AUSTRAliAN PHYSIOTHERAPY

Tim EsmondeJennifercGjn~ey

Joanne ittwerPatricia GoldieClarissa artin

ORIGINAl ARTICLE

Strokerehabilitation:patient activity duringnon-therapy time

Since practice is a crucial variable in motorlearning, this study aimed to quantify the levelof motor activity during inpatient strokerehabilitation andtoinvestigatethe relationshipbetween level ofmotor abi Iitywith performanceof motor activity. Seventeen stroke patientswere observed on nine weekdays in a fast­stream rehabilitation hospital. The resultsshowed that for two-thirds of the day, patientswerenot involved in structured therapyand thatfor half of these observations, the patientswere not engaged in motor activities. Therewas apositive correlation (rho =0.68, P<0.05)betweenpatients' performance ofmotaractivityand atotal Functional Independence Measurescore. It appears thatstrategiesare required sothat, regardless ofthe levelofdisability, patientscan be practising at an appropr~ate level. Ads[Esmonde TJ, McGinley JL, Wittwer JE, GoldiePA and Martin CL: Stroke rehabilitation: patientactivity during non-therapy time. AustralianJournal ofPhysiotheray43: 43-51]

Key words: CerebrovascularDisorders; Motor Activity;Rehabilitation

TJ Esmonde BPhty(Hons), wasformerly aGradeOne physiotherapist at The Royal MelbourneHospital.JL McGinley BAppSc, GradDipNeurosciences,is a postgraduate student at the School ofPhYSiotherapy, La Trobe University, Melbourne

I t has been estimated that 37,000Australians suffer a stroke each year(Anderson et al 1992), accounting

for one ofthe most frequent causes ofhospital admission (Christie 1981).Since it appears that a large proportion(40 per cent) of patients requirespecialised inpatient rehabilitation(Shah 1989), with an average length ofstay in rehabilitation of approximatelytwo months (Dean and Mackey 1992,Shah 1989), the cost to the Australianhealth system of rehabilitationfollowing stroke is substantiaL Theaims ofrehabilitation are best achievedby a multi-disciplinary te"am utilisingthe resources ofseveral healthprofessions (Dombovy et a11986,Feigensonet al 1979). With the highcost of providing such an intensive

~ rehabilitation program, it is crucial thatoptimal conditions are provided forpatients to recover from the disablingconsequences of stroke.

Within the team approach torehabilitation following stroke, theprimary aim of physiotherapy is torehabilitate motor function. Whilevarious methods exist (Ernst 1990),

JE Wittwer, BAppSc, Grad Dip Phty, is anassociate lecturer in the School of Physiotherapy,La Trobe University, MelbournePA GoldieMAppSc, is a senior lecturer in theSchool of Physiotherapy, La Trobe University,Melbourne.

one of the most frequently taughtapproaches in Australia is the MotorRelearning Program (Carr et al 1994).This is an example of a physiotherapyintervention program that has beenbased on principles of motor learning,amongst which practice isa keyvariable. Adaet al (1990) haveemphasised that the rehabilitationenvironment should .be structured topromote independent practice ofrelevant motor activities in order tomaximise benefits during this crucialpost-stroke period.

Although it is expected that anoptimal learning environment shouldexist for patients undertakingrehabilitation following a stroke tofacilitate independent practice ofmotor activities during unstructuredtherapy time. However, very fewAustralian data are available to describehow patients spend their days while inrehabilitation. One recent Australianstudy (Mackey et al 1996) has shownthat patients admitted to rehabilitationfollowing stroke spent less than 20 percent of the day engaged in motor -GL Martin BPhty(Hons), is a Grade 1physiotherapist at the Hampton RehabilitationHospital, Melbourne.Gorrespondence:Mrs P Goldie, School ofPhysiotherapy, Faculty of Health Sciences, LaTrobe University, 625 Swanston Street, CarltonVictoria 3053.

from Pagebehaviour during the weekdays andeven less on the weekends. Studiesfrom other countries have providedsimilar disconcerting evidence whichshows that patients werepredominantly inactive out ofstructured therapy time (Keith 1980,1986 and 1988, Keith and Cowell1987, Lincoln et a11989, Tinson1989). For example, Keith and Cowell(1987) reported that for 63 inpatientswith stroke across three rehabilitationhospitals, 31 per cent of the day wasspent in treatment and 42 per cent ofthe day in passive or inactivebehaviour. Lincoln and colleagues(1989) studied two groups of patientsat five month intervals and found that47 to 59 per cent of the day was spentin inactive or non-task directedactivities. The aim of this study was toinvestigate whether the lack of motoractivity·reported by previous studies(Keith 1980, 1986 and1988,Keith andCowell 1987, Lincolnet all989,Mackey et al 1996, Tinson 1989) wasrepresentative of a stroke rehabilitationunit in the·state ofVictoria. Since datafrom the Australian health system arelimited to only one study, a furtherstudy seemed justified to investigatecurrent practice.

