An Integrated Motor Imagery Program to Improve Functional Task Performance in Neurorehabilitation: A Single-Blind Randomized Controlled Trial

O

ATRTD

Wfr9

pt

iu

K

srtb(

tus

ucR

8aist

fr

TN(ht

sz

SBm

939

RIGINAL ARTICLE

n Integrated Motor Imagery Program to Improve Functionalask Performance in Neurorehabilitation: A Single-Blindandomized Controlled Trial

hamar J. Bovend’Eerdt, PhD, Helen Dawes, PhD, Cath Sackley, PhD, Hooshang Izadi, PhD,
erick T. Wade, FRCP, MD
pfTst

t

M

FrcTeospa

imfmsndst

ts(ppwc

sps

ABSTRACT. Bovend’Eerdt TJ, Dawes H, Sackley C, Izadi H,ade DT. An integrated motor imagery program to improve

unctional task performance in neurorehabilitation: a single-blindandomized controlled trial. Arch Phys Med Rehabil 2010;91:39-46.

Objective: To investigate the feasibility of a motor imageryrogram integrated into physiotherapy and occupationalherapy.

Design: A parallel-group, phase II, assessor-blind random-zed controlled trial comparing motor imagery embedded insual therapy with usual therapy only.Setting: A neurologic rehabilitation center (Oxford, United

ingdom).Participants: Inpatients and outpatients diagnosed with

troke, brain injury, or multiple sclerosis, participating in aehabilitation program with sufficient language skills to under-ake the intervention were recruited (N�30) and assessed ataseline, after 6 weeks (postintervention), and after 12 weeksfollow-up).

Interventions: A motor imagery strategy was developedhat could be integrated into usual therapy, tailored to individ-al goals, and used for any activity. The control group receivedtandard care.

Main Outcome Measures: Goal attainment scaling wassed as the primary outcome measure. Other measures in-luded the Barthel activities of daily living index and theivermead Mobility Index.Results: Compliance with advised treatment was poor in

5% of the therapists and in 72% of the patients. Goalttainment scaling scores significantly improved at post-ntervention and follow-up (F2,27�45.159; P�.001), but noignificant difference was observed between the groups overime (F1,28�.039; P�.845).

Conclusions: Therapist and patient compliance with per-orming the intervention was low, restricting the conclusionsegarding the effectiveness of the integrated motor imagery

From the School of Life Sciences (Bovend’Eerdt, Dawes) and the School ofechnology (Izadi), Oxford Brookes University; Oxford Centre for Enablement,uffield Orthopaedic Trust (Bovend’Eerdt, Dawes, Wade); University of Oxford

Dawes), Oxford; School of Health and Population Sciences, University of Birming-am, Birmingham (Sackley), United Kingdom; and Maastricht University, Maas-richt, The Netherlands (Bovend’Eerdt).

Financial support by Joan Warren (private charity trust).No commercial party having a direct financial interest in the results of the research

upporting this article has or will confer a benefit on the authors or on any organi-ation with which the authors are associated.

Clinical trial registration number: NCT00618.Reprint requests to Thamar J. Bovend’Eerdt, PhD, Dept of Human Movement

ciences, Maastricht University Medical Centre, Universiteitssingel 50, 6229 ER, POox 616, 6200 MD, Maastricht, The Netherlands, e-mail: [email protected].

0003-9993/10/9106-00027$36.00/0doi:10.1016/j.apmr.2010.03.008

rogram. Future studies will need to explore barriers andacilitators to uptake of this intervention in clinical practice.rial recruitment and retention were good. The study demon-trated that imagery could be successfully integrated into usualherapy and tailored for a wide range of functional activities.

Key Words: Imagery (psychotherapy); Randomized con-rolled trials as topic; Rehabilitation.

© 2010 by the American Congress of Rehabilitationedicine

OR PATIENTS WITH neurologic damage or disease, suchas after a stroke or brain injury or in multiple sclerosis,

ehabilitation by a multidisciplinary team is effective,1-5 espe-ially through practicing functional task-specific activities.1,2

here is some evidence that more is better,6 but there is littlevidence to support specific techniques or components of theverall rehabilitation bundle. One potential way of practicingkills is to practice mentally using motor imagery, with theerson imagining himself undertaking the movement withoutctually doing the movement.

Recent systematic reviews7-10 have suggested that motormagery practice may be effective in enhancing motor perfor-ance in sports and in patients after stroke, but the evidence

or using it in people who present with cognitive, sensory, andotor difficulties is limited. Of the 10 studies included in the

ystematic review by Braun et al7—investigating the effective-ess of motor imagery in patients with stroke—8 used stan-ardized motor imagery tasks for all patients, whereas only 2tudies, of which 1 was a single case report, used individuallyailored motor imagery tasks.

