Background and Purpose Stroke unit rehabilitation tends to be directed toward stroke patients with moderately severe disabilities (“the middle group”). Data collected on a stroke rehabilitation unit, however, showed improving outcome over 3 years in patients with a poor prognosis (discharge home: 48% versus 16%, P<.02; discharge Barthel Index score: 9 versus 6, P<.05). The hypothesis that stroke rehabilitation units may improve outcome in severely disabled stroke patients was tested in this study.
Methods A randomized controlled study was undertaken in 71 patients with a poor prognosis who were treated either on a stroke rehabilitation unit (n=34) or on general wards (n=37) to compare outcome between the two groups. Data collected were also compared with those from a methodologically similar study undertaken 3 years ago.
Results Severe stroke patients treated on the stroke rehabilitation unit had a significantly better outcome compared with general wards (mortality: 21% versus 46%, P<.05; discharge home 47% versus 19%, P<.01; median length of hospital stay: 43 versus 59 days, P<.02). The number of stroke unit patients being discharged home had increased significantly from the previous study, with a trend toward improvement in median discharge Barthel Index score.
Conclusions Stroke rehabilitation units may improve outcome in severe stroke patients. This improvement appears to be due to the development of innovative management strategies that reduce mortality and institutionalization and enable caregivers to support more disabled stroke patients at home.
Most studies on stroke unit rehabilitation have concentrated on the “middle group” of stroke patients who have neither a good nor a very bad prognosis.1 2 3 4 5 6 7 8 9 It is believed, although it would be inconceivable to deny any stroke patient adequate treatment solely on the basis of severity of disability, that there may be advantages both for the patient and the hospital service in directing stroke rehabilitation unit resources toward patients most likely to benefit from such input.1 2 3 4 9 This was supported by a recent randomized controlled study that showed no improvement in outcome (except for a shorter length of hospital stay) in the subgroup of patients with very severe deficits and poor prognosis who were treated on a stroke rehabilitation unit.9 Despite these arguments, it is not always appropriate or ethical to exclude patients with a poor prognosis from stroke-unit rehabilitation, and each patient needs to be considered on an individual basis in service settings. Several stroke patients with a poor prognosis were treated on a stroke rehabilitation unit in the 3 years that followed a randomized controlled study on effectiveness in various patient groups,9 and their outcome was monitored by regular data collection and audit.
Data collected over the 3-year period (January 1991 through December 1993) showed a significant improvement in outcome of severely disabled stroke patients treated on the stroke unit, with a 30% increase in the number of patients being discharged home and a 3-point rise in the median Barthel Index score between 1991 and 1993 (Table 1⇓). These observations were of interest but inconclusive because the comparisons were nonrandomized and open to bias. Although all patients were within the same broad prognostic category, a selection bias within the grouping (eg, stroke side or type, level of cognitive impairment, continence, home support, and previous functional status) could not be excluded. Outcome in this patient group (especially destination of discharge) may also have been influenced by the greater and more flexible availability of resources because of changes in community support and social service provision that followed the implementation of the Community Care Act in Britain in 1993. Finally, there was the possibility that developing strategies in the stroke rehabilitation unit may have resulted in improved management, leading to better outcome for this patient group. It is possible that the stroke rehabilitation unit per se may have contributed to the improvement in outcome observed with time in patients with severe strokes and poor prognosis; this hypothesis was tested by undertaking a randomized controlled study in this group of patients.
Subjects and Methods
The 13-bed stroke rehabilitation unit is situated in a modern block and has dedicated physiotherapy and occupational therapy areas on the ward. There is a quiet room for interviews and assessments, a predischarge room, and a large day area for meals and activities. Input is provided by a physician with an interest in stroke, nurses, physiotherapists, occupational and speech therapists, a social services care manager, a pharmacist, a dietitian, and support staff.