It is likely that patients will differ inthe amount of time they spendpractising motor activities, includingboth functional movements andspecific exercises, in non-therapy time.A review of the literature revealed thatlittle is lmown about the individualdifferences between patients in theamount of independent practiceperformed during unstructured .rehabilitation time. One factor whIchmay influence the amount ofindependent practice is the level ofmotor performance or functionalstatus. It is probable that a patient withhigher functional ability has greaterability to perform motor activitiesduring non-therapy time. Since thisissue has not previously beenaddressed, this study also investigatedthe relationship between patients'assessed level·of motor function andtheir observed performance of motoractivity during non-therapy time.

o RIG I NA 1 ART Ie lE

In summary, the purpose of this studywas threefold:(1) to obtain data to describe the

distribution of time inpatients withstroke spend in and out of therapy;

(2) to quantify the amount ofmotoractivity and the specific motoractivities performed duringunstructured therapy time; and

(3) to determine the relationshipbetween patients' level of motorability'with their performance ofmotor activity. It was hypothesisedthat there would be a positivecorrelation between motor statusand observed level of motoractivity.

MethodSubjectsThis study was conducted at a majorrehabilitation hospital in Melbourneduring two separate one-week periods,one month apart, in 1994. Patientswere referred to the rehabilitationhospital from various acute hospitalswithin Melbourne. The study sampleconsisted of those inpatients in therehabilitation hospital who had adiagnosis of stroke and had ,beenreferred to physiotherapy and gaveinformed consent to participate in thestudy.In total there were 17 patients (10

males and seven females) observed overtli.e two combined observation periods.This included nine patients whoparticipated in both observationperiods. The mean (SD) age of the 17patients was 55.4 (13.7) years (range 26to 82 years). The median time fromstroke onset to the beginning of thestudy was 63 days (interquartile range[IQR] 41~3 to 100.5 days) and themedian time from admission inrehabilitation to the beginning of thestudy was 54 days (IQR26 to 67.3days). See Table 1 for characteristics ofthe patient sample.

SettingThe fast stream rehabilitation hospitalis an independent single-storeystructure with two separate 24 bedwards divided by a 50 metre corridor.During the unstructured part ofthe

AUSTRAliAN PHYSIOTHERAPY

day when no therapy was scheduled,patients could be ina number of areas,including outdoor verandahs andgardens, one of three recreationalrooms, a day room, bedrooms andbathrooms. All patients ate together inthe main dining room attached to oneofthe wards. Physiotherapy andoccupational therapy areas werelocated at the far end of the hospitalaway from the ward.areas and speechtherapy, nutrition, neuropsychologyand social work were located withinthe office area between the two wards.Some therapists treated patients on theward but most ofthe therapy wasperformed in the designated therapyareas. Most patients had to betransported to each therapy area fortheir scheduled sessions. During allbreaks the therapy areas wereunsupervised and thus closed.

ApparatusA recording form was used todocument observations of patientbehaviour and location. This form,which was based largely on a methoddesigned by Mackey (1996), wasmodified to include informationregarding patient posture when theywere observed to be inactive.Modifications were also made to suitthe specific environment of thehospital in which the study wasperformed.

ProcedureThis study was approved by both LaTrobe University and the hospital'sethics committee and administration.With the exception of one treatingphysiotherapist, the hospital employeeswere unaware of the specific purposeof this research. Patients were·awarethat the research was to record howmuch time they spent in therapy andwhat sort of activities they performedout of therapy. The observer was awareof the aims of the study and wasunknown to participating patients.There was no attempt made to concealthe intent of the research but theobserver tried to remain as unobtrusiveas possible.

A time-sampling technique ofbehavioural mapping was used toobserve patients' location and motor

AUSTRAliAN PHYSIOTHERAPY ORI GIN A l ART I C lE

Type ofstroke:Haemorrhage 5 30Infarct 12 70Site ofStroke~·Cortical 11 65Subcortical 4 24Brainstem 1 5.5Cerebellar 1 5.5Past MedicalHistory:No previous stroke 15 88Previous stroke 2 12Side ofMotor Deficit:Left 11 65Right 6 35Side ofMotor Dominance:Left 3 18Right 14 82Affected side motor dominant:Yes 7 41No 10 59

* Percentage values rounded.