Neurologic rehabilitation is characterized by a person-cen-ered, multidisciplinary problem-solving process, and one caneriously doubt the usefulness of training standardized taskseg, drinking from a cup) that may not be functional for everyatient. There is often a limited set of skills relevant to everyatient. Most studies focused on only a single skill in patientsith stroke, which is rarely appropriate in clinical practice and

annot be translated to clinical practice freely.Besides a rigorous experimental design, studies should en-

ure that the practiced skills can be translated into clinicalractice. Several recent studies11-14 suggest that motor imageryhould not be used alone but should be implemented in therapy

List of Abbreviations

ADLs activities of daily livingANOVA analysis of varianceRCT randomized controlled trialT1 baselineT2 postintervention
T3 follow-up
Arch Phys Med Rehabil Vol 91, June 2010

mailto:[email protected]

mailto:[email protected]

sttl

ppanatTcadc

mptauTtPstrpra2mw

R

fiwsao

P

ucptqtdi

f

I

tmw

1fipwfgtTafii

fipmtpfrcim

plswgctipw

gtoup

M

pata

Mpt

940 INTEGRATED MOTOR IMAGERY IN NEUROREHABILITATION, Bovend’Eerdt

A

essions, linking physical activities with motor imagery. Fur-her studies are needed to investigate the feasibility and effec-iveness of service-delivered motor imagery in order to trans-ate previous findings to bedside practice.

If motor imagery can be successfully embedded in clinicalractice and if it proves to be effective in the recovery ofatients with neurologic conditions, such as stroke, it could beuseful additional intervention that could be performed by

early everyone to practice almost any activity. This study usedmotor imagery strategy that can be integrated into usual

herapy, tailored to the individual, and used for any activity.he main question of this study was whether motor imageryould be successfully incorporated into standard physiotherapynd occupational therapy in people with neurologic disease oramage. Success was to be judged by (1) compliance and (2)linical effects.

METHODSThis was a phase II exploratory RCT with masked assess-ent and 2 parallel groups. Both groups received task-specific

hysiotherapy and occupational therapy provided by their ownherapists, as usual. Patients in the experimental group weredditionally taught how to incorporate motor imagery intosual therapy and into activities practiced outside therapy.his teaching was undertaken within the allocated treatment

ime (ie, there was no additional therapy contact time).atients in the control group received information on task-pecific practice and on how to undertake it outside therapyo control for attention given to the experimental group. Allehabilitation therapy was delivered by the patient’s ownhysiotherapist and occupational therapist. An independentesearcher performed masked assessments. The study was inccord with the Helsinki Declaration (revised version of000) and was approved by the Oxfordshire Ethics Com-ittee (07/H0605/84), and the trial was registered (http://ww.clinicaltrials.gov, identifier NCT00618085).

andomization and Allocation ConcealmentA series of sequentially numbered opaque envelopes were

lled with a card determining the treatment group. The allocationas determined randomly using a computer. To ensure an even

tudy entry, a block randomization sequence with blocks of 2, 4,nd 6 was used. The envelope, indicating the patient’s group, waspened by the researcher after the baseline assessment.

articipantsEach patient admitted as an inpatient to the rehabilitation

nit was initially screened by a doctor. Possible candidates whoonsented were then screened by the researcher. Patients ex-ressing an interest were given written information, after whichhey were contacted in person by the researcher to answer anyuestions and discuss procedures. Outpatients at the rehabili-ation center (Centre for Enablement, Oxford, United King-om) were recruited through referrals from consultants, phys-otherapists, or occupational therapists.

Patients were eligible for this study if they fulfilled theollowing criteria:

● Were participating in a rehabilitation program for prob-lems arising from a recent event or change in disability,secondary to disease or damage affecting the central ner-vous system (eg, stroke, traumatic brain injury, relapse inmultiple sclerosis).

● Were over 18 years of age.● Had sufficient language skills to undertake the interven-

tion, operationally defined as those able to score positive s

rch Phys Med Rehabil Vol 91, June 2010

on the first 3 items of the Sheffield screening test foracquired language disorders15 (ie, able to understand, re-member, and execute simple commands).

● Had no comorbidity that would interfere with the ability toperform imagery as judged by the clinician or from themedical notes (eg, schizophrenia).

nterventionsAll patients received standard physiotherapy and occupa-

ional therapy, and the amount and content (other than addingotor imagery in the experimental group) of interventionsere not altered.After randomization, each patient was shown 2 films (20 and

5 minutes in length, respectively) on separate occasions. Thelms were shown on a desktop or laptop computer with thearticipants either seated or lying supine in front of the screenith headphones on. There were 2 versions of each film, either

or the control group (physical practice films) or the imageryroup (motor imagery films). The films consisted of still pic-ures, commands, instructions, and exercises spoken by a man.he physical practice films consisted of general informationbout rehabilitation and motor learning. The motor imagerylms consisted of information on motor imagery and the motor

magery strategy.All the therapists received 2 training sessions of 1 hour each

rom the researcher. Therapists were taught how to implementmagery in their therapy sessions. The therapists received arinted version of the imagery strategy, a framework to deliverotor imagery for any task rather than focusing on 1 particular

ask. The researcher sat in on the first therapy session with theatient to help the therapist deliver imagery in a standardizedashion and to support the therapist with the techniques. Theesearcher also sat in on the first session of the control group toontrol for attention given to the experimental group. Allnterventions took place on the institution’s ward, in the treat-ent rooms or the gym.The therapists were asked to perform imagery with the

atients at least 3 times each week for the first 3 weeks and ateast 2 times a week for the last 2 weeks. The total amountpent on imagery during therapy sessions with the therapistsas suggested to be approximately 6.5 hours. The controlroup received standard therapy, and both groups received aomparable amount of therapy (ie, imagery replaced otherherapy). The therapy approach used at the rehabilitation centers client-centered and aimed at functional restoration and inde-endence enhancement. The average length of stay is 7 weeksith a wide range determined by many factors.In the second half of the intervention period, patients in both

roups were specifically encouraged to practice tasks beingaught in their therapy sessions for at least 5 minutes each dayutside their therapy sessions. The imagery group was asked tose imagery, and the control group was encouraged to usehysical practice. Safe tasks were chosen for practice.

