Power Mobility Driving Training for Seniors: A Pilot Studyfhs.mcmaster.ca/powermobility/hall_wheelchair_training_article.pdf · POWER MOBILITY DRIVING TRAINING FOR SENIORS 49 protocols,

Name /astc/17_103 03/14/2005 11:05AM Plate # 0-Composite pg 47 # 1

Allen Press DTPro SystemFile # 03em

47

APPLIED RESEARCHAsst Technol 2005;17:47–56

q 2005 RESNA

Power Mobility Driving Training for Seniors:A Pilot Study

*Karen Hall, BA, MScOT, OTReg (Ont), †Jacqueline Partnoy, BScOT, OTReg (Ont), †SherylTenenbaum, BScOT, OTReg (Ont), and ‡Deirdre R. Dawson, PhD, OTReg (Ont)

*Assistive Technology Clinic, Sunnybrook and Women’s College Health Sciences Centre, Toronto, Ontario, Canada;†Department of Occupational Therapy, Baycrest Centre for Geriatric Care, Toronto, Ontario, Canada; and‡Department of Occupational Therapy, University of Toronto, and Kunin-Lunenfeld Applied Research Unit, Baycrest Cen-tre for Geriatric Care, Toronto, Ontario, Canada

This article describes two power mobility trainingprotocols used with seniors and compares posttrainingdriving performance. Twelve users of power mobilitywere consecutively recruited from two residential fa-cilities in Toronto, Canada. The aim of training at bothsites was to make clients comfortable with and safe atdriving power mobility devices. The content of trainingwas similar, but training protocols differed signifi-cantly in terms of the number of sessions (means of3.43 vs. 9.80; p # .05) and the time frame over whichthe sessions were offered (means of 1.57 vs. 5.10weeks; p # .01). Participants at the two sites differedsignificantly in terms of overall driving performance(p # .05), gender (p # .01), and type of device used(p # .05). Overall, driving performance was signifi-cantly associated with facility, gender, type of deviceused, and training duration (p # .05). When these var-iables were entered into an exploratory hierarchicalregression, facility accounted for 64% of the variancein driving performance. When facility was controlledfor, the correlations between device and duration oftraining with driving performance were no longer sig-nificant. The determinants of driving performance aredifficult to clearly specify as the variable facility en-compasses gender as well as all other differences be-tween the two training protocols. Nevertheless, thesedata provide direction for future research in this area.

Key Words: Power mobility—Wheelchair training—Intensity—Geriatrics—Power-Mobility Indoor DrivingAssessment (PIDA)

Address correspondence and reprint requests to Karen Hall,BA, MScOT, OT Reg (Ont), Assistive Technology Clinic, Sunny-brook and Women’s College Health Sciences Centre, Room UG 20,2075 Bayview Avenue, Toronto (Ontario) M4N 3M5, Canada.

The use of power wheelchairs and scooters bythe elderly has increased dramatically in recentyears. These power mobility devices increase in-dependence and improve the quality of life of in-dividuals with mobility limitations (Miles-Tapping& MacDonald, 1994; Scott-Taplin, Smith, Mc-Laughlin, & Mathews, 1989). However, the safeuse of these devices is a growing concern as thenumber of accidents involving these devices is alsoincreasing, especially among elderly users (Kirby& Ackroyd-Stolarz, 1995; Reed, Yochum, &Schloss, 1993; Ummat & Kirby, 1994). One way toincrease safety and optimize power mobility driv-ing performance is through training elderly clientsin the use of these devices (Kirby, Coughlan, &Christie, 1995). The overall purpose of this articleis to begin a discussion that will ultimately lead todelineating and understanding best practice forpromoting safe and effective use of power mobilityamong elderly users.

Training clients in the use of power mobility de-vices is important, not only to enhance safety anddriving performance, but also to increase satisfac-tion with the devices and conversely decrease aban-donment (Lange, 2000). Given the cost of these de-vices to the health care system, organizations, andindividuals, this latter point it not trivial. In theUnited States, approximately 16 of every 1,000 peo-ple 65 and older use either a power or manualwheelchair (Calder & Kirby, 1990). In Ontario, thenumber of individuals aged 65 and older who re-ceived funding assistance from the Ministry ofHealth and Long-Term Care for power mobility de-vices through the Assistive Devices Program rose



48 ASSISTIVE TECHNOLOGY, VOL. 17, NO. 1

by 340% between 1995 and 2001. Reasons for theincrease in power mobility use by the elderly in-clude technological advances resulting in suitableand effective means of mobility for individuals of allages and ability level (Kirby et al., 1995) and thegrowing percentage of the population aged 65 andolder (Rosenburg & Moore, 1997). The higher prev-alence of chronic conditions and subsequent dis-ability among the elderly also results in an in-creased use of these devices (Wilkins & Park, 1996).

