Systematic literature review of treatment interventions for upper extremity hemiparesis following stroke

Transcription

1 OCCUPATIONAL THERAPY INTERNATIONAL Published online in Wiley InterScience ( Systematic literature review of treatment interventions for upper extremity hemiparesis following stroke MARJORIE L. URTON, Department of Occupational Therapy Education, Rockhurst University, Kansas City, Missouri, USA MOHAMED KOHIA, Department of Physical Therapy Education, Rockhurst University, Kansas City, Missouri, USA JANIS DAVIS, Department of Occupational Therapy Education, Rockhurst University, Kansas City, Missouri, USA MEGAN R. NEILL, Department of Occupational Therapy Education, Rockhurst University, Kansas City, Missouri, USA ABSTRACT: The primary purpose of this review article is to critically analyse the literature from 1999 to 2005 regarding effective interventions for upper extremity hemiparesis following stroke. The researchers narrowed the scope of the review based on inclusion and exclusion criteria, which yielded 11 pertinent studies congruent with the selection criteria. Studies were categorized using Sackett s levels of evidence, level I being the highest degree of certainty and level V the lowest. Grades of recommendations were then developed, grade A being highly recommended, grade B discretionary and grade C not endorsed. Two studies were endorsed as level I grade A, six were level II grade B, and three were level III grade C. Clinical recommendations inferred from the present evaluation are as follows: Electrical stimulation can be used to improve upper limb outcomes in patients with moderate to severe upper limb dysfunction and is feasible for home-based interventions. Therapy that utilizes goal-directed reaching behaviours promotes more typical reaching patterns than non-goal-directed interventions. Reach-to-grasp movements show greater improvement when compensatory trunk movements are reduced. As an addition to regular exercise therapy time, Arm BASIS training may enhance selective movements of the upper extremity (i.e. reaching).

2 12 Urton et al. When performed in conjunction with active neuromuscular stimulation, random and blocked practice may improve pre-motor, motor and total reaction times of the upper extremity. Copyright 2006 John Wiley & Sons Ltd. Key words: CVA, treatment, upper extremity, hemiparesis Introduction Cerebrovascular accident (CVA), or stroke, affects more than 700,000 Americans each year, is the third leading cause of death (National Institute of Health, 2006), and is the primary factor responsible for long-term disabilities in the United States (National Institute of Health, 2006). A prominent impairment in motor functioning associated with stroke is hemiparesis, which affects more than half of all persons who suffer stroke (Gillot et al., 2002). In addition, of those persons who suffer stroke, 69% experience functional motor impairments in the upper limbs (Luke et al., 2004) and approximately 56% continue to complain of marked hemiparesis as long as five years post-cva (Gillot et al., 2003). Considering that such losses in function can severely impact quality of life and that functional independence in numerous activities of daily living is dependent on adequate arm performance (Desrosiers et al., 2005), improvement in motor abilities, notably functional use of the upper extremity, is a primary goal of stroke rehabilitation (Luke et al., 2004; Patten et al., 2003). According to Zorowitz et al. (2002), there are a multitude of stroke rehabilitation theories and many vary in the intervention utilized to address deficits in motor functioning. Central to some theories is that use and training of the affected arm is necessary if function of the extremity is to be recovered, whereas other theories stress the use of compensatory strategies rather than trying to recover complete function (Zorowitz et al., 2002). In addition, some theories emphasize suppression of normal movement patterns in order to facilitate mass movement, such as proprioceptive neuromuscular facilitation (PNF), while others underscore the suppression of synergistic muscle patterns in order to facilitate normal movement patterns, such as neurodevelopmental training (NDT). Moreover, motor learning theory addresses modification and acquisition of new skills via practice and feedback while Rood theory promotes modification in movement patterns via the use of cutaneous sensory stimulation (Zorowitz et al., 2002). Further, whereas many theories are limited to rehabilitation of only the affected limb, a more recent concept incorporates physical constraint of the unaffected limb in order to facilitate use of the paretic limb (Rose and Winstein, 2005). Lastly, other less conventional theories of rehabilitation are also utilized in motor recovery, such as electrical stimulation (de Kroon et al., 2005), acupuncture and pharmacological interventions (Zorowitz et al., 2002). Regardless of the theoretical perspective, the ultimate goal of post-stroke rehabilitation remains consistent: assist individuals in overcoming disability due to stroke (National Institute of Neurological Disorders and Stroke, 2006). In order to do so, however, the deci-