easuresAn independent assessor performed assessments at baseline,

ostintervention (6 weeks), and follow-up (12 weeks). Thessessments were performed in the most convenient setting forhe patient and assessor, which included inpatient, outpatient,nd domiciliary settings.

The short Orientation-Memory-Concentration test16 and theotricity Index17 were performed at baseline to illustrate the

atient characteristics. The short Orientation-Memory-Concen-ration test comprises 6 questions and covers cognitive areas
uch as orientation, concentration on a short task, and learning
http://www.clinicaltrials.gov

http://www.clinicaltrials.gov

aTbh

HvcnralitwmtestTwftt2aeina

wi

so

tias

A

Dtdtwa

uTptTaS

dfptahcae

ea

tttsh

bpf1tddwhma

rgOaM

941INTEGRATED MOTOR IMAGERY IN NEUROREHABILITATION, Bovend’Eerdt

nd recalling some simple information (score range, 0–28).he Motricity Index is an assessment of general motor functionased on the medical research council strength grades, and theigher the score, the less motor-impaired (score range, 0–100).The primary outcome measure was goal attainment scaling.

igh levels of congruent validity (between .62 and .81) witharious outcome measures (ie, Barthel index, self-rated health,linical judgment) in a geriatric population and a physical andeurologic rehabilitation setting have been reported.18,19 Inter-ater reliability was reported as high (between .91 and .95),18-20

nd the sensitivity was also reported as high in various popu-ations.18-20 In the current study, 2 goals (at activity level) weredentified by the physiotherapist and another 2 by the occupa-ional therapist. Goals were chosen and set before the patientas allocated to a group for a period of 6 weeks. The specificethod used has been described in detail.21 Briefly, goal at-

ainment scaling was performed using 5 steps. In step 1, thexpected goals were defined by using a structured approach:pecifying the target activity, the support needed, and the timeo achieve the desired state, and quantifying the performance.hese goals were set to level 0. In step 2, the goals wereeighted for importance and difficulty, after which (step 3)

our more performance levels to each goal were specified: 2hat were better than expected (level �1 was somewhat betterhan expected; level �2 was much better than expected) and

that were worse than the goal (level –1 was the current levelt the time the goal was set; level –2 was much less than thexpected level). In step 4, goal attainment was scored by thendependent assessor at posttreatment and at follow-up. Fi-ally, in step 5, the goal attainment score was calculated bypplying the formula

GAS � 50 �10� (wixi)

��0.7� wi2 � 0.3�� wi�2�

here wi is the weight (importance�difficulty) assigned to the-th goal and xi is the numeric value achieved for the i-th goal.

To cover most domains that could be used in goal attainmentcaling, we employed a variety of standardized secondaryutcome measures:

● The Barthel ADL index, a 10-item generic ADL assess-ment with higher scores representing greater indepen-dence (score range, 0–20).22,23

● The Rivermead Mobility Index, comprising 14 questionsand 1 direct observation with higher scores representingbetter mobility (score range, 0–15).24

● Timed Up and Go, a test of functional mobility measuringthe time to stand up from a chair, walk 3m, turn, walk backto the chair, and sit down again.25

● Action Research Arm Test, an hierarchically organizedtest of arm motor function (19 items) divided in grasping,gripping, pinching, and gross movement (score range,0–57).26

● Nottingham Extended ADL scale, a well validated mea-sure of ability to perform ADL activities in the domains ofmobility, kitchen use, domestic tasks, and leisure activities(score range, 0–63).27

● A custom-developed questionnaire on the patient’s self-efficacy and perceived effort in relation to imagine per-forming tasks. The independent assessor asked the patientto imagine performing 1 of the goal tasks and then askedhow confident the patient was in imagining the task, ratingconfidence on a numeric scale between 0 (not at all con-fident) and 10 (totally confident) for each goal. The per-
ceived effort to imagine the task was investigated by w
asking the patient how much effort it cost to imagine theparticular task, again scored numerically between 0 (noeffort at all) and 10 (maximum effort).

