ArticlesCan differences in management processes explain different outcomes between stroke unit and stroke-team care?
Introduction
Several randomised studies have shown that management of patients in a stroke unit, both in the acute phase and during rehabilitation, is associated with significant reductions in mortality and dependence.1, 2, 3, 4 However, the exact nature of stroke-unit care remains poorly defined because of wide variations in practice in different units. Previous research5 to identify common themes in stroke management has shown that multidisciplinary teamwork, education of staff, and involvement of patients and their relatives characterise effective stroke rehabilitation. Research into the components of acute management has been more modest, but practices such as thrombolysis in selected individuals,6 physiological homoeostasis,7 early prescription of aspirin,8 anticoagulation in patients with atrial fibrillation,9 and early mobilisation10 might be important in reduction of mortality and dependence after a stroke. Additionally, indirect evidence suggests that a higher proportion of patients in stroke units receive such interventions than those in non-specialist settings.2, 10
The processes of stroke management have not been investigated in great detail and most information is either superficial or based on selective case studies.11 Whether processes suspected to be important for favourable outcomes can be replicated in generic settings, thereby reducing the need for geographically defined locations to manage stroke patients, is unknown.1 We do not know how stroke units reduce mortality. Practices such as adequate hydration, maintenance of physiological homoeostasis, and prevention of complications probably contribute substantially,12 but such contributions have not been investigated in randomised trials that include patients from stroke onset with regular and complete follow-up using objective criteria for processes or events.
The translation of trial efficacy of stroke units into clinical effectiveness in mainstream practice requires replication of not only the structure but also the processes associated with a favourable outcome. Objective assessment of processes in specialist management is complex because specialist management consists of several elements working together. It might be possible to overcome some of these problems by categorising processes into theoretically defined and clinically relevant domains, which could then be used to analyse differences between settings.11 Since some processes might be interdependent, this dependency and the relation of the processes to the frequency of complications can be explored by further analysis to identify processes most strongly associated with better outcome.
We have previously reported a randomised controlled trial of 457 patients with acute stroke, which showed that stroke units were more effective than a specialist stroke team (and specialist domiciliary care) in reducing mortality and dependence after stroke.4 The objective of this study was to investigate factors that could explain why mortality and dependence were lower in the stroke unit than in stroke-team care by using prespecified objective criteria to compare processes of management and frequency of complications associated with stroke between different settings.
Section snippets
Patients
The trial4 allocated 457 patients with acute, moderately severe stroke to treatment in a stroke unit (n=152), general medical wards with specialist stroke team support (n=152), or specialist domiciliary stroke care (n=153) at the time of presentation. Patients allocated to the domiciliary-care group of the original trial were excluded from this analysis because a third (51) could not be managed at home and because management processes at home would be difficult to compare with hospital care.
Results
Of 152 patients allocated to the stroke unit, 118 (78%) were admitted within 24 h and 27 (18%) between 24 and 48 h of stroke onset. 124 of 153 (82%) patients allocated to general wards with specialist team support were admitted within 24 h and 20 (13%) between 24 and 48 h of onset. Both groups were well matched for median age (75 vs 77 years), sex (women 47% vs 51%), comorbidity, risk factor profile, and amount of independence before stroke. A similar proportion of patients in the stroke unit
Discussion
Our results show significant differences in frequency of assessment procedures, acute management, early rehabilitation, and secondary complications between patients allocated to the stroke unit and those allocated to stroke-team care on general medical wards—possibly the reason for the higher mortality and dependence seen in non-specialist settings. They also show that specialist aspects of stroke management cannot be replicated in generic settings, even when ongoing specialist stroke-team care
References (32)
- et al.
Alternative strategies for stroke care: a prospective randomised controlled study of stroke unit, stroke team, and domiciliary management of stroke
Lancet
(2000) - et al.
Treatment and secondary prevention of stroke: evidence, costs and effects on individuals and populations
Lancet
(1999) - et al.
Classification and natural history of clincally identifiable subtypes of cerebral infarction
Lancet
(1991) A late step in the right direction for stroke care
Lancet
(2000)- et al.
Effect of blood pressure and diabetes on stroke progression
Lancet
(1994) - et al.
Body temperature in acute stroke: relation to stroke severity, infarct size, mortality and outcome
Lancet
(1996) Collaborative systemic review of the randomised trials of organised inpatient (Stroke Unit) care after stroke
BMJ
(1997)- et al.
Stroke units versus general medical wards I: twelve and 18 month survival, a randomised controlled trial
Stroke
(1998) - et al.
Does the organisation of postacute stroke care really matter?
Stroke
(2001) How do stroke units improve patient outcomes? A collaborative systemic review of the randomised trials
Stroke
(1997)