Review article
Evaluation of stroke rehabilitation by randomized controlled studies: a review de Pedro-Cuesta J, WidCn-Holmqvist L, Bach-y-Rita P. Evaluation of stroke rehabilitation by randomized controlled studies: a review. Acta Neurol Scand 1992: 86: 433-439. 0 Munksgaard 1992.
In all, 22 reports of 20 randomized, controlled rehabilitation studies were evaluated. In 18 of these, the design of the trial was parallel, with a cross-over format being employed in the remaining 2 instances. Seven studies related to intensive rehabilitation during the early post-acute period. In six others, specific techniques - sometimes associated with traditional physiotherapy procedures - were compared: biofeedback, perceptual retraining, and amphetamine treatment. Eight experiments evaluated speech therapy in aphasia. Frequently, methodological considerations limited the interpretation of the results. The review showed that: 1) as regards activities of daily life and motor function, differences as between rehabilitation in stroke units on the one hand and non-rehabilitation or rehabilitation in medical wards on the other, were detected in relatively few quality studies and remained particularly inconclusive insofar as life in the home environment was concerned; 2) rehabilitation for aphasia and perceptual dysfunction proved effective for at least several months after acute stroke; 3) in general, examination of the reports cited revealed no differences in the effects of biofeedback and perceptual retraining vs conventional physiotherapy; 4) rehabilitation, whether administered by specialists or amateur caregivers purpose-trained by specialists, proved equally effective for aphasia. These conclusions constitute a valuable basis for the development and evaluation of modern rehabilitation programs for stroke patients.
While, as a medical technology, stroke rehabilitation is well defined in terms of goals, namely, restoration of brain function and adaptation, it is poorly identified in physical terms, in view of the fact that it embraces a variety of procedures, devices, and organizational and supportive systems (1-2). In affluent countries, this technology is widely available and closely tied in with hospital care of stroke patients (3). It is rather costly however. Accordingly, its true effectivenessis under debate, especially since current knowledge is mostly based on professional experience and less frequently on scientific observation (4). This paper seeks to review a number of strokerehabilitation evaluation studies and, on this basis, proceed to point the way for future research. "Spontaneous", natural recovery of brain function after acute stroke, may well overlap with recovery attributable to rehabilitation. It follows therefore that comparison with appropriate controls is required if the effects of any intervention are to be accurately assessed. Since multiple factors, such as,
J. de Pedro-Cuesta l r 3 , L. Widen-Holmqvist P. Bach-y-Rita lS4
',',
' Unit for Neuroepidemiology and Health Services Research, Huddinge University Hospital, Department of Physical Therapy, Karolinska Institute, Solna, Sweden, Department of Applied Epidemiology, National Centre for Epidemiology, "Carlos Ill" Institute of Health, Madrid, Spain, Department of Rehabilitation Medicine, University of Wisconsin, Madison, USA
Key words: Aphasia; clinical trial(s); epidemiology; randomized-trial(s); rehabilitation; stroke; technology(ies) Jesbs de Pedro-Cuesta, Department of Applied Epidemiology, National Centre for Epidemiology, "Carlos Ill" Institute of Health, c/ Sinesio Delgado, 6, 28029 Madrid, Spain Accepted for publication April 24, 1992
age, education level, time from the acute event, site and nature of lesion, highest level of disturbance of consciousness, degree of functional deficit, etc., have been shown to be related to improvement of functional capacity (5-8), we chose to restrict this review to randomized controlled experiments. Material and methods
From 68 1 publications - letters, abstracts, articles and book chapters - collected on a personal basis by the authors or identified through three Medline searches, focusing on stroke-rehabilitation from 1984-August 1991 and aphasia-therapy or aphasiarehabilitation from 1986-August 1991, we selected 44 evaluation studies relevant for our purposes (7, 9-51), with all single case studies and those where less than 80% of the patients corresponded to stroke survivors being summarily excluded. Of this total of 44,22 studies, in which a control group was included and a formal randomization procedure was de433
de Pedro-Cuesta et al. scribed (9-30), were selected for detailed examination. These 22 reports corresponded to 20 randomized, controlled experiments (9-30), while the remaining 22 dealt with uncontrolled or nonrandomized studies (7, 3 1-5 1). In the randomized controlled trials, a parallel design was to be seen in eighteen instances (9-16, 18-29) and a cross-over mode in the other two (17, 30). Seven studies involved intensive rehabilitation during the early post-acute period (9-16), regardless of whether such therapy had been provided within the formal context of a stroke unit. In six other studies, diverse but specific rehabilitation techniques, such as, biofeedback, perceptual retraining and amphetamine treatment - sometimes linked to traditional physiotherapy (PT) - were compared to conventional, mainly PT-type, procedures (17-2 1). Eight experiments were devoted to the evaluation of speech therapy in aphasia (22-30). In reviewing the randomized controlled studies (9-30) attention was systematically given to: 1) the nature, intensity and duration of the intervention, plus the time elapsed since the last acute-stroke episode; 2) the catchment population generating the patients; 3) patient selection; 4)distribution of potential confounders after randomization; 5 ) outcome measurements; 6) methods of comparison; 7) statis-
tical assessment-methods and power; and lastly, 8) interpretation of results. The terms employed by Balliet - early post-acute and late post-acute - were used to denote, in the former case, intervention during the year immediately following the acute episode and, in the latter case, intervention at any time thereafter (52). Review of methods and results of randomized controlled studies
The characteristics and results of trials, which have sought to evaluate the kind of intensive, early postacute rehabilitation frequently provided in specialized stroke units, are summarized in Table 1. Inconclusive results were reported in two early studies where random variation was not assessed (9), dropout was either important (10) or not investigated (9), and the patients were selected from among those with lower cultural and socio-economic status (10). Several analyses on one set of patients at the beginning of the 1980s uncovered temporary effects - i.e., either during the hospital stay or less than one year after stroke - of intensive rehabilitation in stroke units, coupled with a decrease in mortality (1 1-13). However, this last finding is less convincing, since not only was the time period between stroke onset
Table 1. Summarized data from evaluation of early, intensive or stroke-unit based rehabilitation programs Author Year
Design and interventions
Measured outcome
Study population
Comparison methods
Results
Comments
Rehabilitation vs no rehabilitation t medical hospital care
Neurological status and ADL
Hemiparesis up to 2 months after stroke 82, unselected
Percentages Raw data No statist. assessment
17% with no deficit in the control group
Inconclusive NR drop-out Insensitive methods
Peacock PB et al. 1972 (10)
Intensive rehabilitative vs routine care
Independence and vocational ordinal scales
52 patients (low income and cultural level)
Proportions
No differences
Inconclusive
Garraway WM et al. 1980 (1 1,121
Stroke rehab-unit vs medical wards
ADL and mortality at discharge and 1 year later
31 1 unselected conscious, hemiparetic
Proportions
Higher mortality, and dependence at discharge in wards
Biased mortality estimate Likely temporary effect of stroke unit
Smith ME et al. 1981 (17)
Stroke unit (early short) vs medical wards (late but longer)
Neurological status & ADL at discharge or 16 weeks
152 vs 109 patients (acute)
Proportions
65% ~ ~ 4 9 % independent at end of trial
Stroke Unit effective in hospital environment
Smith DS
Intensive (I) vs conventional (C) vs no rehab (NR)
ADL at 3, 6, 12 months
Acute stroke 46, 43 and 4 4 patients
Average change in ADL (t-test)
I and C better than NR
Rehabilitation I or C: effect one year later
Stevens RS et al. 1984 (15)
Stroke unit vs conventional
Complications, disability and death
112and 116 patients unequal in consciousness
x2 proportions Survival analysis
Better 1 year survival and independence
Bias due to inefficient randomization not excluded
Sivenius J et al. 1985 (16)
Intensive in stroke unit vs conventional
Motor & ADL at 3, 6, and 12 months
95 patients selected
t-test, covariance analysis
30% higher scores at 3/12 months
Persistent effect at 3-12 months
Feldman DJ et al.
1961 (9)
et al.