Coincident with the rise in power mobility use isthe number of accidents involving seniors. Ummatand Kirby (1994) evaluated nonfatal wheelchair-re-lated accidents reported between 1986 and 1990.From these data, the average number of accidentswas estimated at 36,559 per year. The mean age ofpersons reporting accidents was 60.6 years, andpersons in the 80–89 year age group were most fre-quently affected, accounting for almost a quarter ofthe total accidents. Calder and Kirby (1990) iden-tified 770 cases of wheelchair-related deaths be-tween 1973 and 1987. Cases involving individuals60 years and older accounted for 87% of the totalfatalities. Over 80% of these fatal accidents oc-curred indoors, either in the home or an institution-al setting. Of nonfatal accidents reported, over 70%happened in indoor environments (Ummat & Kirby,1994). These statistics highlight the necessity ofpromoting safe driving, particularly in the living en-vironment of seniors who use power mobility.

Although there is a clear trend toward increas-ing use of these devices by seniors, there does notappear to be a corresponding trend in trainingdrivers. Reed et al. (1993) reported that less thanhalf of residents using power scooters in three in-stitutional settings had any type of formal trainingin the use of these devices. Furthermore, limitedliterature is available to inform therapists with re-spect to training clients. Existing research has fo-cused primarily on power mobility driving traininginvolving children (Furumasu, Guerette, & Tefft,1996; Nisbet, Craig, Odor, & Aiken, 1996; Tefft,Guerette, & Furumasa, 1999).

Nevertheless, there are theories and some re-search that point to the direction training shouldtake. Occupational therapy theory suggests thatoptimal mobility arises from the transactional re-lationship between the person, the environment,and the tasks or occupations with which the personis engaged (Law et al., 1996). This approach is sup-ported by the work of a number of researchers whohighlight the importance of using a multifacetedapproach to facilitate independent, optimal, andsafe use of power mobility (Field, 1999; Routhier,Vincent, Desrosiers, & Nadeau, 2001; Ummat &

Kirby, 1994). They have identified the followingfactors as important in power mobility use: theuser’s profile; the wheelchair; the environment;daily activities; social roles; and the activity ofdriving, which includes training. Interestingly, weare aware of no studies that address issues of theamount or intensity of training necessary for op-timal power mobility use.

Intensity of therapeutic interventions is definedby Keith (1997) as ‘‘the duration and frequency ofa procedure within a given time span’’ (p. 1299). Anumber of researchers studying different aspectsof rehabilitation have demonstrated that a greaterintensity of treatment results in better outcomesfor the client compared with less intense treatmentwhen other factors (e.g., diagnosis, age) are takeninto account. For example, Chiodo, Gerety, Mul-row, Rhodes, and Tuley (1992) found that 88% ofelderly nursing home residents who received high-intensity physical therapy improved in physicalfunction compared with 33% and 13% who receivedmoderate and minimal intensity therapy, respec-tively. These changes could not be accounted for byage, cognitive status, level of independence in ac-tivities of daily living, or diagnoses. Similarly, ameta-analysis of nine controlled studies involving1051 stroke patients showed significant improve-ments in activities of daily living and neuromus-cular and functional outcomes as a result of moreintense occupational and physical therapy (Kwak-kel, Wagenaar, Koelman, Lankhorst, & Koetsier,1997).

In summary, to date, limited information existsto inform therapists with respect to training elder-ly clients in the use of power mobility devices. Yetavailable statistics indicate both an increase in thenumber of elderly power mobility users and in thenumber of power mobility accidents. We believethat training new drivers in the use of their powermobility devices will enhance performance, im-prove safety, and increase functional mobility. Theoverall purpose of this study was to begin to ex-plore the impact of training on performance amongusers of power mobility. Specifically, this article

● describes two power mobility driving trainingprotocols used with elderly clients at two long-term care facilities;

● compares the socio-demographic, diagnostic,and mobility-related characteristics of the cli-ents at these facilities; and

● investigates whether there is a significant dif-ference in power mobility driving performancearising from different power mobility training



49POWER MOBILITY DRIVING TRAINING FOR SENIORS

protocols, specifically in terms of training inten-sity.