The researcher asked the therapists (immediately after thereatment period) and the patients (at follow-up) who had usedmagery some custom-developed questions about compliancend the perceived benefits from imagery. These questions arehown later in the discussion.

nalysisGoal attainment scores were calculated for each patient.21

ata were analyzed descriptively first. Independent samples tests were performed on the demographic data and on theescriptive measures (eg, short Orientation-Memory-Concen-ration test and Motricity Index) to investigate whether thereere differences between the control group and imagery group

t baseline.A 3 � 2 (time � group) repeated-measures ANOVA was

sed to analyze the primary and secondary outcome measures.he within-subject factor “time” had 3 levels (ie, baseline,ostintervention, and follow-up), and the between-subject fac-or “group” had 2 levels (ie, control group and imagery group).he significance level for all tests was set at P less than .05, andnalyses were performed with the statistical software packagePSS v 17.0.a

An intention-to-treat analysis was performed, and missingata were imputed with the last observed response carriedorward for all measures except for the Timed Up & Go. Manyatients were not able to perform the Timed Up & Go (ie, timeo stand up from chair, walk 3m, turn, walk back to the chair,nd sit) at baseline, and the repeated-measures ANOVA cannotandle missing data. This means that only the patients thatould perform the Timed Up & Go at baseline were used in thisnalysis. A complete-case analysis employed the same repeat-d-measures ANOVA.

The success of the masked assessments was investigated bymploying a chi-square test, comparing the random allocationnd the masked assessor’s guesses.

Where the assumptions for parametric testing were violated,he nonparametric equivalent was performed. Because the Fest (eg, repeated-measures ANOVA) is usually more robusthan the nonparametric equivalent and because the tests did nothow different results, only parametric results are presentedere.

RESULTSPatients were recruited from the Oxford Centre for Enablement

etween February 14, 2008, and January 15, 2009. Figure 1resents the Consolidated Standards of Reporting Trials flowchartor this study. Of the 77 admitted patients, 7 refused to participate,

was uncooperative because of cognitive problems, 34 pa-ients did not fulfill the criteria (14 did not fulfill the criteria foriagnosis; 17 could not complete the Sheffield screening test; 2id not speak English; 1 was younger than 18y). Seven patientsere only admitted for a short stay, 1 patient was readmitted toospital before being screened, and another patient was read-itted to the rehabilitation unit with another stroke but had

lready been included in the trial.Twenty-six patients provided informed consent and were

andomized either to the imagery (n�13) or the control (n�13)roup. Four outpatients attending outpatient therapy at thexford Centre for Enablement were also included, resulting intotal of 30 patients in this study; 15 patients in each group.any patients (well over 100 patients over the time concerned)
ere treated as outpatients at the Oxford Centre for Enable-

mtmcdawvc

sb

aaIR

tt

idg

dtibutd

w(cbi

t

F intenf


A

ent, but the actual number is unknown. One patient was losto follow-up after the baseline assessment because of develop-

ent of Clostridium difficile diarrhea, and the other patientould not be contacted for the final assessment. The meanuration � SD of the intervention period was 5.3�1.5 weeksnd 5.7�1.2 weeks for the follow-up period. When patientsere discharged during the intervention period, the postinter-ention assessment was performed at discharge, which in someases resulted in a shorter intervention period than 6 weeks.

The baseline characteristics for each group and for the totalample are presented in table 1. No significant differencesetween the 2 groups were found.Table 2 presents group and combined sample data of the 3

ssessments (ie, baseline, postintervention, follow-up) for goalttainment scaling, the Barthel index, the Rivermead Mobilityndex, the Nottingham Extended ADL Index, and the Actionesearch Arm Test.Table 3 presents the answers on the compliance questions by

he therapists (at T2) and the patients (at T3). It is obvious that

ig 1. Consolidated Standards of Reporting Trials flow chart. Anorward.

he therapists used less than the proposed amount of time on i


magery (85% of the responding therapists) and that patientsid not use it as much outside their therapy sessions as sug-ested (72% of the responding patients).The reasons the therapists had not used as much imagery

uring the sessions can be categorized into practical and pa-ient-specific considerations. Practical reasons for not deliver-ng the proposed amount of imagery were that the patient waseing seen by an assistant or a student who was not trained insing imagery, outpatients received fewer therapy sessions,herapists were on vacation, or therapists simply forgot toeliver the imagery.Patient-specific reasons were that the therapist felt imagery

as inappropriate for a patient because of cognitive problemseg, attention, concentration, memory problems)—this oc-urred despite the inclusion criteria, some activities were felt toe inappropriate to practice with imagery (eg, expressive paint-ng), and some patients were uncooperative or unwell.

According to the therapists, the main limiting factor whenhe patients did not spend the proposed amount of time on

tion-to-treat analysis was performed with the last value carried

magery outside the sessions was inadequate cognitive capa-

bt

iTanttpdvl

bRttA(gior

aGm(sw

aIonba

ara27

tg5daci

pdiapsra

iee

AENT

NwtgCAOU


ilities to practice independently such as attention, concentra-ion, and memory deficits.

For the primary outcome measure (ie, goal attainment scal-ng scores), the multivariate approach was used (Pilliai’srace). Goal attainment scaling scores significantly improvedt postintervention and follow-up (F2,27�45.159; P�.001), buto significant difference was observed between the groups overime (F1,28�.039; P�.845). The interaction effect (group �ime) was not significant (F2,27�.085; P�.919). Pairwise com-arisons show that the difference over time was significantlyifferent between each assessment (ie, baseline vs postinter-ention, P�.001; postintervention vs follow-up, P�.003; base-ine vs follow-up, P�.001).