1981 (14)
x2 x2
x2
Stroke rehabilitation and the end of follow-up 36% longer (i.e., 20 days longer) in the group undergoing conventional therapy, but mortality obviously tends to be highest during the three-month period immediately following acute stroke (11, 53). In a study where parametric tests were used for assessment of differences in mean ADL scores (14), significant differences in scores for activities of daily life (ADL) one year after stroke were reported for hospital-based rehabilitation (whether intensive or conventional) as against nonrehabilitation. Furthermore, better survival and independence status one year after stroke was revealed by another trial (15). Unfortunately, in this particular experiment, randomization did not achieve equal distribution of the patients as regards the level of consciousness prevailing at the time of hospitalization, with the result that the better survival rate recorded for the group treated in a stroke unit might well have been attributable, in part at least, to said group's originally enjoying a higher degree of consciousness. In a Finnish study, a comparison was drawn between patients assigned to a stroke unit where intensive rehabilitation was provided and a group treated conventionally (16). The mean score values in motor function and ADL were approximately 30% higher for the group treated in the stroke unit. Moreover, these differences still persisted to a varying extent one year after stroke. Comparative studies of various rehabilitation techniques, with the specific exclusion of those relating
to speech therapy, are set out in summarized form in Table 2. Although a significant degree of improvement from the pre-test status was observed for all groups in the four studies, studies where conventional methods were contrasted against conventional-plus-biofeedback formulae or purely perceptual-retraining schedules, this did not hold true as between the different programs themselves (17- 19) or in the follow-up stage (19). In another study, the effects of exercises consisting of resisted or ballistic finger extension and resisted grasp did not differ from those of hand inactivity (20). In a pilot trial, where physiotherapy followed the administration of a single dose of amphetamine, higher recovery of function in the very early (3 to 10-day) post-acute stroke period was recorded (22). The results of a systematic perusal and evaluation of reports on experiments focusing on aphasia rehabilitation for stroke survivors are summed up in Table 3. No differences in outcome were observed between patients undergoing two hours per week of conventional speech therapy and those receiving counseling, conversation or no therapy at all (2526). However, considerably higher functional-change scores were attained by patients receiving more intense, 8- 10 hour-per-week, individual or groupbased, traditional speech therapy, even beyond 26 weeks after acute stroke (24). Included in five other experiments designed to evaluate speech therapy in aphasia, was an intervention group which had been
Table 2. Summarized data from evaluation of specific rehabilitation techniques Author Year
Design and interventions
lnglis J
Strength Range movement Recovery levels
15/15 patients
t-tests
6 months after
Anova
1984 (17)
Biofeedbackt PT PT. Cross-over (upper limb)
Basmajian J et al. 1987 (18)
Biofeedback SPTS behavioral vs PT (5 weeks)
Upper-limb test Finger osc. test Mood scales
39 patients 0-12 months after stroke
Lincoln NB et al. 1985 (19)
Perceptual retraining vs conventional OT
Visual perception tests & ADL scale
Trombly CA et al. 1986 (20)
Specific exercises (physiotherapy) vs inactivity for finger extension
Crow JL et al. 1989 (21) Crisostomo EA et at. 1988 (22)
et al.
Measured outcome
Study population
Comparison methods
Results
Comments
Both effective Biofeedback additional effect
No differences between specific therapies
x2
Similar effect persist 9 months
No differences between specific therapies
33 patients (27 stroke)
t-test
Differences between groups non-significant
Inconclusive
Range of motion, Halsteads and other tests
23 poststroke patients
Kruskal-Wallis one-way anova
No differences statistically significant
Inconclusive
Biofeedbackt PT vs placebotPT for 6 weeks
ARA & Brunnstrom -Fugl Meyer
20120, 2-8 weeks post stroke
Stratified Mann-Whitney
Effect at end of treatment, not at follow-up
Inconclusive Inefficient randomization
PT t amphetamine vs PTtplacebo (double blind)
Fugl-Meyer days 1-3
Effect
Pharmacological treatment in stroke rehab.
test
acute stroke
414 patients 3-10 days post stroke
t-test covariance
test
Wilcoxon Rank Sum
ARA, Action Research Armtest. OT, Occupational Therapy.
435
de Pedro-Cuesta et al. Table 3. Summarized data from evaluation of aphasia rehabilitation Author Year
Design and interventions
Measured outcome
Meikle M et al. 1979 (23)
3-4 hospital-based ST sessions vs 4 weekly sessions by TV
Reached plateau PICA profile/score Personal relations
31 selected pat. 4-268 weeks after major stroke
Actuarial Wolfran test Covariance analysis
No differences
Low statistical power. Large dropout
Wertz RT et at. 1981 (24)
Individual ST vs group social & recreative (8 hours/48 weeks)
PICA, Word Fluency Measure, Token, and other scales
67 selected aged 4C-80 years, 4 weeks post-acute thromboembolic left
Variance analysis of change scores
Differences in PICA scores No significant dif. other tests
Aprox. 25% higher mean change in ST treated
Lincoln NB et al. 1984 (25)
Two hours week ST vs no therapy during 24 weeks
PICA, FCP BDAE
19 1 non-selected 10 weeks after acute stroke
t-tests for independent data
No significant diferences at 10, 22 and 32 weeks
Important drop-out Non-parametric test
Hartman J et al. 1987 (26)
Conventional ST vs counseling+ and conversation, 2 hourdweek, 6 months
PICA and 8DAE
60 right-handed one month post acute stroke
Covariance analysis of change PICA scores
No differences at 7 band 10 months
No differences
David R et al. 1982 (27)
Speech pathologist vs trained volunteer for 12 weeks
Functional Communication Profile score
155 early, medium and late postonset patients All sorts aphasia
Analysis of variance
20%-100% improvement in scores, (all groups)
Both therapies gave effect, inclusive in late referrals
Shewan CM, Kertesz A 1984 (28)
Language oriented, and stimulation by therapists or trained adult (Three one-year regimes)
WA8 and ACTS base aphasia battery
100 unilateral first stroke, in preceding weeks Controls not randomized
Covariance analysis
All three programs gave better results than those in control group
No differences among treatment variants or treatment providers
Wertz RT et al. 1986 (29)
1 =Hospital SP t NT 2=Home N t N T 3=NTtHospital SP Two-phase programs 12t 12 weeks each
PICA
12 1, selected 2-24 weeks after stroke
Covariance analysis
All effective At 12 weeks program 1 did better than 3 At end no diff.
No differences between treatment by SP or N ; hospital vs home based or deferred
Marshall RC et al. 1989 (30)
SP vs home-TV or 12 weeks deferred treatment, 8-10 hours /week, for 12 weeks
PICA, CADL RCBA
12 1 home-bound males, 2-24 weeks post acute stroke
Covariance analysis, adjusting for severity
SP treated did significantly better than not treated
No differences between treatment by SP or TV, hospital vs home based, or early vs deferred
Comparison methods
Study population
Results
Comments
ACTS, Auditory Comprehension Test for Sentences. BADE, Boston Aphasia Diagnostic Examination. CADL, Communicative Abilities in Daily Living. FCP, Functional Communicative Profile. NT, No Treatment. PICA, Porch Index of CommunicativeAbility. RCBA, Reading Comprehension Battery for Aphasia. SP, Speech Pathologist. TV, Trained Volunteer.
treated by volunteers, who in turn had been purposetrained by professional therapists to help specific patients (23, 27-30). No differences between the effect of rehabilitation environment (27, 29-30), therapeutical techniques (23,28-30), deferred treatment (29-30) or providers (23,27-30) were substantiated. In the sphere of communication scores, improvements of approximately 20% -30% ascribable to aphasia rehabilitation were documented in several of these studies, even for patients referred late after acute stroke (24, 28-30).