18.7

11.8

2.00.4

0.2

< 041

1003521

F == numberofobservations

TOTAL

LOCATION F

Recreational Room 736

Bedroom 520

Corridor 424

Dining Room 335

Outdoors 204

ToiletIBathroom 136

NON-THERAPY 2lSS

Physiotherapy 659Occupational Therapy 414

Speech.Therapy 71

Social Work 13

Neuropsychology 7

Nutrition 2

THERAPY 1166

Table 2.Location of patients during both observation periods.

Frequency Percentage*(N= .• 17)

Characteristic

Table 1.Characteristics of patients.

behaviours (Keith 1988, Lincoln etal1989, Tinson J 989). Potentially therewere 10 therapy and nine non-therapyareas where patients could be locatedwithin the hospital. Motor behaviourwas documented as either active (iemotor activity observed) or inactive (ieno motor activity observed). "Wheninactive, behaviour was categorised aseither passive or·watching (activelyengaged in observing theenvironment). Active motor behaviourwas categorised according to use of theaffected or unaffected upper limb,affected or unaffected lower limb, roll­to-sit, sitting unsupported, sit-to­stand, standing, walking andwheelchair propulsion. Patients wereclassified as sitting unsupported if they

were maintaining an active balancedposition without back support. Theseactivities were not mutually exclusive,as patients could be recordeddisplaying two motor behaviourssimultaneously (eg sitting unsupportedand using the unaffected upper limb orwalking and using the affected upperlimb).

The reliability of using the form wasestablished in· a pilot study by havingtwo observers record the sameobservations fora group of 11 patientsprior to and during the formal study(Esmonde 1994). The coefficient ofagreement was very high for presenceof motor behaviour (K = 048) andlocation (K =0.99).

Observations were made between 9

am and 5 pm, the normal hours of thehospital's rehabilitation program, for atotal of nine weekdays. One observermade all recordings and, to address theissue of observer fatigue, four breaktimes were allocatedwmchdid notcoincide with patients' rest times.Thus, observations were made for sixand a half hours on each day.To commence a sample, the observer

recorded the time and followed apredetermined observation routearound the hospital to ensure allpotential areas of the hospital werecovered. At the first instant theobserver saw a participating patient,the location was recorded (therapy ornon-therapy) and details were recorded

~

figure 1.Proportion of total time spent in therapy and non-therapy activities from a total of 3943observationsreco~·ded over tWQsampling periods.

from Pageas shown in Table 2. VYhen patientswere in therapy, no other details aboutmotor activity were recorded. VVhenpatients were not in therapy, thepresence or absence of motorbehaviours was recorded and thespecific activity or posture,respectively. The observation routewas completed by returning to the startand recording the time to complete thesample. It took approximately 10minutes to complete the route throughthe hospital, however the length oftime to complete one sample wasgenerally determined by the patients'locations throughout the hospital.Samples recommenced after a break ofup to eight minutes. Thus, there wasno set time between each observationsample and this enabled randomsampling ofobservations of eachpatient throughout the day.

In order to investigate therelationship between level of motoractivity and functional motor status,patient data were obtained using allitems of the Modified MotorAssessment Scale (MMAS) (Carr et al1985) and Functional IndependenceMeasure (FIM) (Hamilton etaI1987).The MMAS is a clinical scale of eightmotor tasks which are scored on aseven-point scale (0 to 6). Good inter­rater reliability has been reported forboth the MMAS (Poole eta11988) andthe FIM (Hamiltonetal 1987). Thehospital's senior neurologicalphysiotherapist assessed each of the 17individual patients immediately priorto each week of observation.

Statistical analysisThe number of times patients were intherapy, non-therapy areas or unableto be located was recorded andexpressed as a proportion ofthe totalnumber of observations. The numberof times patients were observed to beactive and inactive was calculated as apercentage of the total number of non­therapy observations. Each posturerecorded when patients were inactiveduring non-therapy observations wasexpressed as a proportion of the totalnumber of observations when patientswere inactive. The number of times

ORIGINAl ARTICLE

Therapy(29.6%)

Unallocated(10.7%)

patients were observed to be in aparticular therapy was ·calculated as apercentage of the total number oftherapy observations. Patientsparticipated in various sessions ofeither individual or groupphysiotherapy. Each session wasexpressed as a proportion of totalphysiotherapy time.