Multivariate analyses of the Barthel index, Rivermead Mo-ility Index, Nottingham Extended ADL Index, and Actionesearch Arm Test show an improvement in both groups over

ime; the within-subject factor “time” (baseline, postinterven-ion, follow-up) was significant for all outcome measures.nalyses did not show significance for the interaction factor

time � group) or the between-subject factor “group” (controlroup, imagery group), suggesting that there was no differencen outcome between the control and imagery group for theseutcome measures. A complete case analysis presents similaresults.

Table 4 presents the results for the Timed Up & Go as wells the walking aids used during the test. For the Timed Up &o, a univariate approach (sphericity assumed) was used. Theain effect of the within-subject factor “time” was significant

F2,16�11.681; P�.001). The group by time effect was notignificant (F2,16�.378; P�.691). The main effect of “group”as not significant (F1,8�1.175; P�.310).Table 5 presents group and total sample data of the 3

ssessments for the confidence and effort components of themagery Questionnaire and has similar results as the otherutcome measures. The within-subject factor “time” was sig-ificant, but the interaction term “time � group” and theetween-subject factor “group” were not, suggesting that the

Table 1: Patient Characteristics per Group and for theTotal Sample

Variable Control Experimental Total

Sex, n (women/men) 5/10 6/9 11/19Age (y) 50.6�16.48 51.2�11.75 50.3�13.88Diagnosis, n

Stroke 14 14 28TBI 1 0 1MS 0 1 1

Time since onset(wk), n�29 21.8�15.17 15.9�17.25 18.9�16.20

SOMCT 21.1�8.14 21.9�3.39 21.4�6.29n�15 n�13 n�28

Motricity IndexUL 60.3�27.75 58.1�35.19 59.2�31.16LL 54.8�29.97 53.5�26.18 54.1�27.66Total 57.5�21.65 55.8�29.02 56.7�25.17

OTE. Values are mean � SD or as indicated. The patient with MSas excluded from the calculation of the time since onset because

his was an outlier (10y); 2 patients randomized to the experimentalroup were unable to complete the short Orientation-Memory-oncentration test.bbreviations: LL, lower limb; MS, multiple sclerosis; SOMCT, shortrientation-Memory-Concentration test; TBI, traumatic brain injury;L, upper limb.

ctive group did not become better at imagining tasks.*b

The success of the masked assessments was investigatedfter each assessment and the reasons of unmasking wereecorded. Of the 29 postintervention assessments (T2), thessessor was unmasked in 3 of the assessments (10.3%). Of the8 follow-up assessments (T3), the assessor was unmasked inof them (25%). The reasons for unmasking were as follows:

● Patient revealed the allocation unintentionally● Staff revealed the allocation unintentionally● Assessor remembered the unmasking from the previous

assessmentWhen the assessor was asked to guess the patient’s alloca-

ion (for the patients who were not unmasked), 57.7% of theuesses were correct for the postintervention assessment (T2;0% correct guesses for the imagery group), not significantlyifferent from chance (�2�.62; P�.433). At the follow-upssessment (T3), 47.6% of the guesses were correct (36.4%orrect guesses for the imagery group), which was not signif-cantly different from chance (�2�1.19; P�.275).

DISCUSSIONThe study demonstrated low compliance of the therapist and

atient with performing motor imagery. As such, we cannotraw conclusions regarding the efficacy of integrating imagerynto usual therapy. Future studies will need to explore barriersnd facilitators to uptake of this intervention into clinicalractice. However, it did show that motor imagery could beuccessfully integrated into usual therapy and tailored for aange of functional activities (ie, when motor imagery wasctually employed).

Several earlier randomized controlled studies using imageryn neurologic rehabilitation have reported significant positiveffects.11,13,28-30 In this current study, we embedded the imag-ry intervention into usual therapy in order to determine its

Table 2: Results of the Primary and Secondary OutcomeMeasures (Mean � SD) on the 3 Assessments: Baseline (T1),

Postintervention (T2), and Follow-Up (T3)

Measure (group) T1 T2 T3

GAS*COEXPTOT

35.8�0.3335.9�0.5435.9�0.45

52.9�13.3153.0�7.5352.9�10.63

57.9�14.4959.3�11.4858.6�12.86

Barthel index*COEXPTOT

12.0�6.7211.9�6.7212.0�6.61

13.5�6.9813.5�6.0013.5�6.39

13.9�6.4415.1�6.2414.5�6.26

RMI*COEXPTOT

6.1�5.546.3�5.386.2�5.37

7.7�5.638.2�5.437.9�5.44

8.4�5.198.9�5.518.7�5.27

NEADL*COEXPTOT

18.3�15.9920.3�14.2619.3�14.92

23.3�17.2224.4�16.7223.9�16.69

28.2�20.0527.8�17.7028.0�18.59

ARAT*COEXPTOT

26.4�22.6926.9�24.5226.6�23.2

30.7�23.3831.7�24.9731.2�23.78

31.5�22.0532.9�25.6032.2�23.49

bbreviations: ARAT, Action Research Arm Test; CO, control group;XP, experimental group; GAS, goal attainment scaling; NEADL,ottingham Extended ADL Index; RMI, Rivermead Mobility Index;OT, total sample.
Significant within-subject factor of time (P�.05) but no significantetween-subject factor of group or interaction time � group (P�.05).

esraett

wehcspptaaptt

ipsnsasc

S

aterm.dgmamdtms

At

NtsTAwr*b


A

ffect compared with additional physical practice. In previoustudies, patients were no longer receiving therapy, patientseceived imagery as a separate addition to physical therapy,nd therapists were employed to deliver the imagery or imag-ry was delivered by tape. We have found that when attemptingo integrate imagery within clinical practice and deliver it byherapists, compliance with the intervention was poor.