Comments
This review shows that: 1) differences in ADL and motor function, which originally came to light as a result ofcomparisons between rehabilitation in stroke units and non-rehabilitation or rehabilitation in medical wards, were based on relatively few quality studies and have shown themselves to be particularly inconclusive insofar as life in the home environment is concerned; 2) rehabilitation for both aphasia (8-10 hours/week) and perceptual dysfunction proved ef-
Stroke rehabilitation fective for at least several months after acute stroke; 3)in general, examination of the reports cited revealed no differences in the effects of specific rehabilitation procedures or programs when compared against their conventional/traditional counterparts; 4) aphasia rehabilitation, whether administered by specialists or amateur caregivers purpose-trained by specialists, proved equally effective; and, 5) pharmacological intervention might prove a promising avenue. Multiple factors, which may have contributed not only to the paucity but also to the somewhat erratic nature and quality of documented, controlled, randomized stroke-rehabilitation studies, have been reviewed elsewhere (57, 58). A problem common to these studies is that in the vast majority of cases the activities or techniques compared were not taskoriented, were based on hospital environment, and did not take into account individual activity-profiles relating to daily life. With regard to the matter of outcome as such, the degree to which motor and sensory scales and language tests reflect functional capacity is not well know. In some studies reviewed, the comparability of the groups covered and the informativeness of the results reported might feasibly have been undermined by: 1) difficulties in masking (9-20, 23-30); 2) inefficient randomization (15,21); 3) a sizeable drop-out factor (10, 23, 25, 26); and, 4) heterogeneity of dysfunction profiles (22, 25, 26). Discordant results, such as those reported for effects of speech therapy (24-26, 29, 30) and intensive rehabilitation (10, 16), may reflect dose-response phenomena (23-26, 29, 30) or patient selection criteria mirroring certain characteristics, e.g., the low educational level (lo), of the catchment population and/or clinical settings. Other problems present included ethical constraints, which may have limited the use of placebo-based control groups, and the fact that quality of life was not assessed. The studies on aphasia (22-30) warrant special attention since they focus on a number of aspects which could well be taken into account in the overall approach to future stroke rehabilitation. For instance, the presence of late recovery, the confusion ensuing from the level of dysfunction at entry, the relevance of psycho-social factors in neurological rehabilitation, and the insensitivity - extensively discussed elsewhere (56-57) - of scales to functional changes in stroke patients have been clearly recognized for the purposes of verbal communication (7, 28-30) and are reflected in the design of these studies. Moreover, the expanding conceptual framework for caregiving, particularly regarding provision of out-patient services to the elderly by trained amateur caregivers (58-60), was taken into account in these experiments. This research model, when read together with the results, suggest that post-acute stroke
rehabilitation could be frequently implemented and assessed within the context of the community. Grounds for such a scheme might just as easily be sought in the well-known fall-off in performance registered by aphasics in hospital-test situations, as it might be in the contrary case, where, as the records show, stroke patients often seem unable to perform ADL at home as efficiently as they do at Day Hospitals (61). When it comes to research into stroke-outcome, the importance of study design and other methodological issues has been stressed (62, 63). Furthermore, there is a body of opinion (54-56, 64) which leans towards greater emphasis being laid on the role of local conditions and personal interests - both relevant for quality of life - as well as the long-run cost of the various alternatives. In these studies, no account has been taken of either cost of treatment or care policy. However, it is precisely these aspects which underlie the design of studies targeted at intensive rehabilitation (9- 16) and aphasia (23-30), areas where a policy-study structure is sometimes to be seen, i.e., large numbers of non-selected patients, concern with global strategies, use of volunteers, etc. There is a need for well-designed studies, where the psycho-social interventive and cost-effective factors relating to stroke rehabilitation are duly taken into account. On the other had, it has been suggested that evaluative studies concerning aphasia and stroke rehabilitation should focus on specific techniques or issues and on small, selected, well-defined groups of patient's, thus gradually building up a hard core of firm fact (65). The dichotomy here might be more apparent than real. Organizational and psycho-social strategies constitute the specific elements of complex healthcare technologies in general (1) and stroke rehabilitation in particular, the effects of which have necessarily to be tested. In practice, resource reallocation and streamlined organization of service-delivery would appear to be major mechanisms of improving health care in those industrial countries bent on using programs with a proven track record (60). In such cases it would appear, paradoxically, that so-called policy-studies, difficult though they may be to implement (62), can prove the most appropriate yardstick of stroke-rehabilitation programs, where reorganization of care for stroke victims, interaction with primary care and family members, and cost-benefit issues are implicated. Bearing the above in mind, we now therefore propose that the following be incorporated as targets into research protocols for a given category of strokepatient rehabilitation programs: 1) the development of hospital/home; function-base stroke-rehabilitation programs, taking advantage of personal interest and motivation, and using amateur caregivers to provide 431
de Pedro-Cuesta et a].
guidance; 2) the development of standard methods of evaluation; 3) the tracing and pinpointing of differences in outcomes by using experimental designs; 4)the introduction of cost-effective appraisals in respect of services furnished to defined populations; and 5) the setting up a formal basis, i.e., policy evaluation, for initiating long-term changes in the provision of stroke-rehabilitation services.
4. SMITHDS, GOLDENBERG E, ASHBURNA et al. Remedial therapy after stroke: a randomised controlled trial. Br Med J 1981: 282: 517-520. 5. STEVENS RS, AMBLERNR, WARRENMD. A randomized
controlled trial of a stroke rehabilitation ward. Age Ageing 1984: 13: 65-75. 6. SIVENIUS J, PYORALAK, HEIKONEN OP, SALONENJT,
RIEKIUNEN P. The significance of rehabilitation of stroke: a controlled trial. Stroke 1985: 16: 928-931.
MW, MONGATN, SPROULE M, YOUNG 17. INGLISJ, DONALD MJ. Electromyographic biofeedback and physical therapy of the hemiplegic upper limb. Arch Phys Med Rehabil 1984: 65:
Acknowledgements To Mar Martin for secretarial assistance. To Martha Messman and Michael Benedict for help with the Enghsh. This study was supported by a grant from the Swedish Academy of Medicine (Lindgren’s Foundation) and funds from the Karolinska Institute and the Swedish Medical Research Council. Partial support for P.B. was provided by a grant from the National Institute for Disability and Rehabilitation Research.
155-159. 18. BASMAJIAN JV, GOWLAND CA, FINLAYSON MA etal.
Stroke treatment: comparison of integrated behavioralphysical therapy vs traditional physical therapy programs. Arch Phys Med Rehab 1987: 68: 267-272.
J, BHAVNANI G. 19. LINCOLNNB, WHITINGSE, COCKBURN An evaluation of perceptual retraining. Int Rehabil Med 1985: 7: 99-110. 20. TROMBLY CA, THAYER-NASON L, BLISSG, GIRARD CA, LYRISTLA, BREXA-HOOSON A. The effectiveness of ther-
References 1. BANTAHD, BEHNEYCJ, WILLEMSJS. Medical technol-
ogy: concepts and concerns. In: Toward rational technology in medicine. Considerations for health policy. New York: Springer 1981: 1-21. 2. WADEDT, LANGTONHEWERR, SKILBECK CE, DAVID
RM. Stroke: recovery and Rehabilitation after stroke. In: A 3.
4. 5.
6. 7. 8.
critical approach to diagnosis, treatment and management London: Chapman and Hall, 1985: 175-284. GRESHAM G, GRANGERCV. Overview-Patient evaluation and treatment program. In: BRANDSTATER ME, BASMAJIAN JV, eds. Stroke rehabilitation. Baltimore: Williams & Wilkins, 1987: 393-423. HACHINSKY V. Controversies in neurology: stroke Rehabilt itation. Arch Neurol 1989: 46: 703. KELLY-HAYES P et al. Time course of functional recovery following stroke: the Framingham study. J Neurol Rehab 1989: 3: 65-70. JONGBLOED L. Prediction of function after stroke: a critical review. Stroke 1986: 17: 765-776. BASSOA, CAPITANIE, VIGNOLOLA. Influence of rehabilitation on language skills in aphasic patients: a controlled study. Arch Neurol 1979: 36: 190-196. WERTZRT. Neuropathologies of speech and language: an introduction to patient management, In: JOHNSDF, ed. Clinical management of neurogenic communicative disorders. Boston: Little Brown 1978: 26.
DJ, LEE PR, UNTERECKER J, LOYDK, RUSK 9. FELDMAN A; A comparison of functionally orientated medHA, TOOLE ical care and formal rehabilitation in the management of patients with hemiplegia due to cerebrovascular disease. J Chron Dis 1961: 15: 297-310. 10. PEACOCKPB, RILEYCHP, LAMPTONTD, RAFFELss, WALKERJS. The Birmingham stroke, epidemiology and rehabilitation study. In: STEWART GT, ed. Trends in epidemiology. Springfield, 111: CC Thomas, 1972: 231-345. WM, AKHTAR AJ, PRESCOTT RJ, HOCKEYL. 11. GARRAWAY Management of acute stroke in the elderly: preliminary results of a controlled trial. Br Med J 1980: 280: 1040-1043.