3 Literature review of treatment interventions 13 sion must be made as to which intervention for treating upper extremity hemiparesis during rehabilitation will be most effective for the patient. The fi rst six months immediately following stroke are crucial to overcoming motor impairment in the upper extremity since regaining functional abilities occurs more rapidly during that time (de Kroon et al., 2005). However, whether in the acute phase or several years post-stroke, it is critical that effective interventions are utilized throughout the treatment process in order to achieve the highest level of functional independence for the patient. The purpose of this review is to critically analyse research literature from 1999 to 2005 that has investigated the effectiveness of clinical rehabilitative interventions for upper extremity hemiparesis following stroke. Identifying appropriate and effective clinical interventions that may contribute to restoration of upper extremity function, as identified by functional outcome measures (i.e. Wolf Motor Arm Test, Functional Independence Measure, etc.), is fundamental to the field of occupational therapy since ameliorating upper extremity function has the potential to enhance overall quality of life through improved performance and greater participation in activities of daily living. Methods Accessible databases included PubMed ( Academic Search Elite ( Academic Search Premier ( CINAHL ( and Health Source: Nursing/Academic Edition ( com), all of which were accessed through EBSCO host databases via Rockhurst University s online library. The following parameters were set for the article search: written in English, peer-reviewed, full-text accessibility, references provided, and a published date between 1999 and Relevant keywords used in the search were hemiparesis, stroke, CVA, treatment, intervention, upper extremity, and upper limb. Articles were selected based on the following inclusion criteria: persons who had previously suffered a CVA (as recent as 10 days to 15 years post-cva), had either unilateral or bilateral lesions resulting in CVA, and were undergoing rehabilitation for unilateral and/or bilateral mild to severe upper extremity hemiparesis. Articles relating to lower extremity hemiparesis and hemiparesis due to diagnoses other than CVA were excluded from this review. Articles that addressed both upper and lower extremity hemiparesis concurrently were also excluded from this review in order to maintain internal consistency and promote generalizability of the fi ndings with respect to upper extremity hemiparesis. Lastly, the cognitive, psychosocial and contextual aspects of disability and their effects on upper extremity dysfunction and performance following stroke were beyond the scope of this review and were not considered in the article selection process, in terms of functional outcomes or in the recommendations cited. The format applied in this review for critical analysis of the articles was equivalent to that incorporated in the article by Toussant and Kohia (2005)

4 14 Urton et al. regarding effectiveness of physical therapy management of hip fractures. Evaluation of the studies and their scientific rigour was based upon Sackett s five hierarchical levels of evidence and three grades of recommendations. Sackett s levels of evidence include levels I V, ranging from a high degree of certainty to a lower degree of certainty, respectively. Studies classified as level I included large, randomized controlled trials with low false positive and false negative errors. A study was considered large if groups contained 15 or more subjects. A classification of level II applied to small, randomized controlled trials with fewer than 15 subjects per group and high false positive and false negative errors. Level III studies were non-randomized, concurrent cohort comparisons between current subjects receiving treatment and those not receiving treatment. Level IV studies were non-randomized, historical cohort comparisons between current subjects receiving treatment and past subjects who had not received treatment. Finally, case studies that lacked experimental control were classified as level V (Toussant and Kohia, 2005). The level of study supported determined the assignment of grade recommendations. When supported by at least a single level I study a grade A recommendation was noted. A grade B recommendation was applied when supported by at least one level II study. The recommendation of grade C was conferred when supported by studies designated as levels III, IV or V (Toussant and Kohia, 2005). Evaluation of the scientific rigour of each study was accomplished using the following six criteria: (1) criteria were listed for inclusion and exclusion of subjects along with operational defi nitions for type of stroke and measure of hemiparesis; (2) treatment protocol was considered replicable; (3) assessment of outcome measures to establish reliability of data obtained was present; (4) examination of the validity of outcome measures was included; (5) assessors were blinded as to the allocation of treatment; and (6) all subjects initially enrolled in the study were accounted for at completion of the study. Scientific rigour for each study was analysed and a Y for yes was noted if it fulfilled the requirements set forth in the aforementioned criteria and an N for no was noted if it did not meet the requirements (Toussant and Kohia, 2005). Results The researchers originally located 24 journal articles related to treatment interventions for upper extremity hemiparesis. The number of articles to be used for the review was then narrowed by excluding literature reviews, case studies, case series, brief reports, articles unrelated to stroke, articles including or only related to lower extremity hemiparesis, articles related solely to hemiplegia, and instrumentation reliability studies. A total of 11 experimental studies that evaluated interventions for upper extremity hemiparesis following CVA were incorporated in the present review. All 11 articles were independently reviewed and classified according to Sackett s levels and are summarized in Table 1.