METHODS

Participants

Participants for this study were recruited over a7-month period (February–August 2002) from theBaycrest Centre for Geriatric Care (Baycrest) andthe Long-Term and Veterans Directorate at theSunnybrook and Women’s College Health ScienceCentre (Sunnybrook). To be eligible, residents ofthese facilities had to be 65 years or older, bedeemed competent to consent by the health careteam, speak and understand English, and be med-ically able to participate in the power mobilitydriving training program. Participants had mobil-ity limitations arising from diagnoses of stroke;chronic obstructive pulmonary disorder (COPD);multiple sclerosis (MS); or a musculoskeletal-re-lated diagnosis (e.g., osteoarthritis). Participantsmay have used a power mobility device sometimein the past (e.g., rented a scooter when on an out-ing); however, to obtain a new device and use itwithin the facility, both facilities required and pro-vided training. Of the 13 people deemed eligible toparticipate, all gave written informed consent toparticipate in the study. One of these was excludedlater because complete data could not be obtained.

Instruments

To assess the power mobility driving perfor-mance of the participants, the Power-Mobility In-door Driving Assessment (PIDA) was used (Dawson,Chan, & Kaiserman, 1994; www.fhs.mcmaster.ca/powermobility; see Appendix for score sheet).This assessment was designed to measure drivingperformance indoors for residents of long-termcare facilities who use power mobility. Up to 30items can be scored, but only items necessary forindividuals to perform within their environmentare scored. For example, if someone never has toaccess his or her bed from the right, this item is nottested. Each item is scored on a 4-point Likert-typescale from 1 (unable to complete task) to 4 (com-pletely independent). The PIDA has content andface validity, moderately good intrarater reliabili-ty, and very good interrater reliability.

To characterize the cognitive level of partici-pants, the Standardized Mini-Mental State Ex-amination (SMMSE) was used (Molloy, Alemaye-hu, & Roberts, 1991). This instrument contains 30items that measure orientation to time and place,immediate recall, short-term memory, calculation,

language, and constructive ability. The SMMSEwas chosen for several reasons. It is the most wide-ly used assessment to measure cognitive impair-ments in seniors, it is easy to administer and score,and it has very good intrarater and interrater re-liability. Typically, a score of 24 of 30 or above in-dicates there is no dementia present.

Attention to extra-personal space was assessedwith the Bells test (Gauthier, Dehaut, & Joanette,1989). This assessment, which detects visual ne-glect, consists of a paper divided into seven col-umns, each of which contains 35 distracter and fivetarget items (bells). Neglect and lack of attentioncan be determined from the number and pattern oftargets circled. The test has good validity, andsome normative data are available (Spreen &Strauss, 1998). It was selected for this study be-cause it is quickly and simply administered andhas been demonstrated in at least one study to besuperior to other tests in terms of detecting visualneglect (Ferber & Karnath, 2001).

Intensity of driver training was measured as thetotal number of training sessions divided by theduration of training in weeks. The number of ses-sions were recorded to include the initial trainingsession to the final session prior to the posttrainingPIDA. The duration of training was measured asbeginning at the initial training session and end-ing on the date of the administration of the post-training PIDA. Socio-demographic characteristicswere determined from a short questionnaire ad-ministered to the participants.

Driver Training Protocol

This article compares two training protocols:those in use at Baycrest and Sunnybrook at thetime this study was conducted. A comparison of thetraining protocols is summarized in Table 1. Bothcenters have used the PIDA for some time as a wayof measuring driving performance. Thus, the skillsrequired by the PIDA are those incorporated intothe training. As such, the content of training at thetwo sites was similar. If a client had no prior ex-perience using a power mobility device, they wereprovided with an orientation to the device in thefirst session. Skills covered in this session includedtransferring to and from the power mobility device,turning the controls on and off, adjusting thespeed, adjusting the seat belt, and driving in all di-rections within the client’s immediate environ-ment (e.g., room or unit of the facility). Subsequentsessions included learning and/or practicing skillslisted on the PIDA. However, clients were nottrained in all items on the PIDA, but rather only




TABLE 1. Comparison of power mobility driving training protocols

Baycrest Sunnybrook

Documented training protocolStandard number of sessionsStandard duration of trainingTrainer

NoTwelve 1-hour sessions3 weeksOTs and OT assistants

YesSix 1-hour sessions2 weeksOT assistants

Pretraining PIDAPosttraining PIDABlind rater for posttraining

PIDA

Not routineRoutine

No

Not routineRoutine

Yes

Note: PIDA 5 Power-Mobility Indoor Driving Assessment; OT 5 occupational therapist.

those skills that were applicable to that client. Forexample, a client who did not use the device in thebathroom was not trained in skills that corre-sponded to items five through eight on the PIDA.