WM, AKHTAR AJ, HOCKEY L, PRESCOTT RJ. 12. GARRAWAY Management of acute stroke in the elderly: follow-up of a controlled trial. Br Med J 1980: 281: 827-829. WM, SMITHDL, AKHTARAJ. 13. SMITHME, GARRAWAY Therapy impact of functional outcome in a controlled trial of stroke rehabilitation. Arch Phys Med Rehab 1982: 63: 2124.
438
apy in improving finger extension in stroke patients. Am J Occupat Ther 1986: 40: 612-617. NB, NOURIFM, DE WEERDTW. The 21. CROWJL, LINCOLN effectiveness of EMG biofeedback in the treatment of arm function after stroke. Int Disabil Studies 1989: 11: 155-160.
EA, DUNCANPW, PROPSTM, DAWSON DV, 22. CRISOSTOMO DAVISJN. Evidence that amphetamine with physical therapy promotes recovery of motor function in stroke patients. Ann Neurol 1988: 23: 94-97. E, TUPPER A et al. Comparative 23. MEIKLEM, WECHSLER trial of volunteer and professional treatments of dysphasia after stroke. Br Med J 1979: 2: 87-89. MJ, WEISSD et al. Veterans Admin24. WERTZRT, COLLINS istration cooperative study on aphasia: a comparison of individual and group treatment. J Speech Hear Res 1981: 24: 580-594. 25. LINCOLNNB, MULLEYGP, JONESAC, MCGUIRKE,
LENDREM W, MITCHELL JRA. Effectivenessof speech therapy for aphasia patients. A. A randomised controlled trial. Lancet 1984: 1: 1197-1200. 26. HARTMAN J, LANDAUWM. Comparison of formal language therapy with supportive counseling for aphasia due to acute vascular accident. Arch Neurol 1987: 44: 646-649. R, ENDERBY P, BAINTON D. Treatment of acquired 27. DAVID aphasia: speech therapists and volunteers compared. J Neurol Neurosurg Psychiatr 1982: 45: 957-961. CM, KERTESZA. Effects of speech and language 28. SHEWAN treatment on recovery from aphasia. Brain Language 1984: 23: 272-279. 29. WERTZRT, WEISS DG, ATEN JL etal. Comparison of
clinic, home, and deferred language treatment of aphasia. Arch Neurol 1986: 4: 653-658. RC, WERTZRT, WEISSD G et al. Home treat30. MARSHALL ment for aphasic patients by trained non-professionals. J Speech Hear Dis 1989: 54: 462-270. 31. BROCKLEHURST JC, ANDREWS K, RICHARDS B, LAYCOCK PJ. How much physical therapy for patients with stroke? Br Med J 1978: 1: 1307-1310. 32. LIND K. A synthesis of studies on stroke rehabilitation. J Chron Dis 1982: 35: 133-149. R, SKILBECK CE, BAINTON 33. WADED T, LANGTON-HEWER D, BRUNS-COX C. Controlled trial of home-care service for acute stroke patients. Lancet 1985: 1: 323-326. JF, DELATEUR BJ, FOWLER RS et al. Stroke: 34. LEHMANN does rehabilitation affect outcome? Arch Phys Med Rehab 1975: 56: 375-389. 35. FEIGENSON JS, GITLOW HS, GREENBERG SD. The disabil-
Stroke rehabilitation ity oriented rehabilitation unit: a major factor influencing stroke outcome. Stroke 1979: 10: 1: 5-7.
JS, MCCARTHY ML, GREENBERG SD, 36. FEIGENSON FEIGENSON WD. Factors influencing outcome and length of stay in a stroke rehabilitation unit. Part 2. Comparison of 318 screened and 248 unscreened patients. Stroke 1977: 8: 6: 657-662. 37. DICKSTEIN R, HOCHERMAN S, PILLART, SHAHAMR. Stroke rehabilitation. Three exercise therapy approaches. Phys Ther 1986: 66: 8: 1233-1238. F, MILLERJM, ROBINSON 38. HANSSTERNP, MCDOWELL M. Effects of facilitation exercise techniques in stroke rehabilitation. Arch Phys Med Rehabil 1970: 66: 527-531. 39. HAMRINE. Early activation in stroke - does it make a difference? Scand J Rehab Med 1982: 14: 102-109. 40. HAMRIN E. One year after stroke - a follow up of an experimental study. Scand J Rehab Med 1982: 14: 111-116. 41. REDINGMJ, MCDOWELLFH. Focused stroke rehabilitation programs improve outcome. Arch Neuroll989: 46: 700703. 42. DOBKIN BH. Focused stroke rehabilitation programs do not improve outcome. Arch Neurol 1989: 46: 701-703. 43. SEALEC, DAVIESP. Outcome measurement in stroke rehabilitation research. Int Disabil Studies 1987: 9: 155-160. 44. HAGENC. Communication abilities in hemiplegia: effects of speech therapy. Arch Phys Med Rehabil 1983: 64: 364-367. 45. STRANDT, ASPLUND K, ERIKSSON S, HAGGE, LITHNER F, WESTERPO. A non-intensive stroke unit reduces functional disability and the need for longterm hospitalization. Stroke 1985: 16: 1: 29-34. 46. LOGIGIAN MK, SAMUELS MA, FALCONER J, ZAGARR. Clinical exercise trial for stroke patients. Arch Phys Med Rehab 1983: 64: 364-367. 47. WAYLONIS GW, KEITHMW, ASEFFJN. Stroke rehabilitation in a midwestern county. Arch Phys Med Rehab 1973: 54: 151-155. 48. MCCANNBC, CUBERTRON RA. Comparison of two systems for stroke rehabilitation in a general hospital. J Am Geriat SOC1976: 24: 211-216. 49. GORDONWA, RUCKDESCHEL HM, EGELKOS et al. Perceptual remediation in patients with right brain damage: A comprehensive program. Arch Phys Med Rehab 1985: 66: 353-359. 50. WOLF SL, BINDER-MACLEOD S. Electromyographic bio-
feedback applications to the hemiplegic patients. Changes in lower extremity neuromuscular and functional status. Phys Ther 1983: 63: 9: 1404-1411. 5 1. WOLFSL, LECRAW DE, BARTONLA. Comparison of motor copy and targeted biofeedback training techniques for restitution of upper extremity function among patients with neurologic disorders. Phys Ther 1989: 69: 719-735. 52. BALLIETR. Neurorehabilitation of post acute brain injury: a key to optimizing functional motor recovery. In: BACH-YRITAP, ed. Traumatic brain injury. New York: Demos, 1988: 119- 131. A. Survival after stroke and transient ischemic at53. TERENT tacks during the 1970s and 1980s. Stroke 1989: 20: 13201326. 54. BOYESON MC, BACH-Y-RITA P. Determinants of brain plasticity. J Neuro Rehabil 1989: 3: 35-37. P, WICABBACH-Y-RITA E. Biological and 55. BACH-Y-RITA psychosocial factors in the recovery from brain damage in humans. Can J Psycho1 1990: 44: 148-165. 56. BACH-Y-RITA P, DE PEDRO-CUESTA J. Neuroplasticity in the aging brain: development of conceptually-based neurologic rehabilitation. In: MOLINAA et al., eds. Rehabilitation medicine. Madrid: Amsterdam: Elsevier, 1991: 5- 12. 57. LINDMARK B. Evaluation of functional capacity after stroke with special emphasis on motor function and activities of daily living. Scand J Rehab Med 1988: Suppl21: 4-40. 58. SHANAS E. The family as a social support system in old age. Gerontol 1979: 19: 169-174. 59. BASSDM, NOELKER LS. The influence of family caregivers on elder’s use of in-home services: an expanded conceptual framework. J Health SOCBehav 1987: 28: 184-196. 60. JOHANSSON LH. The new role of institutional long-term care in Sweden. Dan Med Bull 1987: Spec Suppl 5: 49-53. 61. ANDREWS K, STEWART J. Stroke recovery: he can but does he? Rheumatol Rehab 1979: 18: 43-48. 62. DOMBOVY ML, SANDOKBA, BASFORD JR. Rehabilitation for stroke: a review. Stroke 1986: 17: 363-369. GE. Stroke outcome research. Stroke 1986: 17: 63. GRESHAM 358-360. 64. RIOPELSMITHN, KIELHOFNER G, HAWKINS WATTSJ. The relationships between volition activity pattern, and life satisfaction in the elderly. Am J Occup Ther 1986: 40: 278283. 65. Editorial. Prognosis in aphasia. Lancet 1977: 2: 24.