5 Literature review of treatment interventions 15 TABLE 1: Levels of evidence and other study characteristics Author(s) Experimental Participants Intervention Length of study design and level of evidence Outcome measures Results Cauraugh Randomized 34 stroke subjects Group A: blocked practice the 2 days of 90 minutes Manual dexterity box and With active neuromuscular and Kim controlled clinical with a diagnosis same movement was repetitively training for each of 2 block test, total reaction stimulation, subjects in the blocked (2003) trial, Level II of no more than performed on successive trials weeks with at least 24 time, premotor reaction time, and random practice groups showed two strokes, with combined with active hours of rest between motor reaction time, and improvements during the post-test an 80% upper neuromuscular stimulation. sessions; a session was sustained muscle period in that they exhibited the limit of motor Group B: random practice three sets of 30 contraction ability to move a greater number of recovery, mean different movements on successful active blocks, had faster premotor and time since stroke successive trials along with neuromuscular motor reaction times, and had less 3.2 years active stimulation. Group C: no stimulation trials with variability in the sustained muscular active stimulation assistance three movements contraction task; significant testing control group executed, 10 times/set effects were revealed with respect to total reaction time (blocked and random) and premotor reaction time (blocked unilateral and bilateral conditions; random bilateral condition only); significant limb effects in the form of improved motor reaction time from pre- to post-test were revealed for both groups (blocked bilateral and unilateral; random bilateral only)

6 16 Urton et al. TABLE 1: Continued Author(s) Experimental Participants Intervention Length of study design and level of evidence Outcome measures Results Desrosiers Randomized 41 patients Experimental group received arm 15 20, 45-minute Upper extremity motor Compared to the control group, no et al. (2005) controlled trial, admitted to therapy programme that included therapy sessions, 4 subtest of Fugl-Meyer statistically significant improvements Level I inpatient rehab repetitive unilateral and sessions per week for (motor function), Martin in upper extremity function or unit following symmetrical bilateral tasks in 5 weeks vigorimeter (grip strength), reduction in disability and stroke, subacute addition to usual arm therapy; Box and Block Test (gross impairment were revealed as a result phase, occurring control group received additional manual dexterity), Purdue of the arm training programme. at least 10 days usual arm therapy for similar Pegboard Test (fi ne manual Both groups made improvements, but not more than duration and frequency dexterity), Finger-to-Nose although not significant, suggesting 2 months earlier, Test (motor co-ordination), that the arm training programme is and having TEMPA (arm disability), no more effective than the usual minimal upper French version of the FIM therapy extremity motor (functional independence in function ADLs), and Assessment of Motor and Process Skills (changes in arm integration and use in IADLs) Gabr et al. Randomized, 12 chronic stroke Group A: received ETMS was given in 35 Fugl-Meyer (arm Neither participation in home (2005) controlled, single patients with electromyography-triggered minute increments impairment), Action exercise program nor ETMS blinded, pre-post, palpable muscle neuromuscular stimulation during an 8 week Research Arm Test (grasp, conveyed changes on the Fugl-Meyer cross-over design, contraction in (ETMS) and a home exercise period; home exercises grip, pinch and gross or Action Research Arm Test, Level II their affected wrist programme; Group B: received were performed for an movement), and goniometric however ETMS use increased active extensors but no home exercise programme 8 week period measurements (active wrist affected limb extension movement, time followed by ETMS extension) since stroke ranged from months