There are a few key things to note in comparingthe two training protocols (see Table 1). First, theprotocol at Sunnybrook was documented with spe-cific content for each of the six sessions routinelyprovided. At Sunnybrook only, this included a for-mal session on maintenance and care of the powermobility device. In contrast, the content of thetraining protocol at Baycrest was individualizedand determined by the occupational therapist (OT)as clients’ needs for more or less practice on par-ticular skills emerged. The second thing to note isthat the amount, duration, and intensity of train-ing differed between the two sites. At Sunnybrook,routine practice entailed six 1-hour sessions over 2weeks (3 sessions/week). At Baycrest, routine prac-tice entailed twelve 1-hour sessions over 3 weeks(4 sessions/week). Third, an OT blind to the resultsof the training routinely administered the post-training PIDA at Sunnybrook. Interestingly, nei-ther site routinely did pretraining PIDAs, al-though of the five participants recruited at Bay-crest, three had pretraining scores.

Procedure

Occupational therapy at both sites continued ina routine manner during the course of this study.Clients who received training in the use of powermobility were asked to participate in the studyonly after the posttraining PIDA was administeredand after a decision was made regarding their driv-ing status (i.e., safe/unsafe to drive within the fa-cility). The outcome of training was determinedprior to recruitment so that clients’ decisions toparticipate in the study could in no way impacttheir driving training, evaluation, or decisions re-garding driving status. The OTs then asked the cli-ents if they could be approached regarding a re-

search study. With their verbal consent they wereapproached by the principal investigator (PI) andinvited to participate in the study. Upon writtenconsent, the PI obtained the participants’ trainingdocumentation and conducted an interview withparticipants. The interviews took approximately20–30 minutes and included socio-demographicquestions, the SMMSE, and the Bells test. Thestudy received ethics approval from the Sunny-brook, Baycrest, and University of Toronto scien-tific and ethics review boards.

Planned Analyses

Descriptive statistics were done on all variablesin order to characterize the sample, and differenc-es on these variables across the two sites were cal-culated. Interval and ratio data (e.g., trainingamounts and posttraining PIDA scores) were an-alyzed using t tests. Nominal data were analyzedwith Fisher’s exact x2 (e.g., gender and type of de-vice). Correlational analyses were run between thedependent variable, driving performance, as mea-sured by the posttraining PIDA and the descrip-tive variables. Nonparametric correlations wereused due to the categorical nature of some data andthe small sample size. The Kendall t-b was select-ed as it controls for data that have the same rank.Based on the results, an analysis of variance (AN-OVA) was conducted to determine how much of thevariance in posttraining driving performancecould be accounted for by the independent vari-able, facility. Additionally, partial correlationswere run between the power mobility trainingcharacteristics found to differ between the twosites and posttraining driving performance whilecontrolling for facility. This was done to determineif differences in the training protocol would still becorrelated with driving performance when control-ling for the variation explained by the facility var-iable. A significance level of .05 was set, althoughdifferences and correlations with p values # .10




TABLE 2. Participant characteristics

Baycrest(n 5 5)

Sunnybrook(n 5 7)

AgeM (SD)Range

Ratio of males:females*

86.60 (9.53)73–97

0:5

81.71 (5.94)76–91

7:0

Education completed, High school$ High school

1 (20%)4 (80%)

5 (71.4%)2 (28.6%)

DiagnosisStrokeCOPDMSMusculoskeletal

1 (20%)0 (0%)0 (0%)4 (80%)

1 (14.3%)1 (14.3%)1 (14.3%)4 (57.1%)

SMMSE score, M (SD)Bells test score, M (SD)

24.80 (5.50)28.75 (6.75)

25.71 (2.50)30.43 (3.26)

Note: COPD 5 chronic obstructive pulmonary disorder;MS 5 multiple sclerosis; SMMSE 5 Standard Mini-MentalState Examination.

*p # 0.01.

are discussed as potentially representing a trendin the data. All analyses were conducted usingSPSS 10.0.