439
Evaluation of stroke rehabilitation by randomized controlled studies: a review de Pedro-Cuesta J, WidCn-Holmqvist L, Bach-y-Rita P. Evaluation of stroke rehabilitation by randomized controlled studies: a review. Acta Neurol Scand 1992: 86: 433-439. 0 Munksgaard 1992.
In all, 22 reports of 20 randomized, controlled rehabilitation studies were evaluated. In 18 of these, the design of the trial was parallel, with a cross-over format being employed in the remaining 2 instances. Seven studies related to intensive rehabilitation during the early post-acute period. In six others, specific techniques - sometimes associated with traditional physiotherapy procedures - were compared: biofeedback, perceptual retraining, and amphetamine treatment. Eight experiments evaluated speech therapy in aphasia. Frequently, methodological considerations limited the interpretation of the results. The review showed that: 1) as regards activities of daily life and motor function, differences as between rehabilitation in stroke units on the one hand and non-rehabilitation or rehabilitation in medical wards on the other, were detected in relatively few quality studies and remained particularly inconclusive insofar as life in the home environment was concerned; 2) rehabilitation for aphasia and perceptual dysfunction proved effective for at least several months after acute stroke; 3) in general, examination of the reports cited revealed no differences in the effects of biofeedback and perceptual retraining vs conventional physiotherapy; 4) rehabilitation, whether administered by specialists or amateur caregivers purpose-trained by specialists, proved equally effective for aphasia. These conclusions constitute a valuable basis for the development and evaluation of modern rehabilitation programs for stroke patients.
While, as a medical technology, stroke rehabilitation is well defined in terms of goals, namely, restoration of brain function and adaptation, it is poorly identified in physical terms, in view of the fact that it embraces a variety of procedures, devices, and organizational and supportive systems (1-2). In affluent countries, this technology is widely available and closely tied in with hospital care of stroke patients (3). It is rather costly however. Accordingly, its true effectivenessis under debate, especially since current knowledge is mostly based on professional experience and less frequently on scientific observation (4). This paper seeks to review a number of strokerehabilitation evaluation studies and, on this basis, proceed to point the way for future research. "Spontaneous", natural recovery of brain function after acute stroke, may well overlap with recovery attributable to rehabilitation. It follows therefore that comparison with appropriate controls is required if the effects of any intervention are to be accurately assessed. Since multiple factors, such as,
J. de Pedro-Cuesta l r 3 , L. Widen-Holmqvist P. Bach-y-Rita lS4
',',
' Unit for Neuroepidemiology and Health Services Research, Huddinge University Hospital, Department of Physical Therapy, Karolinska Institute, Solna, Sweden, Department of Applied Epidemiology, National Centre for Epidemiology, "Carlos Ill" Institute of Health, Madrid, Spain, Department of Rehabilitation Medicine, University of Wisconsin, Madison, USA
Key words: Aphasia; clinical trial(s); epidemiology; randomized-trial(s); rehabilitation; stroke; technology(ies) Jesbs de Pedro-Cuesta, Department of Applied Epidemiology, National Centre for Epidemiology, "Carlos Ill" Institute of Health, c/ Sinesio Delgado, 6, 28029 Madrid, Spain Accepted for publication April 24, 1992
age, education level, time from the acute event, site and nature of lesion, highest level of disturbance of consciousness, degree of functional deficit, etc., have been shown to be related to improvement of functional capacity (5-8), we chose to restrict this review to randomized controlled experiments. Material and methods
From 68 1 publications - letters, abstracts, articles and book chapters - collected on a personal basis by the authors or identified through three Medline searches, focusing on stroke-rehabilitation from 1984-August 1991 and aphasia-therapy or aphasiarehabilitation from 1986-August 1991, we selected 44 evaluation studies relevant for our purposes (7, 9-51), with all single case studies and those where less than 80% of the patients corresponded to stroke survivors being summarily excluded. Of this total of 44,22 studies, in which a control group was included and a formal randomization procedure was de433
de Pedro-Cuesta et al. scribed (9-30), were selected for detailed examination. These 22 reports corresponded to 20 randomized, controlled experiments (9-30), while the remaining 22 dealt with uncontrolled or nonrandomized studies (7, 3 1-5 1). In the randomized controlled trials, a parallel design was to be seen in eighteen instances (9-16, 18-29) and a cross-over mode in the other two (17, 30). Seven studies involved intensive rehabilitation during the early post-acute period (9-16), regardless of whether such therapy had been provided within the formal context of a stroke unit. In six other studies, diverse but specific rehabilitation techniques, such as, biofeedback, perceptual retraining and amphetamine treatment - sometimes linked to traditional physiotherapy (PT) - were compared to conventional, mainly PT-type, procedures (17-2 1). Eight experiments were devoted to the evaluation of speech therapy in aphasia (22-30). In reviewing the randomized controlled studies (9-30) attention was systematically given to: 1) the nature, intensity and duration of the intervention, plus the time elapsed since the last acute-stroke episode; 2) the catchment population generating the patients; 3) patient selection; 4)distribution of potential confounders after randomization; 5 ) outcome measurements; 6) methods of comparison; 7) statis-
tical assessment-methods and power; and lastly, 8) interpretation of results. The terms employed by Balliet - early post-acute and late post-acute - were used to denote, in the former case, intervention during the year immediately following the acute episode and, in the latter case, intervention at any time thereafter (52). Review of methods and results of randomized controlled studies
The characteristics and results of trials, which have sought to evaluate the kind of intensive, early postacute rehabilitation frequently provided in specialized stroke units, are summarized in Table 1. Inconclusive results were reported in two early studies where random variation was not assessed (9), dropout was either important (10) or not investigated (9), and the patients were selected from among those with lower cultural and socio-economic status (10). Several analyses on one set of patients at the beginning of the 1980s uncovered temporary effects - i.e., either during the hospital stay or less than one year after stroke - of intensive rehabilitation in stroke units, coupled with a decrease in mortality (1 1-13). However, this last finding is less convincing, since not only was the time period between stroke onset
Table 1. Summarized data from evaluation of early, intensive or stroke-unit based rehabilitation programs Author Year
Design and interventions
Measured outcome
Study population
Comparison methods
Results
Comments
Rehabilitation vs no rehabilitation t medical hospital care
Neurological status and ADL
Hemiparesis up to 2 months after stroke 82, unselected
Percentages Raw data No statist. assessment
17% with no deficit in the control group
Inconclusive NR drop-out Insensitive methods
Peacock PB et al. 1972 (10)
Intensive rehabilitative vs routine care
Independence and vocational ordinal scales
52 patients (low income and cultural level)
Proportions
No differences
Inconclusive
Garraway WM et al. 1980 (1 1,121
Stroke rehab-unit vs medical wards
ADL and mortality at discharge and 1 year later
31 1 unselected conscious, hemiparetic
Proportions
Higher mortality, and dependence at discharge in wards
Biased mortality estimate Likely temporary effect of stroke unit
Smith ME et al. 1981 (17)
Stroke unit (early short) vs medical wards (late but longer)
Neurological status & ADL at discharge or 16 weeks
152 vs 109 patients (acute)
Proportions
65% ~ ~ 4 9 % independent at end of trial
Stroke Unit effective in hospital environment
Smith DS
Intensive (I) vs conventional (C) vs no rehab (NR)
ADL at 3, 6, 12 months
Acute stroke 46, 43 and 4 4 patients
Average change in ADL (t-test)
I and C better than NR
Rehabilitation I or C: effect one year later
Stevens RS et al. 1984 (15)
Stroke unit vs conventional
Complications, disability and death
112and 116 patients unequal in consciousness
x2 proportions Survival analysis
Better 1 year survival and independence
Bias due to inefficient randomization not excluded
Sivenius J et al. 1985 (16)
Intensive in stroke unit vs conventional
Motor & ADL at 3, 6, and 12 months
95 patients selected
t-test, covariance analysis
30% higher scores at 3/12 months
Persistent effect at 3-12 months
Feldman DJ et al.
1961 (9)
et al.