Patients are not admitted directly to the stroke rehabilitation unit but are assessed with the Orpington Prognostic Score (OPS) by physiotherapists before transfer from general medical wards. The OPS ranges from 1.6 (best prognosis) to 6.8 (worst prognosis) and is a clinical score based on motor deficit, proprioception, balance, and cognition.10 11 On the basis of previous studies,9 12 13 14 patients scoring between 3 and 5 on the OPS are automatically accepted to the stroke rehabilitation unit. Patients whose score is greater than 5 are assessed on an individual basis and are transferred to the stroke rehabilitation unit at the discretion of the assessor. This decision is based on subjective assessment, and no objective criteria are applied.9
The stroke rehabilitation unit has a well-established philosophy of rehabilitation. Patients are assessed comprehensively by members of the multidisciplinary team on arrival. The emphasis is on identifying problems affecting activities of daily living and mobility in the context of the home environment and support available. An appropriate management strategy is designed, and long-term goals as well as the time required to achieve these goals are agreed on in consultation with patients and their relatives.
Patients have an individualized rehabilitation program reviewed on a daily basis. In addition to formal therapy sessions, nursing staff are trained to reinforce therapy on the ward under the guidance of the therapists. Relatives are encouraged to be present during nursing and therapy sessions to appreciate the patients’ abilities and to learn specific skills. The entire rehabilitation team meets formally once a week to discuss the progress of each patient, at which time achievement objectives or time frames may be modified. These decisions are communicated to the patients and their relatives by the relevant member of the multidisciplinary team.
Medical and nursing care, physiotherapy, speech and occupational therapy input, dietary and pharmacy advice, and social worker support are freely available to stroke patients on general medical wards. However, significant differences of the wards from the stroke rehabilitation unit include (1) emphasis on acute rather than rehabilitation care, (2) deficiencies in multidisciplinary planning and goal setting, (3) lack of patient/caregiver involvement, (4) inadequate support or counseling facilities, and (5) undue nihilism about stroke outcome.
Data on the stroke rehabilitation unit were collected in “real time” with a multidisciplinary, integrated data-collection system (Orpington Stroke Management System) comprising well-validated and frequently used assessments in the major domains of stroke rehabilitation.15 Members of each discipline were required to enter demographic, medical, nursing, therapy, and psychosocial data appropriate to their specialty as a part of their normal work activities. The quality, completeness, and accuracy of data collection have been shown to be comparable to those of conventional methods.15 The system has also been used previously for data collection in other studies.9 12 13 14 The databases of the integrated system were searched to provide the relevant data for the year-by-year comparison of stroke patients within the poor prognostic category (OPS >5).
The Controlled Trial
The randomized controlled trial was limited to stroke patients with an OPS >5 and referred for rehabilitation on a stroke unit. After stabilization on general medical wards, patients with severe deficits were assessed for stroke type,16 neurological deficits, and functional disabilities. Patients were referred for specialized rehabilitation as soon as they were medically stable. Requests were received by a central stroke office, where patients in this group were randomized to the stroke rehabilitation unit or to continue to be treated on general wards. The process of randomization was not influenced by availability of beds because the unit had the flexibility to accommodate extra patients. The median duration between stroke and randomization (which took place at the time of referral) was 9 days. Neither the staff of the stroke rehabilitation unit nor those of general wards were aware of the comparative study, although there was potential for bias because of the involvement of the consultant on the stroke rehabilitation unit (L.K.).
Patients were treated according to existing practices in both settings. Data on demography, stroke type and side, and neurological and function deficits on initial assessment were collected for both groups. Outcome was measured at the time of hospital discharge and included mortality, destination of discharge, Barthel Index score, and length of hospital stay.
Group homogeneity was analyzed using the χ2 test for multigroup comparisons of sex, stroke side, stroke type, mortality, and destination of discharge. Multigroup comparison of medians was undertaken using the Kruskal-Wallis test. Age and average length of hospital stay were compared using one-way ANOVA. Data in the controlled study were compared using the χ2 test, t test, or Mann-Whitney test.
The randomized controlled study was undertaken over a 1-year period. Seventy-three of the 76 patients with severe strokes (OPS >5) referred to the central office were randomly allocated to the stroke rehabilitation unit (n=36) or remained on general medical wards (n=37). Of the three patients referred but not randomized, assessment showed a cerebral tumor in one patient and cerebral metastases in another. One patient refused transfer to the stroke rehabilitation unit because it was in a different hospital. Data could not be collected for two patients on the stroke rehabilitation unit because, as out-of-district residents, they were transferred to other hospitals.