7 Literature review of treatment interventions 17 Author(s) Experimental Participants Intervention Length of study design and level of evidence Outcome measures Results Lewis and Non-randomized, 6 subjects, 1 Subjects completed unilateral 5 sessions per Fugl-Meyer (motor Baseline measures varied between Byblow repeated measures, month to 4 years training sessions performed with week for 4 weeks for a impairment of upper limb). subjects. Little, if any, beneficial (2004) Level III post mono- the affected limb and bilateral total of 20 training Motor performance of upper effect of bimanual practice on task hemispheric- training sessions performed sessions limb (movement performance was revealed. stroke with simultaneously with both limbs smoothness, speed, path Neurophysiological investigations impairment of in order to examine directness, intralimb were inconsistent between subjects upper limb neurophysiological and synchony, grasp quality and function relative behavioural adaptations joint angles at target). to the non- between subjects. Each subject Excitability of contralateral involved side was assigned three specific and ipsilateral motor tasks to be performed during pathways each training session for the duration of the study Michaelsen Randomized 28 subjects with Experimental group (TR) 60-trial training Temporal characteristics of The trunk restraint group had less and Levin controlled trial, arm paresis who received reach-to-grasp training period on Day 1 and arm trajectory, including anterior trunk displacement, used (2004) Level II had been with trunk restraint and the 10 trials on Day 2 to velocity and movement time, more elbow extension, and had discharged from a control group (C) received reach- measure retention trunk displacement, trunk improved interjoint co-ordination rehabilitation to-grasp training with only verbal rotation, elbow extension, following the training session; only center and were instructions not to move the shoulder flexion, shoulder the TR group maintained increases 7 94 months post- trunk during the training session horizontal adduction and in range of motion during the stroke interjoint co-ordination retention test 24 hours after training Nadeau et al. Randomized, 18 subjects, one Experimental group was given One tablet was Actual Amount of Use Test Results of the study were (2004) double-blind, or more years donepezil 5 mg and control administered daily for (AAUT), Box and Block Test inconclusive, however differential placebo-controlled, post-stroke, with group received a placebo. Both 2 weeks and then two (BBT), Wolf Motor Function gains on the WMFT, but not on parallel-group trial, moderate upper- groups instructed to take 1 tablets administered Test (WMFT) to measure other measures, for subjects taking Level II limb paresis and tablet/day for 2 weeks, then 2 daily for 4 week for a time, Fugl-Meyer Motor donepezil approached statistical undergoing tablets/day for 4 weeks. During total of 6 weeks Scale-Upper Extremity signifi cance (p =.067) constraint- last 2 weeks both groups scale, the MAL (fi ngerinduced therapy underwent constraint-induced tapping test), Stroke Impact for upper-limb therapy in conjunction with Scale (SIS) Version 2.0, dysfunction medication Caregiver Strain Index (CSI), Geriatric Depression Scale (GDS), and two probes measurements (number of cards fl ipped in 20 secs), and grip strength using a dynamometer

8 18 Urton et al. TABLE 1: Continued Author(s) Experimental Participants Intervention Length of study design and level of evidence Outcome measures Results Platz et al. Single-blind, 60 anterior Group A: no augmented exercise 4 weeks, all patients Fugl-Meyer (arm motor General effects of augmented (2005) multicentre circulation therapy time; Group B: received usual ability, sensation, and exercise therapy on recovery of randomized ischaemic stroke augmented exercise therapy standard rehabilitation passive joint motion/pain), motor control was not established, control patients with time as Bobath therapy; Group therapy, patients with Action Research Arm Test rather effects of the type of therapy trial, Level I severe (incomplete) C: augmented exercise therapy augmented exercise (grasp, grip, pinch, and supplemental to regular exercise arm paresis, 3 time as Arm BASIS training, a received an aditional gross movement), Ashworth therapy time was revealed. As weeks to 6 months repetitive, systematic technique 20 units at 45 Scale (resistance to passive augmented exercise therapy, Arm post-stroke that trains all segments of the minutes/unit movement and spasticity) BASIS training was superior to limb across the full range of Bobath therapy. Only Arm BASIS motion training, and not Bobath, as augmented therapy revealed effects. Arm BASIS training enhanced motor control (i.e. selective arm movements) but had no effect on recovery of motor function (i.e. object handling), however motor improvement was less for patients with more severe impairment and longer time post-cva. Both Arm BASIS and Bobath produced mild negative effects on passive joint motion and slightly increased joint pain, but these effects were greater for Bobath than for Arm BASIS Porter and Non-randomized, 12 volunteers All participants acted as their 14 days of c-splint use Fugl-Meyer Assesment Significant improvement in affected Lord (2004) Concurrent from Stroke own control and were required to concurrent with 10 upper extremity sections, upper limb function at 3 month cohort, pre-post Foundation of wear an upper-limb constraint (c- days of the affected Motor Assessment Scale post-treatment on the MAS and in design, Level III New Zealand, splint) that immobilzed only the upper limb exercise (MAS) upper limb sections, Grip Strength was demonstrated, who were at least wrist and fi ngers with a programme Nine-Hole Peg Test and grip but not on the Fugl-Meyer 1 year post-stroke concurrent intensive exercise strength to measure upper Assessment. Overall, upper limb prior to programme for the affected limb function and function improved despite low commencement of upper limb for 4.5 hrs per day, impairment; Modified restraint use, which suggests that the study and had consisting of 10 exercises per Ashworth Scale (muscle practice may be more crucial than an ongoing day, with 2 breaks during the tone of elbow flexors); MAL use of a restraint disability 4.5 hrs (rating scale to assess person s ability to perform functional tasks)