Poor compliance with advised treatment is a major world-ide problem and an important modifier of health system

ffectiveness leading to poor health outcomes and increasedealth care costs.31 Various factors influence compliance, in-luding social and economic factors, the health care team/ystem, the characteristics of the disease, disease therapies, andatient-related factors.31 This study attempted to increase com-liance, in therapists as well as patients, by involving theherapists in the development of the intervention, by employingpatient-tailored intervention with individual treatment goals,

nd by using a multidisciplinary approach. Unfortunately, com-liance was still poor, which was explained by practical (eg,herapists on vacation) and patient-specific issues (eg, cogni-ive problems of the patient).

Table 3: Compliance Questions for t

Question

Have you spent the proposed amount of time on motorimagery over the past 6 weeks?

In your opinion, has the patient spent the proposed amountof time on motor imagery outside the therapy sessions?

Have you used motor imagery over the past 6 weeks?

Table 4: Results of the TUG in Seconds (Mean � SD) and theAids Used on the 3 Assessments: Baseline (T1), Postintervention

(T2), and Follow-Up (T3)

Measure(group) T1 T2 T3

TUG*CO 14.8�5.21 (n�5) 15.6�12.15 (n�7) 34.7�37.21 (n�10)EXP 12.1�3.87 (n�5) 18.4�16.07 (n�9) 13.4�9.01 (n�9)TOT 13.4�4.56 (n�10) 17.2�14.10 (n�16) 24.6�29.11 (n�9)

AidsCO Stick (n�1) Stick (n�1) Stick (n�2)

3-Rollator (n�1) Quad stick (n�1)4-Rollator (n�1)

EXP 4-Rollator (n�1) Stick (n�2) Frame (n�1)4-Rollator (n�1)

OTE. Repeated-measures ANOVA of the TUG is performed only onhe patients able to perform the test at baseline (n�10), but the tablehows the mean time in seconds � SD of all patients completing theUG over the 3 assessments.bbreviations: CO, control group; EXP, experimental group; frame,alking frame; 3-rollator, 3-wheeled rollator; 4-rollator, 4-wheeled

ollator; TOT, total sample; TUG, Timed Up & Go.
*b


It has been suggested that access to therapy is necessary butnsufficient in itself for a successful treatment. Enhancing com-liance may have a far greater impact on its success,31 so futuretudies investigating the effectiveness of motor imagery ineurorehabilitation (or any other intervention for that matter)hould increase the effectiveness of compliance first. Barriersnd facilitators to employing motor imagery in clinical practicehould be investigated, and therapists and patients should re-eive more support and training to optimize compliance.31

tudy LimitationsThe conclusions that can be drawn from this specific study

re also limited by several other factors. The number of pa-ients was too small to detect anything other than a very largeffect. The power to detect a small treatment effect in strokeehabilitation (only a small treatment effect is expected ofotor imagery), at alpha equal to .05, has been shown to be

09, suggesting that a much larger population is required toemonstrate a significant treatment effect.32 Second, we choseoal attainment scaling because it individualizes the outcomeeasured for each patient, because it is appropriate for evalu-

ting complex interventions, and because it is supposed to beore responsive than standard measures (eg, Barthel in-

ex).33,34 However, the absence of significant effects on any ofhe secondary outcome measures also suggests a lack of aajor effect. Rockwood et al33 reported that goal attainment

caling was the most responsive measure in a geriatric popu-

Table 5: Results of the Imagery Questionnaire (Mean � SD) onthe Confidence and Effort Components on the 3 Assessments:

Baseline (T1), Postintervention (T2), and Follow-Up (T3)

Measure (group) T1 T2 T3

Confidence*CO 8.1�1.87 8.6�1.59 9.0�1.06EXP 8.0�1.79 7.8�2.26 8.7�0.98TOT 8.1�1.80 8.2�1.96 8.9�1.01

Effort*CO 4.1�2.51 2.7�2.73 2.5�2.33EXP 3.8�2.28 3.1�1.77 3.1�2.35TOT 3.9�2.36 2.9�2.27 2.8�2.32

bbreviations: CO, control group; EXP, experimental group; TOT,otal sample.

herapists (T2) and the Patients (T3)

Who? Answer

Occupational therapist (n�13) Yes (n�1)No, less time (n�12)

Physiotherapist (n�14) Yes (n�3)No, less time (n�11)

Occupational therapist (n�8) Yes (n�1)No, less time (n�6)No, more time (n�1)

Physiotherapist (n�14) Yes (n�3)No, less time (n�10)No, more time (n�1)