Two methods were required toempirically investigate the followingrelationships between ~) motor activityand motor performance (using theMMAS); and ii) motor activity withfunctional status (using a total FIM).First, specific motor activities (egwalking, sitting unsupported, or use ofaffected upper limb) were individuallyexpressed as a proportion of the totalnumber of times patients wereobserved to be active. Each motoractivity was correlated with theappropriate item ofthe MMAS. Forexample, the proportion of walkingobservations was correlated with Item

AUSTRAliAN PHYSIOTHERAPY

Non-therapyActive(29.4%)

Non-therapyInactive(30.3%)

5 (walking) scores ofthe MMAS.Second, the proportion of total motoractivity performed by each patient wascalculated from the number ofobservations the patients were activedivided by the number of times thepatient was observed. This percentageof motor activity was correlated withthe total FIM score (Bunch andDvonch 1994). These analyses wereconducted using data from 17 patientsduring the first one-week observationperiod only. In this way, each patientwas only represented once in thestatistical analysis. Calculations wereperformed usingStatViewSE+Graphics™ on an Apple PowerMacintosh.

ResultsOver the nine days of observation, atotal of 325 sampling periods wereconducted, representing an average of36.1 sampling periods per day. There

AUSTRAliAN PHYSIOTHERAPY ORIGINAl ARTICLE

Figure 2.Box plot distribution of motcractivities performed during non-therapy time.U=Unaffected; A=Affected. Each box plot shows the median at the waist of the boxh,alue shown at base of graph), 25th and 75th percentiles (lower and lipper limits of box),the 10th and 90th percentHes(whiskers) and individual values which lie outside theselimits.

The results of this study have shownthat patients were involved instructured therapy for one third of the9 am to 5 pm day. This appears to behigher than the 12 per cent reportedby Mackey et al (1996), the 12.9 percent reported by Tinson (1989) andthe 19 per cent reported by Keith(1986) but comparable with thefindings ofother studies (Keith 1980and 1988, Keith and Cowell 1987,Lincolnet al 1989). Physiotherapy

observed to be in seated postures,either in a wheelchair (52.7 per cent)or sitting supported (30.9 per cent).Relatively few observations·were madein lying postures, either in supine (12.4per cent) or sidelying (3.5 percent).

Patients were observed in variouslocations throughout the hospital whennot involved in therapy, with thehighest number ofobservations in therecreational room (Table 2).

Therapy observationsOf the one third of the day spent intherapy, most observations (18.7 percent) were recorded in physiotherapy(Table 2). Most of physiotherapytreatment time was allocated toindividual treatment (77.2 per .cent),with the remaining time spent in groupactivities such as balance (9.6 per cent),hydrotherapy (6.8 per cent),gymnasium (3.3 percent) andrelaxation (3.1 per cent).

The relationship between levelof functional performance andobserved level of motor activityduring non-therapy fime.There was a wide range of individualdifferences for the MMAS items andthe total FIM score (Table 3). Patientsalso varied widely in the percentage oftime involved in any motor activityranging from 9.6 per cent to 78 percent of non-therapy time (Figure 3).In general,. there was a moderately

strong correlation between level offunctional performance andcorresponding observed level of motoractivity (Table 4).

Discussion

Non-therapy observationsWhen patients were observed in non­therapy time (66.9 per cent of the day)they were active for approximately halfof the observations (49.2 per cent) andengaged in inactive watchingbehaviour (39.3 per cent) and inactivepassive behaviour (11.5 per cent) forthe other half of the observations.

Figure 2 presents box and whiskerplots of each category of motor activityrecorded during non-therapy time(66.9 per cent of the day). rrhepredominant motor activity involvedthe unaffected upper limb and sittingunsupported. Figure 2 also highlightsthe predominance of use of theunaffected side, particularly for theupper limb.