9 Literature review of treatment interventions 19 Author(s) Experimental Participants Intervention Length of study design and level of evidence Outcome measures Results Ring and Single-blind, 22 patients with Patients were divided into 2 sub- 6 weeks total. All Modified Ashworth (muscle Type I (no active voluntary motion) Rosenthal randomized, moderate to groups: those with no active patients in the study tone at shoulder, elbow, patients demonstrated statistically (2005) controlled severe voluntary motion at the wrist and attended their regular wrist, fi ngers and thumb), significant reductions in shoulder clinical trial, hemiparesis of fi ngers (Type I) and those with therapy sessions 3 days AROM (shoulder forward and fi nger spasticity and statistically 2 2 Factorial the upper partial active voluntary range of per week for at least fl exion/abduction, elbow and significant improvements in tests of design, Level II extremity and motion (Type II) and then 3 hrs per day. The wrist flexion/extension, and hand function (Box and Blocks and were 3 6 months assigned to the treatment or treatment group was thumb opposition), Box and Jebsen-Taylor tests). No statistically post-stroke and control group. EG: received required to wear the Blocks test and 3 Jebsen- signifi cant changes in active motion were admitted to standard treatment regime and neuroprosthetic WHO Taylor subtests requiring were revealed for this group. Type II a day hospital for standard occupational and at home for 10 minutes grasp and release to (partial active voluntary motion) outpatient physical therapy treatment 2 /day, progressing to measure functional use of patients demonstrated statistically rehabilitation modalities in addition to a wrist- up to 50 minutes 3 / the hand significant improvements in hand orthosis (WHO) equipped day over the fi rst 2 shoulder, wrist, fi ngers and thumb with a neuroprosthetic functional weeks and remained spasticity, statistically significant electrial stimulation device CG: at the latter level of gains in active motion (shoulder received only standard treatment use until the end of flexion, wrist extension and wrist regime and standard the 6th week flexion) and statistically significant occupational and physical improvements on the functional therapy treatment modalities hand tests. Of patients with pain and oedema in the neuroprosthesis groups, all showed improvement by the study s end, although changes were not statistically significant Trombly and Randomized 14 participants Experiment I: participants 10 trials under each OPTO-TRAK/3020 motion Goal-directed action produced Wu (1999) controlled trial, who were post- randomly assigned to either AB condition; total time 1 analysis system (upper significantly smoother, faster, more repeated measures, stroke, non- or BA sequence (A = goal hour extremity smoothness, forceful, and more preplanned counter-balanced hospitialized and directed action and B = rote speed, displacement, movement than did rote exercise. design, Level II were able to exercise) Experiment II: velocity and planning Functional specificity of the context reach and grasp participants randomly assigned strategy of the action); did not significantly affect with their to one of 6 sequences ABC, Ashworth Scale (spasticity); organization of the reach hemiparetic arm, BCA, CAB, ACB, BAC, and CBA Perception of Joint Position months where Condition A = natural Sense Test (sensory post-stroke context (active telephone), awareness of impaired Condition B = partial context limb); Upper Extremity (detached telephone receiver), Subtest of the Fugl-Meyer and Condition C = simulated Motor Function Test (control condition (a stick) of movement)

10 20 Urton et al. TABLE 1: Continued Author(s) Experimental Participants design and level of evidence Intervention Length of study Outcome measures Results Waller and Non-randomized, 22 subjects who Non-progressive repetitive 3 times per week for 6 Fugl-Meyer Upper Extremity No statistical differences between Whitall two groups pretest- were at least 6 bilateral arm training with weeks for a total of 18 Test (impairment measure), groups were observed at baseline. (2005) post-test, Level III months post- rhythmic auditory cueing training sessions AROM and PROM using Both groups made improvements unilateral stroke, (BATRAC) programme goniometer, Chatillion Force after BATRAC in the Fugl-Meyer all right-hand Dynamometer (arm Upper Extremity Test and the dominant, with at strength), Baseline UMAQS. Subjects with left least minimal Hydraulic Hand hemispheric lesions but not those antigravity Dynamometer (arm with right lesions made movement in the strength), Wolf Motor Arm improvements in the Wolf Motor shoulder of the Test (upper extremity Arm Test (time and weight), in paretic arm function), and University of strength measures of paretic elbow Maryland Arm flexion, shoulder extension, shoulder Questionnaire for Stroke abduction and non-paretic wrist (UMAQS) to measure daily flexion, wrist extension and shoulder use of the upper extremity abduction (fi ne and gross motor activities)