Patient (n�12) Yes, daily (n�2)Yes, weekly (n�3)Yes, sporadically (n�5)No (n�2)

he T


lt(marpRo

tdsn

tdslmtaTp

tvurtnamthaMc

iepepvmtwi

tpatpfs

Of

1

1

1

1

1

1

1

1

1

1

2

2

2

2


ation compared with the Barthel index and Instrumental Ac-ivities of Daily Living, among others. However, our own studyTJ Bovend’Eerdt, unpublished data, 2010) suggests low agree-ent between the scoring of the independent masked assessor

nd the treating clinician and thus cannot have resulted in aesponsive measure. More detailed information regarding thesychometric properties of goal attainment scaling in blindCTs is required before it can be successfully used as anutcome measure in blind RCTs.Third, although we put considerable effort into standardizing

he training and practice of imagery, it is likely that the actualose delivered was small. A meta-analysis by Kwakkel et al35

uggests that at least 16 hours of augmented exercise therapy iseeded to find a beneficial effect.Fourth, it should be noted that imagery replaced normal

herapy in the imagery group and that overall there was littleifference in overall therapy dose or, indeed, in outcome. Thistudy did not investigate the effect of additional practice de-ivered by imagery. We had hoped that patients would practiceore, and it is possible that eventually they did, but in the short

erm this was not evident. Participants performing motor im-gery did not report practicing more than the control group.he groups appear to have performed the same amount ofractice, which may have affected outcome.This study was carried out in a specialist neurologic rehabili-

ation center that particularly admits the most complex and se-erely affected patients with any neurologic condition. The pop-lation included in this RCT was heterogeneous and may not beepresentative of most neurologic patients. This may have affectedhe uptake of the intervention because participants had complexeeds. Uptake may be better in other settings. Although the meange in our population is somewhat lower compared with otherotor imagery studies,11,28-30,36 we do not think this has nega-

ively affected our results. If anything, it would be expected toave benefited our study because imagery capacity decreases withge.37 Also, our Motricity Index scores and short Orientation-emory-Concentration test scores do not show any major con-

ern that could explain our results.

CONCLUSIONSThe novelty of this study compared with the other motor

magery studies is that the treatment involved teaching a strat-gy to the patient by therapists, rather than focusing on 1articular task. In principle, teaching a strategy should be moreffective and efficient because it can be used to help with manyroblems. However, our results suggest that we did not teach itery successfully—the patients who received training were noore confident at imagery after training than the control pa-

ients, and the therapists and patients reported not complyingith the intervention. Moreover, the benefits of strategy train-

ng may take time to accrue.The main recommendation arising from this study relates to

he observed poor compliance with imagery. Considering thereviously reported benefits of imagery and our poor compli-nce with the intervention, the techniques for training patients/herapists to use imagery in clinical practice need to be im-roved and reconsidered. Future research should establishactors affecting compliance and develop a more effectivetrategy for training patients in its use.

Acknowledgments: We thank the patients and the therapists at thexford Centre for Enablement, Charlotte Winward and Emad El-Yahya

or their help in this study, and Joan Warren for her financial support.

References1. Multiple sclerosis: management of multiple sclerosis in primary

and secondary care: clinical guideline. London: National Institutefor Clinical Excellence; 2003.

2. Rehabilitation following acquired brain injury: national clinicalguidelines. London: British Society of Rehabilitation MedicineRoyal College of Physicians; 2003.

3. National clinical guidelines for stroke. London: Royal College ofPhysicians; 2004.

4. Stokes M. Physical management of neurological rehabilitation.2nd ed. London: Elsevier Mosby; 2004.

5. Turner-Stokes L, Disler PB, Nair A, Wade DT. Multi-disciplinaryrehabilitation for acquired brain injury in adults of working age.Cochrane Database Syst Rev 2005(3):CD004170.

6. Van Peppen RP, Kwakkel G, Wood-Dauphinee S, Hendriks HJ,Van der Wees PJ, Dekker J. The impact of physical therapy onfunctional outcomes after stroke: what’s the evidence? Clin Re-habil 2004;18:833-62.

7. Braun SM, Beurskens AJ, Borm PJ, Schack T, Wade DT. Theeffects of mental practice in stroke rehabilitation: a systematicreview. Arch Phys Med Rehabil 2006;87:842-52.

8. Martin KA, Moritz SA, Hall CR. Imagery use in sport: a literaturereview and applied model. Sports Psychol 1999;13:245-68.

9. Sharma N, Pomeroy VM, Baron JC. Motor imagery: a backdoor tothe motor system after stroke? Stroke 2006;37:1941-52.

0. Zimmermann-Schlatter A, Schuster C, Puhan MA, Siekierka E,Steurer J. Efficacy of motor imagery in post-stroke rehabilitation:a systematic review. J Neuroeng Rehabil 2008;5:8.

1. Liu KP, Chan CC, Lee TM, Hui-Chan CW. Mental imagery forpromoting relearning for people after stroke: a randomized con-trolled trial. Arch Phys Med Rehabil 2004;85:1403-8.

2. Malouin F, Richards CL, Doyon J, Desrosiers J, Belleville S.Training mobility tasks after stroke with combined mental andphysical practice: a feasibility study. Neurorehabil Neural Repair2004;18:66-75.