1981 (14)
x2 x2
x2
Stroke rehabilitation and the end of follow-up 36% longer (i.e., 20 days longer) in the group undergoing conventional therapy, but mortality obviously tends to be highest during the three-month period immediately following acute stroke (11, 53). In a study where parametric tests were used for assessment of differences in mean ADL scores (14), significant differences in scores for activities of daily life (ADL) one year after stroke were reported for hospital-based rehabilitation (whether intensive or conventional) as against nonrehabilitation. Furthermore, better survival and independence status one year after stroke was revealed by another trial (15). Unfortunately, in this particular experiment, randomization did not achieve equal distribution of the patients as regards the level of consciousness prevailing at the time of hospitalization, with the result that the better survival rate recorded for the group treated in a stroke unit might well have been attributable, in part at least, to said group's originally enjoying a higher degree of consciousness. In a Finnish study, a comparison was drawn between patients assigned to a stroke unit where intensive rehabilitation was provided and a group treated conventionally (16). The mean score values in motor function and ADL were approximately 30% higher for the group treated in the stroke unit. Moreover, these differences still persisted to a varying extent one year after stroke. Comparative studies of various rehabilitation techniques, with the specific exclusion of those relating
to speech therapy, are set out in summarized form in Table 2. Although a significant degree of improvement from the pre-test status was observed for all groups in the four studies, studies where conventional methods were contrasted against conventional-plus-biofeedback formulae or purely perceptual-retraining schedules, this did not hold true as between the different programs themselves (17- 19) or in the follow-up stage (19). In another study, the effects of exercises consisting of resisted or ballistic finger extension and resisted grasp did not differ from those of hand inactivity (20). In a pilot trial, where physiotherapy followed the administration of a single dose of amphetamine, higher recovery of function in the very early (3 to 10-day) post-acute stroke period was recorded (22). The results of a systematic perusal and evaluation of reports on experiments focusing on aphasia rehabilitation for stroke survivors are summed up in Table 3. No differences in outcome were observed between patients undergoing two hours per week of conventional speech therapy and those receiving counseling, conversation or no therapy at all (2526). However, considerably higher functional-change scores were attained by patients receiving more intense, 8- 10 hour-per-week, individual or groupbased, traditional speech therapy, even beyond 26 weeks after acute stroke (24). Included in five other experiments designed to evaluate speech therapy in aphasia, was an intervention group which had been
Table 2. Summarized data from evaluation of specific rehabilitation techniques Author Year
Design and interventions
lnglis J
Strength Range movement Recovery levels
15/15 patients
t-tests
6 months after
Anova
1984 (17)
Biofeedbackt PT PT. Cross-over (upper limb)
Basmajian J et al. 1987 (18)
Biofeedback SPTS behavioral vs PT (5 weeks)
Upper-limb test Finger osc. test Mood scales
39 patients 0-12 months after stroke
Lincoln NB et al. 1985 (19)
Perceptual retraining vs conventional OT
Visual perception tests & ADL scale
Trombly CA et al. 1986 (20)
Specific exercises (physiotherapy) vs inactivity for finger extension
Crow JL et al. 1989 (21) Crisostomo EA et at. 1988 (22)
et al.
Measured outcome
Study population
Comparison methods
Results
Comments
Both effective Biofeedback additional effect
No differences between specific therapies
x2
Similar effect persist 9 months
No differences between specific therapies
33 patients (27 stroke)
t-test
Differences between groups non-significant
Inconclusive
Range of motion, Halsteads and other tests
23 poststroke patients
Kruskal-Wallis one-way anova
No differences statistically significant
Inconclusive
Biofeedbackt PT vs placebotPT for 6 weeks
ARA & Brunnstrom -Fugl Meyer
20120, 2-8 weeks post stroke
Stratified Mann-Whitney
Effect at end of treatment, not at follow-up
Inconclusive Inefficient randomization
PT t amphetamine vs PTtplacebo (double blind)
Fugl-Meyer days 1-3
Effect
Pharmacological treatment in stroke rehab.
test
acute stroke
414 patients 3-10 days post stroke
t-test covariance
test
Wilcoxon Rank Sum
ARA, Action Research Armtest. OT, Occupational Therapy.
435
de Pedro-Cuesta et al. Table 3. Summarized data from evaluation of aphasia rehabilitation Author Year
Design and interventions
Measured outcome
Meikle M et al. 1979 (23)
3-4 hospital-based ST sessions vs 4 weekly sessions by TV
Reached plateau PICA profile/score Personal relations
31 selected pat. 4-268 weeks after major stroke
Actuarial Wolfran test Covariance analysis
No differences
Low statistical power. Large dropout
Wertz RT et at. 1981 (24)
Individual ST vs group social & recreative (8 hours/48 weeks)
PICA, Word Fluency Measure, Token, and other scales
67 selected aged 4C-80 years, 4 weeks post-acute thromboembolic left
Variance analysis of change scores
Differences in PICA scores No significant dif. other tests
Aprox. 25% higher mean change in ST treated
Lincoln NB et al. 1984 (25)
Two hours week ST vs no therapy during 24 weeks
PICA, FCP BDAE
19 1 non-selected 10 weeks after acute stroke
t-tests for independent data
No significant diferences at 10, 22 and 32 weeks
Important drop-out Non-parametric test
Hartman J et al. 1987 (26)
Conventional ST vs counseling+ and conversation, 2 hourdweek, 6 months
PICA and 8DAE
60 right-handed one month post acute stroke
Covariance analysis of change PICA scores
No differences at 7 band 10 months
No differences
David R et al. 1982 (27)
Speech pathologist vs trained volunteer for 12 weeks
Functional Communication Profile score
155 early, medium and late postonset patients All sorts aphasia
Analysis of variance
20%-100% improvement in scores, (all groups)
Both therapies gave effect, inclusive in late referrals
Shewan CM, Kertesz A 1984 (28)
Language oriented, and stimulation by therapists or trained adult (Three one-year regimes)
WA8 and ACTS base aphasia battery
100 unilateral first stroke, in preceding weeks Controls not randomized
Covariance analysis
All three programs gave better results than those in control group
No differences among treatment variants or treatment providers
Wertz RT et al. 1986 (29)
1 =Hospital SP t NT 2=Home N t N T 3=NTtHospital SP Two-phase programs 12t 12 weeks each
PICA
12 1, selected 2-24 weeks after stroke
Covariance analysis
All effective At 12 weeks program 1 did better than 3 At end no diff.
No differences between treatment by SP or N ; hospital vs home based or deferred
Marshall RC et al. 1989 (30)
SP vs home-TV or 12 weeks deferred treatment, 8-10 hours /week, for 12 weeks
PICA, CADL RCBA
12 1 home-bound males, 2-24 weeks post acute stroke
Covariance analysis, adjusting for severity
SP treated did significantly better than not treated
No differences between treatment by SP or TV, hospital vs home based, or early vs deferred
Comparison methods
Study population
Results
Comments
ACTS, Auditory Comprehension Test for Sentences. BADE, Boston Aphasia Diagnostic Examination. CADL, Communicative Abilities in Daily Living. FCP, Functional Communicative Profile. NT, No Treatment. PICA, Porch Index of CommunicativeAbility. RCBA, Reading Comprehension Battery for Aphasia. SP, Speech Pathologist. TV, Trained Volunteer.
treated by volunteers, who in turn had been purposetrained by professional therapists to help specific patients (23, 27-30). No differences between the effect of rehabilitation environment (27, 29-30), therapeutical techniques (23,28-30), deferred treatment (29-30) or providers (23,27-30) were substantiated. In the sphere of communication scores, improvements of approximately 20% -30% ascribable to aphasia rehabilitation were documented in several of these studies, even for patients referred late after acute stroke (24, 28-30).