RESULTS

Characteristics of the participants are shown inTable 2. There were no statistically significant dif-ferences between participants at the two sites ex-cept in terms of gender (p # .01), as all partici-pants at Baycrest were women and all at Sunny-brook were men. However, participants at Bay-crest showed a trend toward being older, onaverage by 5 years, and being more educated with80% having finished high school compared with29% of Sunnybrook participants.

Table 3 shows the mobility-related characteris-tics of the 12 participants. There were no signifi-cant differences between the participants at thetwo sites in terms of their previous driving expe-rience (motor vehicle or power mobility). However,there was a trend for participants at Sunnybrookto have more motor vehicle driving experience(100% compared with 60% at Baycrest) and forparticipants at Baycrest to have more power mo-bility driving experience (100% compared with43% at Sunnybrook). The two groups were signif-icantly different in terms of the type of device used.All participants from Baycrest used power wheel-chairs, whereas only 29% (2/7) of users at Sunny-brook used power wheelchairs (p # .05).

Data on the training protocol are also shown in

Table 3. As per the protocol, the amount of trainingprovided at Baycrest was significantly more thanthat provided at Sunnybrook, with an average of9.80 sessions compared with 3.43 sessions (p #.05). The duration of training was also significant-ly greater at Baycrest, averaging 5.1 weeks com-pared with 1.57 weeks (p # .01). However, the in-tensity of training (sessions/week) was not signif-icantly different. Data on the length of each train-ing session was available for only one participantat Sunnybrook, thus disallowing comparison.

In terms of driving performance, there was a sig-nificant difference between the two groups, withdrivers at Sunnybrook having a mean posttrainingPIDA score of 96.84 of 100% compared with 86.85of 100% among Baycrest drivers (p # .05). Therewas no difference in posttraining PIDA scores be-tween users of specifically power wheelchairs at ei-ther site—there were no scooter users at Baycrestfor comparison. However, when the participants atboth sites were pooled, a significant difference ondriving performance was found between the driv-ers of wheelchairs and scooters. Drivers of wheel-chairs had a mean posttraining PIDA score of89.08 of 100%, and drivers of scooters had a meanscore of 97.71 of 100% (p # .01). Pretraining PIDAscores were only available for three of the partici-pants, thereby disallowing analyses of this vari-able.

Table 4 shows the relationships between socio-demographic, morbidity, and power mobility char-acteristics and driving performance. Four vari-ables were found to be significant at the .05 level:gender, facility, device used, and duration of train-ing. Men at Sunnybrook driving scooters withshorter durations of training had higher posttrain-ing PIDA scores. In addition, there appeared to bea trend for a higher number of training sessions tobe associated with lower training scores. However,all of these variables are interrelated as the sam-ple at Sunnybrook are all men, all the scooter driv-ers are at Sunnybrook, the average duration oftraining at Sunnybrook was significantly shorter,and there were fewer training sessions at Sunny-brook.

Based on these results, an ANOVA was con-ducted with driving performance as the dependentvariable and facility as the independent variable.In this analysis, facility accounted for 64% of thevariance in driving performance. To determine iftraining protocol characteristics were still corre-lated with driving performance while controllingfor facility, partial correlations were run betweenthe power mobility characteristics and driving per-formance. When facility was controlled for, the cor-




TABLE 3. Power mobility–related characteristics

Baycrest(n 5 5)

Sunnybrook(n 5 7)

Ratio of wheelchair:scooter users** 5:0 2:5

Previous driving experienceMotor vehiclePower mobility

3 (60%)5 (100%)

7 (100%)3 (42.9%)

Actual power mobility trainingNumber of sessions

M (SD)**Range

9.80 (2.77)5–12

3.43 (2.44)1–6

Duration in weeksM (SD)*Range

5.10 (1.34)3–6.5

1.57 (1.24)0.5–4

Intensity (sessions/week)M (SD)Range

1.91 (0.23)1.67–2.2

2.21 (0.57)1.5–3

Session length (minutes)M (SD)Range

54.87 (9.93)15–90

a

a

Power mobility driving performance (posttraining PIDA scores)PIDA score

M (SD)**Range

86.85 (5.18)81.7–95.0

96.84 (3.11)92.0–100

Wheelchair usersM (SD)Range

86.85 (5.18)81.7–95.0

94.66 (3.76)92.0–97.3

Scooter usersM (SD)Range

N/AN/A

97.71 (2.77)93.0–100

Note: PIDA 5 Power-Mobility Indoor Driving Assessment.aData available for only 1 of 7 participants.*p # .01. **p # .05.

relations between device and duration of trainingwith driving performance were no longer signifi-cant.