3. Page SJ, Levine P, Leonard A. Mental practice in chronic stroke:results of a randomized, placebo-controlled trial. Stroke 2007;38:1293-7.

4. Page SJ, Levine P, Sisto SA, Johnston MV. Mental practicecombined with physical practice for upper-limb motor deficit insubacute stroke. Phys Ther 2001;81:1455-62.

5. Blake H, McKinney M, Treece K, Lee E, Lincoln NB. An eval-uation of screening measures for cognitive impairment afterstroke. Age Ageing 2002;31:451-6.

6. Wade DT, Vergis E. The short Orientation-Memory-Concentra-tion test: a study of its reliability and validity. Clin Rehabil1999;13:164-70.

7. Collen FM, Wade DT, Bradshaw CM. Mobility after stroke:reliability of measures of impairment and disability. Int DisabilStud 1990;12:6-9.

8. Joyce BM, Rockwood KJ, Matekole CC. Use of goal attainmentscaling in brain injury in a rehabilitation-hospital. Am J Phys MedRehabil 1994;73:10-4.

9. Stolee P, Stadnyk K, Myers AM, Rockwood K. An individualizedapproach to outcome measurement in geriatric rehabilitation. JGerontol A Biol Sci Med Sci 1999;54:M641-7.

0. Rockwood K, Stolee P, Fox RA. Use of goal attainment scaling inmeasuring clinically important change in the frail elderly. J ClinEpidemiol 1993;46:1113-8.

1. Bovend’Eerdt TJ, Botell RE, Wade DT. Writing SMART reha-bilitation goals and achieving goal attainment scaling: a practicalguide. Clin Rehabil 2009;23:352-61.

2. Collin C, Wade DT, Davies S, Horne V. The Barthel ADL index:a reliability study. Int Disabil Stud 1988;10:61-3.

3. Wade DT, Collin C. The Barthel ADL Index: a standard measure
of physical disability? Int Disabil Stud 1988;10:64-7.

2

2

2

2

2

2

3

3

3

3

3

3

3

3


A

4. Collen FM, Wade DT, Robb GF, Bradshaw CM. The RivermeadMobility Index: a further development of the Rivermead MotorAssessment. Int Disabil Stud 1991;13:50-4.

5. Tyson SF, DeSouza LH. Reliability and validity of functionalbalance tests post stroke. Clin Rehabil 2004;18:916-23.

6. Platz T, Pinkowski C, van Wijck F, Kim IH, di Bella P, JohnsonG. Reliability and validity of arm function assessment with stan-dardized guidelines for the Fugl-Meyer Test, Action ResearchArm Test and Box and Block Test: a multicentre study. ClinRehabil 2005;19:404-11.

7. Green J, Forster A, Young J. A test-retest reliability study of theBarthel index, the Rivermead Mobility Index, the Nottingham Ex-tended Activities of Daily Living Scale and the Frenchay ActivitiesIndex in stroke patients. Disabil Rehabil 2001;23:670-6.

8. Page SJ. Imagery improves upper extremity motor function in chronicstroke patients: a pilot study. Occup Ther J Res 2000;20:200-15.

9. Page SJ, Levine P, Leonard AC. Effects of mental practice onaffected limb use and function in chronic stroke. Arch Phys MedRehabil 2005;86:399-402.

0. Page SJ, Levine P, Sisto S, Johnston MV. A randomized efficacyand feasibility study of imagery in acute stroke. Clin Rehabil2001;15:233-40.

1. Adherence to long-term therapies: evidence for action. Geneva:
World Health Organization; 2003. a

2. Matyas TA, Ottenbacher KJ. Confounds of insensitivity and blindluck: statistical conclusion validity in stroke rehabilitation clinicaltrials. Arch Phys Med Rehabil 1993;74:559-65.

3. Rockwood K, Howlett S, Stadnyk K, Carver D, Powell C, StoleeP. Responsiveness of goal attainment scaling in a randomizedcontrolled trial of comprehensive geriatric assessment. J ClinEpidemiol 2003;56:736-43.

4. Turner-Stokes L, Williams H, Johnson J. Goal attainment scaling:does it provide added value as a person-centred measure forevaluation of outcome in neurorehabilitation following acquiredbrain injury? J Rehabil Med 2009;41:528-35.

5. Kwakkel G, van Peppen R, Wagenaar RC, et al. Effects ofaugmented exercise therapy time after stroke: a meta-analysis.Stroke 2004;35:2529-39.

6. Dijkerman HC, Ietswaart M, Johnston M, MacWalter RS. Doesmotor imagery training improve hand function in chronic strokepatients? A pilot study. Clin Rehabil 2004;18:538-49.

7. Mulder T, Hochstenbach JB, van Heuvelen MJ, den Otter AR.Motor imagery: the relation between age and imagery capacity.Hum Mov Sci 2007;26:203-11.

Supplier
. SPSS Inc, 233 S Wacker Dr, 11th Fl, Chicago, IL 60606.

Documents

An Integrated Motor Imagery Program to Improve Functional Task Performance in Neurorehabilitation: A Single-Blind Randomized Controlled Trial