11 Literature review of treatment interventions 21 The research design, type of stroke (ischaemic vs. haemorrhagic), location of stroke (unilateral hemispheric lesions vs. bilateral hemispheric lesions), time since onset of stroke, and severity of upper extremity dysfunction varied across studies. The majority of studies included only participants who had suffered a unilateral stroke, however one study included a single subject with bilateral lesions (Trombly and Wu, 1999) and two studies did not delineate the location of the stroke (Gabr et al., 2005; Porter and Lord, 2004). Moreover, two studies directly referred to the type of stroke and included only participants who suffered an ischaemic, or non-hemorrhagic, stroke (Platz et al., 2005; Ring and Rosenthal, 2005), whereas all others were non-specific. Time since stroke varied significantly between subjects and across studies, from as recent as 10 days to 175 months post-stroke. Level of upper extremity dysfunction ranged from mild to severe deficits in motor function. The articles reviewed included eight experimental studies that utilized randomized, controlled trial design, whereas the remaining three were quasi-experimental studies that applied either the repeated measures or pre-test/post-test designs. Critical analysis of the studies revealed two studies categorized as level I, six as level II, and three as level III. All level I and level II studies included at least one experimental group and one control group. Level III studies included at least one group that acted as its own control on either a repeated measures or pre-/posttest design. Accordingly, all level I studies received a recommendation of grade A, level II studies grade B and level III studies grade C. At all levels and grades of articles reviewed, a wide spectrum of treatment interventions was studied, including electrical stimulation, exercise (goal-directed vs. rote), constraintinduced therapy, and arm training programmes. For purposes of this review, cognitive impairment secondary to stroke and relative to performance, participation and improved upper extremity function was not considered in the studies reviewed or the overall fi ndings. Studies that examined similar treatment interventions, such as constraint-induced therapy and arm training programmes, are compared herein in order to delineate conclusive and/or contradictory fi ndings. Results of studies that examined the use of constraint-induced therapy to remedy upper limb dysfunction revealed similar fi ndi ngs. Sp eci fically, Nadeau et al. (2004) and Porter and Lord (2004) both found that constraint-induced therapy was effective in the short-term, however long-term effects were inconclusive since the outcomes were not sustained longer than 2 3 months following treatment. Conversely, results of studies that examined arm-training programmes in rehabilitation of the upper extremity were contradictory in terms of functional upper extremity outcomes. One study found that Arm BASIS training, a repetitive, systematic technique that trains all segments of the limb across the full range of motion, heightens motor control (i.e. selective arm movements) but has no effect on recovery of motor function (i.e. object handling). However, such fi ndings were only significant when the Arm BASIS training was supplemental to standard treatment (Platz et al., 2005). In contrast, the study by Desrosiers et al. (2005) found that reduction of impairment, disability and functional

12 22 Urton et al. outcomes in the sub-acute phase of stroke using an arm-training programme based on repetition of unilateral and symmetrical bilateral practice was no greater than that accomplished by usual therapy. The study by Lewis and Byblow (2004), which examined the effectiveness of a short-term bilateral armtraining programme subsequent to unilateral arm-training to enhance upper limb motor performance, revealed little, if any, beneficial effects of bimanual practice on task performance of the affected upper limb. Finally, the study by Waller and Whitall (2005) found improved arm function in right-hand dominant patients with left hemisphere lesions following use of a bilateral arm training programme with rhythmic auditory cueing (BATRAC). In summation, although results of studies that examined arm training programmes yielded discrepant results, such fi ndings may be related to differences across training programmes, environmental contexts (i.e. time of day and location of testing), and/or characteristics of the participants, including time since stroke, level of upper extremity function, motivation, perception, and cognitive impairment post-cva. However, since the cognitive aspects of disability, inherent characteristics of the participants and contextual factors associated with participation are beyond the scope of this review, these factors were not considered in the recommendation process. Relative to the aforementioned studies, and in conjunction with pertinent fi ndi ngs f rom t he ot her studies reviewed, recommendations for rehabilitation of the hemiparetic upper extremity are as follows: Grade A recommendation For patients with severe paresis in the upper extremity, augmented exercise therapy time using the Arm BASIS training protocol can be used to enhance performance of selective arm movements, or upper limb motor control (Platz et al., 2005). Grade B recommendations 1. For patients with moderate to severe hemiparesis, upper limb outcomes may improve with use of daily home neuroprosthetic functional electrical stimulation supplemental to standard outpatient rehabilitation (Ring and Rosenthal, 2005). 2. Goal-directed reach (object present) enables persons with stroke to display more typical reach patterns than non goal-directed reach (no object present) (Trombly and Wu, 1999). 3. Electromyography-triggered neuromuscular stimulation (ETMS) can be used to increase active extension of the affected limb and is an attainable homebased intervention (Gabr et al., 2005). 4. For patients with chronic stroke, greater improvements in reach-to-grasp movements may be obtained by reducing compensatory trunk movements