Baseline patient characteristics, stroke type and severity, and admission Barthel Index scores of patients transferred to the stroke rehabilitation unit were comparable to those of patients treated on general wards (Table 2⇓). There were no significant differences in premorbid functional ability or available family support between the two groups. The majority of patients had been independent with no significant functional disability before the stroke. CT scanning was undertaken in 62 (87%) patients. The number of patients who underwent scanning was comparable between the two settings. The majority of patients in both settings had a total anterior circulation infarct and were severely disabled (Table 2⇓).
An “on-treatment” analysis showed that mortality was significantly reduced in patients treated on the stroke rehabilitation unit and significantly more patients were discharged home compared with general wards (Table 3⇓). Patients treated on the stroke rehabilitation unit had a higher median discharge Barthel Index score compared with those on general medical wards, although this failed to achieve statistical significance. The hospital length of stay was significantly shorter in patients treated on the stroke rehabilitation unit (Table 3⇓).
Of the two patients who were transferred to other hospitals from the stroke rehabilitation unit, one patient was eventually discharged (destination and functional status not known). No information was available for the other patient. A “worst-case” analysis (assuming that the discharged patient was institutionalized and the other patient had died) continued to show significant differences in mortality and discharge home between the two settings.
Comparative data from the previous study showed that the number of patients being discharged home from the stroke rehabilitation unit had increased over the years (Table 3⇑). There was a trend toward an improvement in the median Barthel score for activities of daily living of severely disabled patients treated on the stroke rehabilitation unit, although this did not achieve statistical significance (Table 3⇑). The possibility of type II error due to an inadequate sample size, however, cannot be excluded. In contrast to the previous study, the length of hospital stay decreased significantly for patients on general wards, but there were no significant changes in mortality, destination of discharge, or functional ability at discharge in this group (Table 3⇑).
Patient selection can significantly influence the effectiveness of stroke rehabilitation units, and the greatest benefits of these units are seen in patients with moderately severe deficits and an intermediate prognosis rather than in patients with mild or very severe strokes.3 4 5 9 17 The present study shows that stroke rehabilitation units may improve outcome in even severely disabled stroke patients. It is possible that developments in community services and general improvement in care with time across all settings may have contributed to the better outcome, but these factors alone do not explain the improvement in outcome observed with time in severely disabled stroke patients treated on the stroke unit.
This study suggests that stroke rehabilitation units are a dynamic environment in which developing strategies and expertise can result in improvements in outcome of even severely disabled stroke patients. In addition to differences in therapy input and medical and nursing care between the stroke rehabilitation unit and general medical wards,9 13 14 improvements in management practices on the stroke rehabilitation unit (since undertaking the previous study) that may have contributed include (1) enhancement of rehabilitation skills over time due to “in-service” staff training and an increase in confidence, which have resulted in improved multidisciplinary functioning with blurring of professional boundaries and greater involvement with patients and their caregivers, and (2) better liaison with patients and their families resulting in their increased involvement in goal setting and discharge planning; more adequate counseling and support from ward staff, social workers, and voluntary agencies; “hands-on” training to manage disability before discharge; and follow-up support from a social worker for 3 months after discharge.
There is a possibility of bias being introduced because of the involvement of the consultant (L.K.) in providing care to one group of patients. This bias was unavoidable in the service circumstances in which the study was undertaken, but it is unlikely to have influenced outcome (other than the length of hospital stay) because of the involvement of other professionals in the rehabilitation and decision-making process. Nonreferral (and noninclusion) of some patients with severe stroke may have been another potential source of error. Hospital audit, however, suggested that any intervention (other than supportive terminal care) would have been inappropriate in those severe stroke patients not referred for rehabilitation. These limitations need to be acknowledged but are unlikely to have had a significant influence on the results of this study.