When subjects were inactive (passiveor watching) they were mostly

'"01,)

Upper Limb Lower Limb ~

~

060 ~

U A U A bJ).~

'E .~.~ = ~

0 0tI) ::s 'a

50 0 0bJ)

bJ) ..d=. d ()

en 0 :.a ...... ~

c =. ~1,)

0 ~1,)

~ ..d-; 40

~

~

~tI) ~0

(l) 0(/)

~0

0 8 0"-t-0Q)0)

12c:Q)u 10s,.;..Q)

0-

00

23.2 1.3 0 0 0 17.5 0 4.3 8.7 0

were 3,943 individual observationsmade of patients' location andbehaviour during these samples.Patients were not involved in therapyfor approximately 60 per cent of totalobservations made over the recordingperiod (Figure 1). These were theoccasions when details of patients'behaviour and location were recorded.Patients were located in therapy for30.3 per cent of total observations.There was only a total 10.7 per cent ofall observations when patients wereunable to be located during theobservation period, with the majorityofthese when patients were on dayleave from the hospital. These 10.7 percent of observations were excludedfrom all following calculations and dataanalyses.

o RI GIN A l t\ RTI C LE AUSTRAliAN PHYSIOTHERAPY

10 20 30 40 50 60 70 80 90 100Percentage of Motor Activity Observed'

Figure 3.Relationship between percentage of motor activity and total functionallndependel1ceMeasure (FIM) score (rho =0.61, P<0.01).

3-61-60-60-60-60-6

Range

47-122

5/65/64/65/63/61/6

MedianScore

102/126

Although we do not have control dataabout activity patterns of persons of asimilar age, either in a rehabilitationsetting or ina home setting, it appearsthat some of the observed behaviourswould be detrimental to achievingtreatment goals, and indeed should beavoided. Such examples include self....

ClJ 0 00

0 0

00 00

0 00

o

o

130

120

Test

TotalFIM

MMAS Item 3 - Balanced SittingMMAS Item 4 - Sit to StandMMAS Item 5 - WalkingMMAS· Item 6 - Upper Limb FunctionMMAS Item 7 - Hand FunctionMMAS Item 8 - Advanced Hand Function

Table 3.Scores obtained. on Modified Motor AssessmentScale (MMAS) .items and totalFunctional Independence Measure fFIM).

occasions in which lower limb activitywas observed involved patientspropelling theirwheelchairs"Althoughit appears that the lower limbs wereused less than the upper limbs, it mustbe noted that the former were alsoused bilaterally in activities such asstanding up, standing and walking.

From Page 41accounted for more than half of theobservations ofstructured therapy·timeand, of this time,almost one quarterwas spent in group activities, such asbalance class and hydrotherapy. Withonly one third of the·dayspent intherapy, it is clear that patients have. alarge amount of each day with noformal structure.

An analysis of the level of motoractivity observed ·in unstructured timerevealed a disappointing pattern,withpatients inactive for half of this time.When inactive, patients were observedto be mainly sitting supported (83.6per cent of observations) and watchingthe surrounding environment. As theproportion of inactivityreported inthis study was also similar to thatreported by previous researchers,(Keith 1980, 1986 and 1988, Keith andCowell 1987, Lincoln et al 1989,Mackey et al 1996, Tinson 1989), itappears that it is common for patientswith stroke to be inactive for asubstantial part of their day. Since ithas been emphasised that skill inperforming a motor task increases as adirect function of the amount ofpractice (Carr .and Shepherd 1987a and1987b), it is argued strongly thatpatients should be making maximumuse of the available time to practisemotor tasks (Ada etal 1990). Thefindings of our study and previousresearch suggest this is not occurring.This appears to be in conflict with theprimary goal ofaphysicalrehabilitation program which aims toimprove motor skills.

During the unstructured time whenmotor activity was observed, patientswere frequently found to be sittingunsupported (median 17.5 per cent oftotal motor activity during non­therapy observations) with relativelyfewer observations of walking (median8.7 per cent) or standing (median 4.3per cent). There was a large differencebetween the use·of the affected(median 1.3 per cent) and unaffectedupper limbs (median 23.2 per cent).Similarly, there was a trend for patientsto perform fewer isolated movementsof the affected lower limb than of theunaffected lower limb. Most of the

AUSTRAliAN PHYSIOTHERAPY ORIGINAL ARTICLE

propulsion in a wheelchair and thedisproportionately high ratio ofunilateral to bilateral upper limbmovements.

The data from our study support thehypothesis that there would be anassociation between the patient's levelof motor ability and the observation ofmotor activity in unstructured time. Ofthe seven correlations examined, fivewere found to be moderately strongwith Spearman rho values rangingfrom 0.55 to 0.73. It is interesting tonote that two of the correlations werenot significant. Closer inspection ofthe data revealed a truncation of rangeeffect in the balanced sitting item(Item 3) of the MMAS with no patientsscoring below three. Similarly therewas a truncation of range in thenumber of observations of sit-to-standactivity, with only 11 observationsrecorded during non-therapy time withthe sampling technique used.Therefore, caution must be used ininterpreting that no relationshipsexisted, since true relationships may bemasked. The moderately strongrelationship between observed walkingactivity and the walking item of theMMAS indicates that patients withhigher levels of walking ability werefound to be walking more frequently.In this sample only half of the patientshad adequate skill in walking topractise independently out of therapy.

