies.
In this issue (pp 1702–6) the BIOMED European Study of Stroke Care Group present data from 11 centres in several Central and West European countries.3 The group have previously shown wide variation in early (3 month) mortality and dependency, even after adjustment for case mix and other confounding factors. Three month mortality ranged from 42% in one UK centre to 19% in France.4 Centres involved in this study may not be typical of stroke care within their countries; their involvement in a research project of this type suggests they are already particularly interested in stroke care. Indeed a UK audit suggests wide variations in stroke care provision within a single country.5 However, the national differences noted by the BIMOED group are consistent with data from other sources. For example, in the International Stroke Trial there were wide national differences in outcome, not fully explained by case mixed variables, and the UK again had poor outcome.6
What explains these differences? One needs to be cautious that they are not, at least partly, explained by differences in case mix including hospital referral practice, ethnic and other differences in stroke subtype, and socio-economic differences. The latter was not controlled for in the BIOMED project, while some residual confounding by stroke severity is likely despite attempts to control for it. However, such confounding is unlikely to fully account for these differences, and the wide variations in patterns of stroke care seen throughout Europe appear to be important. Many of these have been described previously by the European BIOMED Group.7 One feature that varies widely across Europe is the intensity with which the acute episode of stroke is managed. In many units (e.g. Dijon, France and Kuopio and Turku, Finland), almost all patients are admitted directly to a neurology unit, while in others admission is to a dedicated acute stroke unit (e.g. Florence, Italy and Copenhagen, Denmark). In contrast many patients in the UK are admitted to a general medical ward. Fifty percent in Dijon are admitted to an intensive care unit while in other countries the figure is zero. There are wide variations in the use of physiological monitoring, and attempted control of cerebral blood flow, fever, blood glucose, and hydration, and in the treatment of severe cerebral oedema with drugs, craniectomy, or hypothermia. There are also wide variations in availability and use of early brain imaging and other investigational facilities, as well as total medical, nursing, and therapist time utilised per patient. Importantly total cost is not directly related to outcome. For example, in one UK centre overall cost was one of the highest, but outcome was one of the worst.7 The great majority of stroke care cost is due to bed occupancy, including nursing and therapy costs. Therefore improved acute care—if it results in reduced length of stay—is likely to be very cost effective.
This paper extends the previous findings to focus on longer term (12 month) survival and outcome, and again demonstrates large differences in both outcome and resource utilisation. For example, availability and support from family and friends differed widely between countries. A number of variables were identified as being associated with death and dependency, and these included male gender, pre-stroke handicap, and at presentation the presence of coma, incontinence, swallowing problems, or weakness.
How can we use this and other data to improve stroke care? A number of messages are emerging from different studies. Stroke care should be in specialised units and should be delivered by appropriately trained teams. Increasing evidence suggests early admission to a specialised unit, early brain imaging and diagnosis, and attention to early physiological care is associated with improved outcome,8,9,10 although the evidence for benefit of most specific acute interventions is not well founded.11 However, many questions remain. We understand little about which components of rehabilitation therapy are beneficial and what the dose–response relationship is for these. There are also many uncertainties in acute care. For example, how aggressive should we be with blood pressure lowering and other physiological interventions in the immediate post stroke period? What is the role of interventions such as craniectomy, hypothermia, and extending the time window for thrombolysis? Many of these issues are being addressed in randomised controlled trials, but others are difficult to test in this way, and answering them all in sufficiently powered trials is a major challenge. Until then comparative data provided by this and other studies will provide useful clues to best management. It will also act as a useful reminder to those countries with poor outcome, such as the UK, of the need for an improvement in stroke services.
Large differences in outcome and resource utilisation
If you wish to reuse any or all of this article please use the link below which will take you to the Copyright Clearance Center’s RightsLink service. You will be able to get a quick price and instant permission to reuse the content in many different ways.