13 Literature review of treatment interventions 23 via physical trunk restraint rather than solely practising such actions (Michaelsen and Levin, 2004). 5. Active neuromuscular stimulation as an adjunct to either blocked (same movement performed repetitively on successive trials) or random practice (performance of different movements on successive trials) can be used to improve upper extremity function (Cauraugh and Kim, 2003). Grade C recommendation Bilateral arm training can be utilized by patients with left hemispheric lesions to improve upper extremity function (Waller and Whitall, 2005). Evaluation of the scientific rigour for each study is summarized in Table 2. Only two of the 11 studies met all of the evaluation criteria for scientific rigour (Gabr et al., 2005; Platz et al., 2005). Additionally, only Gabr et al. (2005) and Platz et al. (2005) addressed both the reliability and validity of outcome measures; only one other study addressed reliability and not validity of the outcome measures (Trombly and Wu, 1999). One study received a no for replication because the article did not explicate the specific schedule of exercise tasks performed by the participants during treatment. Lastly, only half of the studies reviewed utilized blinding as a means of control. Discussion and conclusions Based on Sackett s rules for evaluating the research and the criteria for establishing scientific rigour, specific clinical implications were identified despite the broad scope of interventions included in this review. There is evidence that certain interventions for treatment of upper extremity hemiparesis following stroke are effective. However, the clinical recommendations purported in this review should be carefully appraised since the degree of hemiparesis, time since onset of stroke, location of stroke, level of upper limb motor and functional impairment, and severity of stroke varies across studies. In addition, because individual characteristics such as motivation, cognitive level and perception can affect participation and performance in the rehabilitation process, and may be adversely affected following stroke, such factors should be considered prior to implementing a treatment intervention based solely on the fi ndings of this review. Lastly, more than half of the studies were low power studies, levels II and III and grades B and C respectively, while only two were high power level I, grade A studies. Clinical recommendations inferred from the present evaluation are as follows: Electrical stimulation can be used to improve upper limb outcomes in patients with moderate to severe upper limb dysfunction and is a feasible home-based intervention.

14 24 Urton et al. TABLE 2: Evaluative Criteria for Studies Reviewed Author(s) Inclusion and Treatment Reliability of Validity of Blind Account for exclusion criteria can be outcome outcome assessment of attrition and operational replicated measures measures outcome defi nition of stroke assessed investigated Cauragh and Kim, 2003 Y Y N N N Y Desrosiers et al., 2005 Y Y N N N Y Gabr et al., 2005 Y Y Y Y Y Y Lewis and Byblow, 2004 Y Y N N N Y Michaelsen and Levin, 2004 Y Y N N N Y Nadeau et al., 2004 Y Y N N Y Y Platz et al., 2005 Y Y Y Y Y Y Porter and Lord, 2004 Y N N N N N Ring and Rosenthal, 2005 Y Y N N Y N Trombly and Wu, 1999 Y Y Y N N Y Waller and Whitall, 2005 Y Y N N Y N Y = yes; N = No

15 Literature review of treatment interventions 25 Therapy that utilizes goal-directed reaching behaviours promotes more typical reaching patterns than non-goal-directed interventions. Reach-to-grasp movements show greater improvement when compensatory trunk movements are reduced. As an addition to regular exercise therapy time, Arm BASIS training may enhance selective movements of the upper extremity (i.e. reaching). When performed in conjunction with active neuromuscular stimulation, random and blocked practice may improve pre-motor, motor and total reaction times of the upper extremity. Considering that the present review incorporated diverse studies regarding treatment for upper extremity hemiparesis, it is difficult to make defi nitive conclusions regarding the effectiveness of one particular intervention over another. In addition, long-term effects of treatment for upper extremity hemiparesis could not be established based on the limited length of intervention sessions for the studies included in this review. Moreover, the complex nature of functional upper extremity losses at different stages following stroke complicates the treatment process and limits the generalizability of research fi ndings to small, specific populations as operationally defi ned in each of the studies in this review. Lastly, the ubiquitous effects of stroke on the body and person as a whole, including global motor function, cognitive impairment, patient outlook and adjustment to disability, are important determinants of rehabilitation outcomes that must be considered in the treatment process yet were excluded from the present review. Therapists should proceed with caution when incorporating the recommendations from this review if the patient does not fall within the parameters of those identified in the studies reviewed. In the future, larger, randomized controlled trials that compare one specific intervention across multiple stages of upper extremity hemiparesis post-stroke is needed to determine at which level of disability a particular intervention is most effective. For instance, studies that examine the effects of bilateral arm training with rhythmic auditory cueing (BATRAC) on functional ability during upper extremity dressing in patients with a functional independence measure (FIM) of three would provide a more specific measure of outcomes as identified by the FIM score. Longitudinal studies that examine patients over an extended period of time, such as 3, 5 and 10 years post-intervention, would also be useful in determining whether the effects of a specific intervention are sustained over time, or whether they are merely beneficial in the short-term. Likewise, studies that examine the use of one specific intervention at different time intervals, such as 3 months, 6 months, or 1 year post-cva would provide insight into which intervention(s) would be most beneficial during a specific period of the rehabilitation process. Finally, studies that consider the cognitive aspects of participation in the treatment process, such as attention and cooperation, would also provide a more defi nitive measure of upper extremity rehabilitation potential based on the patient s willingness and motivation to