This is likely to explain why walkingwas observed less frequently thansitting. Further evidence of arelationship between motor ability andfrequency of observed motor activity isprovided by the significant correlationbetween the total FIM and percentageof motor activity.

Although moderately strongrelationships were found betweenobserved motor activity of the affectedupper limb and the correspondingitems of the MMAS, these positivecorrelations must be interpreted withcaution. As expected, these positive~relationships indicate that patients withhigher scores were more likely to beobserved using their affected upperlimb than patients with lower scores.However, since the method ofrecording did not specify the nature ofthe observed upper limb activity, thepossibility exists that, even thoughpatients were using their affectedupper limbs, they were not doing so atthe level of which they were capable.For example, a patient who was able,under test conditions, to pick up asmall object such as a pen cap (MMASItem 8 [advanced hand activities],Levell) may only be using their armto stabilise a book or dinner plate.Therefore, one possible explanation isthat the concept of learned non-use(Taub 1980), defined by Barton andWolf (1993) as "the discrepancy

between inherent neurological abilityand functional ability", is occurring inthis population. The potential todevelop learned non-use is ofparticular concern in the early stagesfollowing stroke, especially for thosepatients with very limited movement inthe upper limb. Clearly, furtherresearch is required to investigatewhether patients perform motoractivities which are appropriatelychallenging.

It is also difficult to interpret themarked disparity between thefrequency of use of the affected andunaffected upper limbs. It could bepartly attributed to both the primarymotor deficit and learned non-use.Furthermore, the effect of motordominance may also have contributed.The design of this study does notpermit further examination of therelative contribution of each of thesefactors, however, this is an area whichalso requires further investigation.

In attempting to derive clinicalimplications from these data, severalpoints must be considered. Patients inthis study were inactive for longperiods of the day when out of therapy.When active, one of the mostfrequently recorded motor behaviourswas unsupported sitting, which was nota demanding task for this population

From Page 49(MMAS Item 3 [sitting], median =5).This ,strongly suggests that thesepatients were not effectively using timeavailable out of therapy to practiseappropriately challenging functionalmotor tasks in line with the goals ofaphysical rehabilitation program.

There are a number of options whichmay prove effective in, achieving thegoal of increasing the frequency ofpractice of appropriate motor skills.Patients need to be given structuredroutines which enable effectiveindependent practice of motor tasks.This concept has been advocated byCarr and Shepherd (1987a and 1990)and described in detail by Adaet al(1990). Rehabilitation staff need toreinforce the concept that thepredominant activity each day shouldbe to work towards the goals ofrehabilitation, whether in therapy ornot. If each day was considered 'a 9 amto 5 pm working day, even with theinclusion of two one-hour rest periods,the proportion of motor activity shouldincrease to higher levels than observedto date. There should be anexpectation that the patient must takeresponsibility for the learningprogram. Education, monitoring andencouragement within a multi­disciplinary team framework areimportant in promoting this concept.For example, it may be desirable forteam members to assist the patient towalk between therapy sessions ratherthan only in the closed environment ofa treatment area. In this way, patientslearn to adapt to the task demands ofwalking with other pedestrians, over avariety of surfaces, in different lightingconditions and with the challenge ofobstacles as they occur in theenvironment (Bassille and Bock 1995).By encouraging mobility 'activitieswithin the rehabilitation setting,patients are likely to take moreresponsibility for thei~~"own mobilitythroughout the day. '/'-

Therapists must also consider thebest means of evaluating theeffectiveness of such changes totraditional rehabilitation practice. Ajudgment may be made about how

ORIGINAL ARTICLE

much practice is expected in non­therapy time for each individualpatient and then strategiesimplemented to monitor whether theseactivity levels are being achieved.Further research must focus onexamining the potential benefits of theprovision ofintensive structuredpractice for motor outcome of strokepatients. Given the pressure to reducelength of stay without compromisingoutcome, strategies to improve motorfunction more efficiently warrantfurther investigation.