Comments
This review shows that: 1) differences in ADL and motor function, which originally came to light as a result ofcomparisons between rehabilitation in stroke units and non-rehabilitation or rehabilitation in medical wards, were based on relatively few quality studies and have shown themselves to be particularly inconclusive insofar as life in the home environment is concerned; 2) rehabilitation for both aphasia (8-10 hours/week) and perceptual dysfunction proved ef-
Stroke rehabilitation fective for at least several months after acute stroke; 3)in general, examination of the reports cited revealed no differences in the effects of specific rehabilitation procedures or programs when compared against their conventional/traditional counterparts; 4) aphasia rehabilitation, whether administered by specialists or amateur caregivers purpose-trained by specialists, proved equally effective; and, 5) pharmacological intervention might prove a promising avenue. Multiple factors, which may have contributed not only to the paucity but also to the somewhat erratic nature and quality of documented, controlled, randomized stroke-rehabilitation studies, have been reviewed elsewhere (57, 58). A problem common to these studies is that in the vast majority of cases the activities or techniques compared were not taskoriented, were based on hospital environment, and did not take into account individual activity-profiles relating to daily life. With regard to the matter of outcome as such, the degree to which motor and sensory scales and language tests reflect functional capacity is not well know. In some studies reviewed, the comparability of the groups covered and the informativeness of the results reported might feasibly have been undermined by: 1) difficulties in masking (9-20, 23-30); 2) inefficient randomization (15,21); 3) a sizeable drop-out factor (10, 23, 25, 26); and, 4) heterogeneity of dysfunction profiles (22, 25, 26). Discordant results, such as those reported for effects of speech therapy (24-26, 29, 30) and intensive rehabilitation (10, 16), may reflect dose-response phenomena (23-26, 29, 30) or patient selection criteria mirroring certain characteristics, e.g., the low educational level (lo), of the catchment population and/or clinical settings. Other problems present included ethical constraints, which may have limited the use of placebo-based control groups, and the fact that quality of life was not assessed. The studies on aphasia (22-30) warrant special attention since they focus on a number of aspects which could well be taken into account in the overall approach to future stroke rehabilitation. For instance, the presence of late recovery, the confusion ensuing from the level of dysfunction at entry, the relevance of psycho-social factors in neurological rehabilitation, and the insensitivity - extensively discussed elsewhere (56-57) - of scales to functional changes in stroke patients have been clearly recognized for the purposes of verbal communication (7, 28-30) and are reflected in the design of these studies. Moreover, the expanding conceptual framework for caregiving, particularly regarding provision of out-patient services to the elderly by trained amateur caregivers (58-60), was taken into account in these experiments. This research model, when read together with the results, suggest that post-acute stroke
rehabilitation could be frequently implemented and assessed within the context of the community. Grounds for such a scheme might just as easily be sought in the well-known fall-off in performance registered by aphasics in hospital-test situations, as it might be in the contrary case, where, as the records show, stroke patients often seem unable to perform ADL at home as efficiently as they do at Day Hospitals (61). When it comes to research into stroke-outcome, the importance of study design and other methodological issues has been stressed (62, 63). Furthermore, there is a body of opinion (54-56, 64) which leans towards greater emphasis being laid on the role of local conditions and personal interests - both relevant for quality of life - as well as the long-run cost of the various alternatives. In these studies, no account has been taken of either cost of treatment or care policy. However, it is precisely these aspects which underlie the design of studies targeted at intensive rehabilitation (9- 16) and aphasia (23-30), areas where a policy-study structure is sometimes to be seen, i.e., large numbers of non-selected patients, concern with global strategies, use of volunteers, etc. There is a need for well-designed studies, where the psycho-social interventive and cost-effective factors relating to stroke rehabilitation are duly taken into account. On the other had, it has been suggested that evaluative studies concerning aphasia and stroke rehabilitation should focus on specific techniques or issues and on small, selected, well-defined groups of patient's, thus gradually building up a hard core of firm fact (65). The dichotomy here might be more apparent than real. Organizational and psycho-social strategies constitute the specific elements of complex healthcare technologies in general (1) and stroke rehabilitation in particular, the effects of which have necessarily to be tested. In practice, resource reallocation and streamlined organization of service-delivery would appear to be major mechanisms of improving health care in those industrial countries bent on using programs with a proven track record (60). In such cases it would appear, paradoxically, that so-called policy-studies, difficult though they may be to implement (62), can prove the most appropriate yardstick of stroke-rehabilitation programs, where reorganization of care for stroke victims, interaction with primary care and family members, and cost-benefit issues are implicated. Bearing the above in mind, we now therefore propose that the following be incorporated as targets into research protocols for a given category of strokepatient rehabilitation programs: 1) the development of hospital/home; function-base stroke-rehabilitation programs, taking advantage of personal interest and motivation, and using amateur caregivers to provide 431
de Pedro-Cuesta et a].
guidance; 2) the development of standard methods of evaluation; 3) the tracing and pinpointing of differences in outcomes by using experimental designs; 4)the introduction of cost-effective appraisals in respect of services furnished to defined populations; and 5) the setting up a formal basis, i.e., policy evaluation, for initiating long-term changes in the provision of stroke-rehabilitation services.
4. SMITHDS, GOLDENBERG E, ASHBURNA et al. Remedial therapy after stroke: a randomised controlled trial. Br Med J 1981: 282: 517-520. 5. STEVENS RS, AMBLERNR, WARRENMD. A randomized
controlled trial of a stroke rehabilitation ward. Age Ageing 1984: 13: 65-75. 6. SIVENIUS J, PYORALAK, HEIKONEN OP, SALONENJT,
RIEKIUNEN P. The significance of rehabilitation of stroke: a controlled trial. Stroke 1985: 16: 928-931.
MW, MONGATN, SPROULE M, YOUNG 17. INGLISJ, DONALD MJ. Electromyographic biofeedback and physical therapy of the hemiplegic upper limb. Arch Phys Med Rehabil 1984: 65:
Acknowledgements To Mar Martin for secretarial assistance. To Martha Messman and Michael Benedict for help with the Enghsh. This study was supported by a grant from the Swedish Academy of Medicine (Lindgren’s Foundation) and funds from the Karolinska Institute and the Swedish Medical Research Council. Partial support for P.B. was provided by a grant from the National Institute for Disability and Rehabilitation Research.
155-159. 18. BASMAJIAN JV, GOWLAND CA, FINLAYSON MA etal.
Stroke treatment: comparison of integrated behavioralphysical therapy vs traditional physical therapy programs. Arch Phys Med Rehab 1987: 68: 267-272.
J, BHAVNANI G. 19. LINCOLNNB, WHITINGSE, COCKBURN An evaluation of perceptual retraining. Int Rehabil Med 1985: 7: 99-110. 20. TROMBLY CA, THAYER-NASON L, BLISSG, GIRARD CA, LYRISTLA, BREXA-HOOSON A. The effectiveness of ther-
References 1. BANTAHD, BEHNEYCJ, WILLEMSJS. Medical technol-
ogy: concepts and concerns. In: Toward rational technology in medicine. Considerations for health policy. New York: Springer 1981: 1-21. 2. WADEDT, LANGTONHEWERR, SKILBECK CE, DAVID
RM. Stroke: recovery and Rehabilitation after stroke. In: A 3.
4. 5.
6. 7. 8.
critical approach to diagnosis, treatment and management London: Chapman and Hall, 1985: 175-284. GRESHAM G, GRANGERCV. Overview-Patient evaluation and treatment program. In: BRANDSTATER ME, BASMAJIAN JV, eds. Stroke rehabilitation. Baltimore: Williams & Wilkins, 1987: 393-423. HACHINSKY V. Controversies in neurology: stroke Rehabilt itation. Arch Neurol 1989: 46: 703. KELLY-HAYES P et al. Time course of functional recovery following stroke: the Framingham study. J Neurol Rehab 1989: 3: 65-70. JONGBLOED L. Prediction of function after stroke: a critical review. Stroke 1986: 17: 765-776. BASSOA, CAPITANIE, VIGNOLOLA. Influence of rehabilitation on language skills in aphasic patients: a controlled study. Arch Neurol 1979: 36: 190-196. WERTZRT. Neuropathologies of speech and language: an introduction to patient management, In: JOHNSDF, ed. Clinical management of neurogenic communicative disorders. Boston: Little Brown 1978: 26.
DJ, LEE PR, UNTERECKER J, LOYDK, RUSK 9. FELDMAN A; A comparison of functionally orientated medHA, TOOLE ical care and formal rehabilitation in the management of patients with hemiplegia due to cerebrovascular disease. J Chron Dis 1961: 15: 297-310. 10. PEACOCKPB, RILEYCHP, LAMPTONTD, RAFFELss, WALKERJS. The Birmingham stroke, epidemiology and rehabilitation study. In: STEWART GT, ed. Trends in epidemiology. Springfield, 111: CC Thomas, 1972: 231-345. WM, AKHTAR AJ, PRESCOTT RJ, HOCKEYL. 11. GARRAWAY Management of acute stroke in the elderly: preliminary results of a controlled trial. Br Med J 1980: 280: 1040-1043.