16 26 Urton et al. achieve desired functional upper extremity outcomes relative to the chosen intervention. References Cauraugh JH, Kim SB (2003). Stroke motor recovery: Active neuromuscular stimulation and repetitive practice schedules. Journal of Neurology, Neurosurgery and Psychiatry 74(11): de Kroon JR, Ijzerman MJ, Chae J, Lankhorst GJ, Zilvold G (2005). Relation between stimulation characteristics and clinical outcome in studies using electrical stimulation to improve motor control of the upper extremity in stroke. Journal of Rehabilitation 37: Desrosiers J, Bourbonnais D, Corriveau H, Gosselin S, Bravo G (2005). Effectiveness of unilateral and symmetrical bilateral task training for arm during the subacute phase after stroke: A randomized controlled trial. Clinical Rehabilitation 19: Gabr U, Levine P, Page S (2005). Home-based electromyography-triggered stimulation in chronic stroke. Clinical Rehabilitation 19 (7): Gillot AJ, Holder-Walls A, Kurtz JR, Varley NC (2003). Perceptions and experiences of two survivors of stroke who participated in constraint-induced movement therapy home programs. American Journal of Occupational Therapy 57(2): Lewis GN, Byblow WD (2004). Neurophysiological and behavioural adaptations to a bilateral training intervention in individuals following stroke. Clinical Rehabilitation 18(1): Luke C, Dodd KJ, Brock K (2004). Outcomes of the Bobath concept on upper limb recovery following stroke. Clinical Rehabilitation 18: Michaelsen SM, Levin MF (2004). Short-term effects of practice with trunk restraint on reaching movements in patients with chronic stroke: A controlled trial. Stroke 35(8): Nadeau SE, Behrman AL, Davis SE, Reid K, Wu SS, Stidham BS, Helms KM, Gonzalez-Rothi LJ (2004). Donepezil as an adjuvant to constraint-induced therapy for upper-limb dysfunction after stroke: An exploratory randomized clinical trial. Journal of Rehabilitation Research and Development 41(4): National Institute of Health (2006). Available from: aboutstroke/01.html [Accessed 15 March 2006]. National Institute of Neurological Disorders and Stroke (2006). Available from: ninds.nih.gov/ disorders/stroke/stroke.htm [Accessed 15 March 2006]. Patten C, Kothari D, Whitney J, Lexell J, Lum PS (2003). Reliability and responsiveness of elbow trajectory tracking in chronic poststroke hemiparesis. Journal of Rehabilitation Research and Development 40(6): Platz T, Eickhof C, van Kaick S, Engel U, Pinkowski C, Kalok S, Pause M (2005). Impairmentoriented training or Bobath therapy for severe arm paresis after stroke: A single blind, multicentre randomized controlled trial. Clinical Rehabilitation 19(7): Porter K, Lord S (2004). Constraint-induced movement therapy for people following stroke in an outpatient setting. Journal of Physiotherapy 32(3): Ring H, Rosenthal N (2005). Controlled study of neuroprosthetic functional electrical stimulation in sub-acute post-stroke rehabilitation. Journal of Rehabilitation Medicine 37(1): Rose DK, Winstein CJ (2005). The co-ordination of bimanual rapid aiming movements following stroke. Clinical Rehabilitation 19(4): Toussant EM, Kohia M (2005). A critical review of literature regarding the effectiveness of physical therapy management of hip fracture in elderly persons. Journal of Geriontology 60A(10):

17 Literature review of treatment interventions 27 Trombly CA, Wu C (1999). Effect of rehabilitation tasks on organization of movement after stroke. American Journal of Occupational Therapy 53(4): Waller SM, Whitall J (2005). Hand dominance and side of stroke affect rehabilitation in chronic stroke. Clinical Rehabilitation 19: Zorowitz RD, Gross E, Polinski DM (2002). The stroke survivor. Disability and Rehabilitation 24(13): Address correspondence to Mohamed Kohia, PT, PhD, Rockhurst University, 1100 Rockhurst Road, Kansas City, Missouri Tel: (816) ; Fax: (816) Mohamed. Kohia@rockhurst.edu