The positive correlation betweenfunctional motor ability and observedactivity indicates that patients of lowerfunctional abilities are less active out oftherapy. Therapists need to employinnovative thinking to ensure that thepractice for this group is safe,adequately challenging and sufficientlyfrequent. For example, for a patient:vho is not independent with transfers,Increasing bed and chair height mayenable both independent transfers andmore frequent practice ofstanding up.Consideration must also be given tocognitive and behavioural deficits andtheir effect on motor activity levels.One of the crucial elements whichmust be considered in independentpractice is the safety of the patient.Clearly, 'the ideal of practising undersupervision is becoming difficult in thecurrent cost-cutting climate. Onep~otentialstrategymay be to involvefamily and friends in providingassistance with initiation and executionof strucmred practice programs.

\Vhen inactive, patients wererecorded to be seated for more than 80per cent ofobservations. Thisprolonged period ofimmobilityincreases the likelihood of thedevelopment of length-associatedchanges of the musculoskeletal system(Ada and Canning 1990), thus creatingfurther complications for theretraining of functional movement(Ada and Canning 1990). Therefore,therapists need to consider the bestmethods of preventing these changes.

The data from this study showed thatpatients spent more than 50 per cent ofnon-therapy time in the recreation

AUSTRAliAN PHYSIOTHERAPY

room and bedroom. This informationsuggests that if changes were possibleat this hospital, then these locationsshould be targeted. For example in therecreation room, a table ofboardgames modified for patients withlimited upper limb control mayencourage more frequent and desirablemotor activity than chairs groupedaround a television. In order toincrease the amount ofindependentpractice, therapists must consider thefeatures of their own clinicalenvironments and how they maypositively or negatively influencemotor behaviour levels.

Although these data were collectedfrom a relatively small sample ofpatients undertaking inpatientrehabilitation following stroke, one ofthe strengths of this smdy's internalvalidity lies in the intense datacollection during the period of thestudy. A total of3,943 observationswere made, giving avery detailedpicture ofpatient activity on weekdays.Clearly, it is not known from this studyhow much activity occurs prior to 9 amand after 5 pm, but until this isexamined, it cannot be assumed todiffer substantially. Data provided byMackeyet al (1996) indicated thatthere was even less motor activity onthe weekends than on weekdays.Similarly, data provided by Lincoln etal (1996) indicated that patients werepredominantly sitting or lying downduring non-therapy time whichincluded observation periods from 6am to 10 pm. Despite the relativelysmall number of patients, their largevariation in functional ability reflectsthe wider stroke population. Since thisstudy was conducted in a fast streamrehabilitation hospital, the findings canonly be generalised to similar settings.Activity levels may be even worse inslow stream rehabilitation settings~

This study confirms the resultsreported by the only other study(Mackey et al 1996) of activity levels instroke rehabilitation within theAustralian health systema

In summary, from the accumulationof knowledge to date, it appears thatchange is required to ensure thatpatients utilise the crucial learning

AUSTRAliAN PHYSIOTHERAPY

period in the first few weeks poststroke. Physiotherapists need to beinnovative in assisting their patients tomaximise the opportunities for motorlearning through independent practiceduring their stay as inpatients inrehabilitation. Further research will berequired to evaluate the effectivenessof any systematic changes to theutilisation of the large amount ofunstructured timewmch is availableduring inpatient rehabilitationfollowing stroke.

AcknowledgementsThis study was completed in partialfulfilment of the requirements of theBachelor of Physiotherapy withHonours at the School ofPhysiotherapy, Faculty of HealthSciences, La Trobe University,Melbourne. The authors thank thestaff and patients of the HamptonRehabilitation Hospital for theirinvaluable participation in thisresearch.

ReferencesAda L· and Canning C (1990): Anticipating and

avoiding muscle shortening. In Ada LandCanning C (Eds): Key Issues inNeurologicalPhysiotherapy. Oxford: ButterworthHeinemann, pp. 219-236.

AdaL, Canning Cand WestwoodP (1990): Thepatient as an active learner. In Ada LandCanning C (Eds): Key Issues in NeurologicalPhysiotherapy. Oxford: ButterworthHeinemann, pp. 99-124.

Anderson CS,JamrozikKD, Burvill PW, ChakeraTM, Johnson GA and Stewart-WynneEG(1993): Ascertaining the true incidence ofstroke: Experience from the PerthCommunityStroke Study 1989-1990.MedicalJournal.ofAustralia 158: 80",84.