DISCUSSION

Our primary finding is that posttraining drivingperformance was significantly better at one sitethan the other. As there were significant differ-ences in the participant characteristics betweenthe sites, we cannot isolate this finding to differ-ences in the training protocol, but rather suggestthis finding be used to direct future research.

The most obvious difference between the twogroups was gender and, related to this, associatedpast roles. All participants from Sunnybrook weremen and also all veterans of the Armed Forces. Al-

though there was no significant difference betweenthe two facilities in terms of past motor vehicledriving experience, the difference may have beenclinically relevant (60% at Baycrest compared with100% at Sunnybrook). Many of the male veterans,in addition to driving a car for several decades, alsohad experience driving other vehicles, includingairplanes, army tanks, and/or jeeps. One partici-pant stated that learning to drive a power wheel-chair was easy for him because he had severalyears experience flying an airplane that had sim-ilar controls to his power wheelchair. This overallpast experience by the veterans at Sunnybrookmay have enhanced their power mobility drivingabilities relative to the participants at Baycrest.

The type of device used may also have affecteddriving performance. Again, this variable (device)




TABLE 4. Associations between participant char-acteristics and training variables with driving per-

formance (posttraining PIDA scores)

VariableCorrelation

(p value)

Socio-demographic characteristicsAgeGenderEducationPast car driving experience

20.250 (.267)0.645 (.012)

20.451 (.780)20.440 (.860)

Morbidity characteristicsMMSE scoreBells test score

20.147 (.526)20.150 (.530)

Power-mobility characteristicsFacilityPast power mobility driving experienceDevice usedNumber of training sessionsDuration of training in weeksIntensity of training

0.645 (.012)0.435 (.089)0.645 (.012)

20.413 (.070)20.469 (.038)

0.244 (.301)

Note: Significant associations are bolded. PIDA 5 Pow-er-Mobility Indoor Driving Assessment.

was subsumed in facility as all the scooter driverswere at Sunnybrook. Conceivably a higher level ofperceptual and fine motor skills are required tomaneuver a power wheelchair via a joystick (thetype of control used by all wheelchair drivers inthis study). In contrast, scooters are controlledthrough a tiller bar with more natural directionalcontrols. The resulting performance differencemay arise from the fact that driving a scooter maytap into previously learned procedural skills, suchas riding a bicycle. Further, a scooter providesmore physical freedom than does a wheelchair.This results in much more physical ease with vi-sual checking of the environment, which is neces-sary to prevent accidental collisions.

Our second major finding is that there are sub-stantially different approaches to training peoplein the use of power mobility, even between twosites that are closely related in a number of ways.Both are University of Toronto teaching sites, bothuse the PIDA as their primary measure of drivingperformance, and both have therapists who havebeen involved in the instrument development ofthe PIDA and/or the Power-mobility CommunityDriving Assessment (PCDA; Letts, Dawson, &Kaiserman-Goldstein, 1998). Nevertheless, onesite has a much more structured approach to train-ing than the other and delivers less training overa shorter period of time. Although these charac-teristics were not correlated with driving perfor-

mance once other differences between the two fa-cilities were controlled for, this may have been anissue related to sample size, and merits further in-vestigation. As such, factors related to the drivingtraining protocol across the two sites warrant dis-cussion. Duration of training and the number ofsessions were significantly different between thetwo sites. Power mobility users at Sunnybrook re-ceived fewer sessions on average over a shortertime period. A possible explanation for this is thatthe male veterans, driving scooters, were able todevelop adequate driving skills in a shorter timeperiod, and accordingly their training was trun-cated. However, these results merit further explo-ration in relation to what is known about acquisi-tion and retention of procedural memories.