WM, AKHTAR AJ, HOCKEY L, PRESCOTT RJ. 12. GARRAWAY Management of acute stroke in the elderly: follow-up of a controlled trial. Br Med J 1980: 281: 827-829. WM, SMITHDL, AKHTARAJ. 13. SMITHME, GARRAWAY Therapy impact of functional outcome in a controlled trial of stroke rehabilitation. Arch Phys Med Rehab 1982: 63: 2124.
438
apy in improving finger extension in stroke patients. Am J Occupat Ther 1986: 40: 612-617. NB, NOURIFM, DE WEERDTW. The 21. CROWJL, LINCOLN effectiveness of EMG biofeedback in the treatment of arm function after stroke. Int Disabil Studies 1989: 11: 155-160.
EA, DUNCANPW, PROPSTM, DAWSON DV, 22. CRISOSTOMO DAVISJN. Evidence that amphetamine with physical therapy promotes recovery of motor function in stroke patients. Ann Neurol 1988: 23: 94-97. E, TUPPER A et al. Comparative 23. MEIKLEM, WECHSLER trial of volunteer and professional treatments of dysphasia after stroke. Br Med J 1979: 2: 87-89. MJ, WEISSD et al. Veterans Admin24. WERTZRT, COLLINS istration cooperative study on aphasia: a comparison of individual and group treatment. J Speech Hear Res 1981: 24: 580-594. 25. LINCOLNNB, MULLEYGP, JONESAC, MCGUIRKE,
LENDREM W, MITCHELL JRA. Effectivenessof speech therapy for aphasia patients. A. A randomised controlled trial. Lancet 1984: 1: 1197-1200. 26. HARTMAN J, LANDAUWM. Comparison of formal language therapy with supportive counseling for aphasia due to acute vascular accident. Arch Neurol 1987: 44: 646-649. R, ENDERBY P, BAINTON D. Treatment of acquired 27. DAVID aphasia: speech therapists and volunteers compared. J Neurol Neurosurg Psychiatr 1982: 45: 957-961. CM, KERTESZA. Effects of speech and language 28. SHEWAN treatment on recovery from aphasia. Brain Language 1984: 23: 272-279. 29. WERTZRT, WEISS DG, ATEN JL etal. Comparison of
clinic, home, and deferred language treatment of aphasia. Arch Neurol 1986: 4: 653-658. RC, WERTZRT, WEISSD G et al. Home treat30. MARSHALL ment for aphasic patients by trained non-professionals. J Speech Hear Dis 1989: 54: 462-270. 31. BROCKLEHURST JC, ANDREWS K, RICHARDS B, LAYCOCK PJ. How much physical therapy for patients with stroke? Br Med J 1978: 1: 1307-1310. 32. LIND K. A synthesis of studies on stroke rehabilitation. J Chron Dis 1982: 35: 133-149. R, SKILBECK CE, BAINTON 33. WADED T, LANGTON-HEWER D, BRUNS-COX C. Controlled trial of home-care service for acute stroke patients. Lancet 1985: 1: 323-326. JF, DELATEUR BJ, FOWLER RS et al. Stroke: 34. LEHMANN does rehabilitation affect outcome? Arch Phys Med Rehab 1975: 56: 375-389. 35. FEIGENSON JS, GITLOW HS, GREENBERG SD. The disabil-
Stroke rehabilitation ity oriented rehabilitation unit: a major factor influencing stroke outcome. Stroke 1979: 10: 1: 5-7.
JS, MCCARTHY ML, GREENBERG SD, 36. FEIGENSON FEIGENSON WD. Factors influencing outcome and length of stay in a stroke rehabilitation unit. Part 2. Comparison of 318 screened and 248 unscreened patients. Stroke 1977: 8: 6: 657-662. 37. DICKSTEIN R, HOCHERMAN S, PILLART, SHAHAMR. Stroke rehabilitation. Three exercise therapy approaches. Phys Ther 1986: 66: 8: 1233-1238. F, MILLERJM, ROBINSON 38. HANSSTERNP, MCDOWELL M. Effects of facilitation exercise techniques in stroke rehabilitation. Arch Phys Med Rehabil 1970: 66: 527-531. 39. HAMRINE. Early activation in stroke - does it make a difference? Scand J Rehab Med 1982: 14: 102-109. 40. HAMRIN E. One year after stroke - a follow up of an experimental study. Scand J Rehab Med 1982: 14: 111-116. 41. REDINGMJ, MCDOWELLFH. Focused stroke rehabilitation programs improve outcome. Arch Neuroll989: 46: 700703. 42. DOBKIN BH. Focused stroke rehabilitation programs do not improve outcome. Arch Neurol 1989: 46: 701-703. 43. SEALEC, DAVIESP. Outcome measurement in stroke rehabilitation research. Int Disabil Studies 1987: 9: 155-160. 44. HAGENC. Communication abilities in hemiplegia: effects of speech therapy. Arch Phys Med Rehabil 1983: 64: 364-367. 45. STRANDT, ASPLUND K, ERIKSSON S, HAGGE, LITHNER F, WESTERPO. A non-intensive stroke unit reduces functional disability and the need for longterm hospitalization. Stroke 1985: 16: 1: 29-34. 46. LOGIGIAN MK, SAMUELS MA, FALCONER J, ZAGARR. Clinical exercise trial for stroke patients. Arch Phys Med Rehab 1983: 64: 364-367. 47. WAYLONIS GW, KEITHMW, ASEFFJN. Stroke rehabilitation in a midwestern county. Arch Phys Med Rehab 1973: 54: 151-155. 48. MCCANNBC, CUBERTRON RA. Comparison of two systems for stroke rehabilitation in a general hospital. J Am Geriat SOC1976: 24: 211-216. 49. GORDONWA, RUCKDESCHEL HM, EGELKOS et al. Perceptual remediation in patients with right brain damage: A comprehensive program. Arch Phys Med Rehab 1985: 66: 353-359. 50. WOLF SL, BINDER-MACLEOD S. Electromyographic bio-
feedback applications to the hemiplegic patients. Changes in lower extremity neuromuscular and functional status. Phys Ther 1983: 63: 9: 1404-1411. 5 1. WOLFSL, LECRAW DE, BARTONLA. Comparison of motor copy and targeted biofeedback training techniques for restitution of upper extremity function among patients with neurologic disorders. Phys Ther 1989: 69: 719-735. 52. BALLIETR. Neurorehabilitation of post acute brain injury: a key to optimizing functional motor recovery. In: BACH-YRITAP, ed. Traumatic brain injury. New York: Demos, 1988: 119- 131. A. Survival after stroke and transient ischemic at53. TERENT tacks during the 1970s and 1980s. Stroke 1989: 20: 13201326. 54. BOYESON MC, BACH-Y-RITA P. Determinants of brain plasticity. J Neuro Rehabil 1989: 3: 35-37. P, WICABBACH-Y-RITA E. Biological and 55. BACH-Y-RITA psychosocial factors in the recovery from brain damage in humans. Can J Psycho1 1990: 44: 148-165. 56. BACH-Y-RITA P, DE PEDRO-CUESTA J. Neuroplasticity in the aging brain: development of conceptually-based neurologic rehabilitation. In: MOLINAA et al., eds. Rehabilitation medicine. Madrid: Amsterdam: Elsevier, 1991: 5- 12. 57. LINDMARK B. Evaluation of functional capacity after stroke with special emphasis on motor function and activities of daily living. Scand J Rehab Med 1988: Suppl21: 4-40. 58. SHANAS E. The family as a social support system in old age. Gerontol 1979: 19: 169-174. 59. BASSDM, NOELKER LS. The influence of family caregivers on elder’s use of in-home services: an expanded conceptual framework. J Health SOCBehav 1987: 28: 184-196. 60. JOHANSSON LH. The new role of institutional long-term care in Sweden. Dan Med Bull 1987: Spec Suppl 5: 49-53. 61. ANDREWS K, STEWART J. Stroke recovery: he can but does he? Rheumatol Rehab 1979: 18: 43-48. 62. DOMBOVY ML, SANDOKBA, BASFORD JR. Rehabilitation for stroke: a review. Stroke 1986: 17: 363-369. GE. Stroke outcome research. Stroke 1986: 17: 63. GRESHAM 358-360. 64. RIOPELSMITHN, KIELHOFNER G, HAWKINS WATTSJ. The relationships between volition activity pattern, and life satisfaction in the elderly. Am J Occup Ther 1986: 40: 278283. 65. Editorial. Prognosis in aphasia. Lancet 1977: 2: 24.
439