Barton LA and WolfSL (1993): Learned non-usein thehemiplegicupper extremity. In GordonWA(Ed.): Advances in Stroke Rehabilitation.Boston: Andover MedicalPllblishers,pp.79-87.

o RIG IN A L ARTie LE

Bassile CC and Bock C (1995): Gait training. InCraik RL and Oatis CA (Eds): Gait Analysis.St Louis: Mosby, pp. 420-43 5~

Bunch WHand Dvonch VM (1994): The value ofFunctional Independence Measure scores.American Journal of Physical Medicine andRehabilitation 73: 40-43.

Carr]HandShepherd RB (1987a): AmotorIearningmodel for rehabilitation. In CarrJH,ShepherdRB, Gordon], Gentile AM and Held JM(Eds): Movement Science Foundations forPhysicalTherapy in Rehabilitation. London:Heinemann Physiotherapy, pp. 31-91.

Carr JH and Shepherd RB (1987b): A MotorRelearning Programmefor Stroke. (2nded.)London: Heinemann.

CarrJH and Shepherd RB (1990): Amotorlearningmodel for rehabilitation of the movement­disabled. In AdaL and CanningC (Eds): KeyIssuesin NeurologicalPhysiotherapy. Oxford:Butterworth Heinemann, pp. 1-24.

CarrJH, Mungovan SF, Shepherd RB,Dean CMand Nordholm LA (1994): Physiotherapy instroke rehabilitation: Bases for Australianphysiotherapists' choice of treatment.Physiotherapy Theory andPractice. 10: 201-209.

CarrJH, Shepherd RB, Nordholm L and Lynne D(1985): Investigation of a new MotorAssessment Scale for stroke patients. PhysicalTherapy. 65: 175-180.

Christie D (1981): Stroke in Melbourne, Australia:An epidemiological study. Stroke. 12:467-469.

DeanCMand MackeyFH (1992): MotorAssessment Scale scores as a measure ofrehabilitation outcome following stroke.Australian Journal ofPhysiotherapy 38: 31-35.

DombovyML, SandockBAand BasfordJR (l986):Rehabilitation for stroke: A review. Stroke17: 363-369.

Ernst E (1990): A review of stroke rehabilitationand physiotherapy. Stroke 21: 1081-1085.

Esmonde TJ (1994): Stroke rehabilitation: Patientactivity during non-therapy time.Unpublished honours thesis, La TrobeUniversity, pp. 97-106.

FeigensonJS, GitlowHS and GreenbergSD (1979):The disability oriented rehabilitation unit - amajor factor influencing stroke outcome.Stroke 10: 5-8.

Hamilton BB, Granger CV, SherwinFS, ZieleznyMandTashmanJS (1987): AunifonnnationaIdata system for medical rehabilitation. InFuhrer MJ (Ed.): Rehabilitation OutcomesAnalysis and Measurement. Baltimore: PaulBrookes, pp. 137-150.

Keith RA (1980): Activity patterns ofa strokerehabilitation unit. Social Science andMedicine.14A: 575-585.

Keith RA (1986): Organizational effectiveness inthe rehabilitation hospital: A cross-nationalcomparison. Rehabilitation Psychology 31:131-141.

KeithRA (1988): Observations in the rehabilitationhospital: Twentyyearsofresearch. ArchivesofPhysical Medicine and Rehabilitation 69:625-631.

KeithRA and Cowell KS (1987): Time use ofstroke patients in three rehabilitationhospitals. Social Science and Medicine 24:529-533.

LincolnNB, Garnlen R and Thomason H (1989):Behavioural mapping of patients on a strokeunit. International Disability Studies. 11 :149-154.

Lincoln NB,Willis D, Philips SA, Juby LC andBerman P (1996): Comparison ofrehabilitation practice on hospital wards forstroke patients. Stroke 27: 18~23.

Mackey F, Ada L,Heard R and Adams R (1996):Stroke rehabilitation: Are highly structuredunits more conducive to physical activity thanless structured units? Archives of PhysicalMedicine and Rehabilitation. (In press).

Poole J and Whitney S (1988): Motor assessmentscale for stroke patients: concurrent validityand inter-rater reliability. Archives ofPhysicalMedicine and Rehabilitation 69: 195-197.

Shah S (1989): Admissions, patterns of utilizationand disposition of cases of acute stroke inBrisbanehospitals. MedicalJournalofAustralia150: 256-260.

Taub ·E (1980): Somatosensory deafferentationresearch with monkeys: Implications forrehabilitation medicine. In Ince LP (Ed.):Behavioural Psychology In RehabilitationMedicine: Clinical Applications. Baltimore:Williams and Wilkins, pp. 371-401.

Timson DJ (1989): How stroke patients spendtheir days. Internationa{Disability Studies 11:45-49.