It is possible that the finding of an improved outcome associated with stroke-unit rehabilitation in severe stroke patients may be due to external factors rather than improved practice on such units. This is unlikely because poor outcome is well documented in this patient group2 3 4 9 16 and is thought to be refractory to stroke rehabilitation unit intervention.17 Several randomized controlled trials in the past have either excluded this group of patients1 17 18 or shown no differences in outcome between specialist and nonspecialist settings.3 9 17 18 Despite changes in community care provisions and general improvements in standard of care, there were no differences in outcome (except for hospital length of stay) in severe stroke patients treated on general wards between this and a previous study (Table 3⇑). In contrast, there was a significant difference in the number of patients being discharged home between the two time points for the stroke rehabilitation unit. There also appeared to be an improvement in median discharge Barthel Index scores, although its interpretation is limited by the possibility of a type II error.
The improvement in outcome of severe stroke patients treated on the stroke rehabilitation unit may have important implications for patient selection for these units. The issues regarding “triage” to improve the effectiveness of stroke unit rehabilitation have been discussed in several reports.1 3 4 9 10 17 The limitations of such criteria are well recognized, and pragmatism has been recommended in their application to clinical situations.9 10 This caution has been justified in this study, and it is likely that the types of patient who will benefit from specialized rehabilitation will change depending on the expertise developed within these units. Despite this limitation, the concept of prognostic grouping of stroke patients undergoing rehabilitation remains important. Prognostic grouping reflects the resource requirements and likely outcome of patients, which are important in planning health services as well as monitoring their delivery (health benefit groups). On a more local level, prognostic grouping can help in developing and refining critical pathways of care for stroke patients. Such stratification of stroke patients will also continue to be of importance in assessing the impact of interventions in rehabilitation, as seen in this and other studies.9 12 13 14
This study suggests that developing strategies on stroke rehabilitation units may play a significant role in the management of stroke patients. Although there is considerable work yet to be undertaken before the “black box” of stroke unit interventions can be unraveled, it would appear that specialist training, individualized rehabilitation programs, caregiver education and hands-on training, seamless professional collaboration, and innovative planning of postdischarge care can significantly alter outcome even in the more disabled stroke patients.
- Received May 2, 1995.
- Revision received July 12, 1995.
- Accepted July 17, 1995.
- Copyright © 1995 by American Heart Association
Garraway WM, Akhtar AJ, Prescott RJ, Hockey L. Management of acute stroke in the elderly: preliminary results of a controlled trial. Br Med J. 1980;280:1040-1043.
Prescott RJ, Garraway WM, Akhtar AJ. Predicting functional outcome following acute stroke using a standard clinical examination. Stroke. 1982;13:641-647.
Ebrahim S. Clinical epidemiology of stroke. Oxford, UK: Oxford University Press; 1990.
Indredavik B, Bakke F, Solberg R, Rokseth R, Haahein LL, Holme I. Benefit of a stroke rehabilitation unit: a randomized controlled trial. Stroke. 1991;22:1026-1031.
Kalra L, Dale P, Crome P. Improving stroke rehabilitation: a controlled study. Stroke. 1993;24:1462-1467.
Kalra L, Dale P, Crome P. Evaluation of a clinical score for prognostic stratification of elderly stroke patients. Age Ageing. 1994;23:492-499.
Kalra L. Does age affect benefits of stroke unit rehabilitation? Stroke. 1994;25:346-351.
Kalra L. The influence of stroke unit rehabilitation on functional recovery from stroke. Stroke. 1994;25:821-825.
Kalra L, Yu G, Wilson K, Roots P. Medical complications during stroke rehabilitation. Stroke. 1995;26:990-994.
Kalra L, Fowle AJ. An integrated system for multidisciplinary assessments in stroke rehabilitation. Stroke. 1994;25:2210-2214.
Dennis M, Langhorne P. So stroke units save lives: where do we go from here? BMJ. 1994:309:1273-1277.
Stroke Unit Trialists’ Collaboration. A systemic overview of specialist multidisciplinary team (stroke rehabilitation unit) care for stroke inpatients. In: Warlow C, Van Gijn J, Sandercock P, eds. Electronic publication in Stroke Module of Cochrane Database of Systematic Reviews. London, UK: BMJ Publishing Group; Disk issue 1, 1995.