Doyon and Ungerlieder (2002) summarize re-search findings that suggest there are three stagesin the acquisition of motor skills. The first is anearly, fast learning stage where there is great im-provement over one training session. A second pro-posed stage is a consolidation phase occurring atabout 6 hours after the original training, providedthere is no interference from other tasks that needto be learned in that window. The final stage is aslower stage that occurs across sessions in whichthe behavior becomes well learned and can be re-trieved readily over long periods without practice.Utilizing these stages in designing a training pro-tocol could mean that for persons previouslyskilled in driving, virtually no training is required,as the previously learned skills (e.g., riding a bi-cycle and other transferable driving skills) arereadily retrieved (particularly for scooter drivers).However, for those with little previous driving ex-perience, training will need to occur across a num-ber of sessions so that the behavior becomes welllearned. Zanetti et al. (2001) have shown that oth-er procedural skills can be retained followingtraining even in patients with mild dementia.Their training protocol was 1 hour/day, 5 days/week for 3 weeks. These thoughts are offered pure-ly in a speculative manner to prompt further re-search in this area.

To return to the data in this study, it is impor-tant to recognize that this was a pilot study with avery limited sample size done in the context of rou-tine clinical care. Although the therapists reportedstandard protocols prior to the beginning of thestudy and adhered to these as much as possible,the demands of clinical practice do not always per-mit following the type of protocol that is necessaryfor rigorous research. This problem is further high-lighted by the fact that a pretraining PIDA is notroutinely done. Therefore, we do not know the




baseline driving performance of most of the partic-ipants—a significant flaw in this endeavor thatmust be remedied in future studies. Furthermore,posttraining driving performance was assessed byan OT blind to the training process only at one site(Sunnybrook). In contrast, at Baycrest the thera-pist involved in the training assesses posttrainingdriving performance. This is standard practice inmany clinical departments but may, in fact, haveimpacted scoring the PIDA at Baycrest. Cliniciansinvolved in training individuals in the use of powermobility devices should consider having a personblind to the training process administer the post-training assessment to ensure an unbiased evalu-ation of the driving. This is particularly importantgiven the incidence of accidents and crashes usingthese devices.

Three final comments about the sample anddata follow. First, we did not find an associationbetween age and driving performance nor a statis-tically significant difference in age between thetwo sites; however, it is worth noting that partici-pants at Baycrest were on average older by 5 years,a potentially clinically significant difference. Lit-erature in the area of rehabilitation for olderadults indicates the ‘‘young old’’ do better in re-habilitation than the ‘‘old old.’’ For example, Fal-coner, Naughton, Stasser, and Sinacore (1994)compared the performance of older adults involvedin rehabilitation following a stroke. The oldest pa-tients (75 and older) performed significantly poor-er in motor function skills and required more carethan ‘‘younger’’ groups (65–74 and ,65).

Second, most of the participants in this studyhad a number of medical issues. This study ana-lyzed only the medical diagnosis that most im-pacted mobility. Wells, Seabrook, Stolee, Borrie,and Knoefel (2003) suggest that disability amongseniors is multicausal and seniors with multiplemedical problems have a greater degree of func-tional impairment than those with a single medi-cal issue. Delving more deeply into the potentialimpact of comorbity on power mobility driving per-formance would be an interesting area for futureresearch.

Third, one study participant was later removedfrom the data set due to incomplete data, as earlierindicated. On administration of the posttrainingPIDA, this individual’s driving performance wasconsidered so unsafe that the testing was stoppedand the PIDA was not completed. Had the com-pleted items been prorated to generate a summaryscore, this individual would have achieved a scoreof 57.4 out of 100% on the PIDA, well below that ofall other participants. This participant received

the standard six training sessions, but due to ill-ness, training occurred over a period of 5 weeks, amuch longer duration than all other participantsat Sunnybrook. One wonders if a different protocolbased on our understanding of the acquisition ofprocedural memory would have resulted in im-proved driving performance.

Despite the limitations of the study, we are en-couraged by the fact that clinicians are developingstandard treatment protocols that lend themselvesto investigation through a research process. Thisstudy represents an ideal situation where re-searcher and clinician have worked together. Thisresearch would not have been done without theopenness of the clinicians. Collaborations, such asthis one, allow clinicians to change practice basedon empirical data and researchers to undertakeclinically meaningful research and further inves-tigations based on the needs of the clinic.

To conclude, successful and safe driving is a mul-tifactorial process involving personal, environmen-tal, and task-related characteristics. This studybegan investigation of the personal and trainingfactors that impact on driving performance. Pre-liminary findings suggest a number of factors con-tribute to driving performance including gender,type of device, and time period over which trainingis provided. Further research is needed to expli-cate these factors more clearly and should be un-dertaken in the context of what we know about theacquisition and retention of motor skills. Our find-ings suggest that power mobility driving trainingis one factor that can have a significant impact onoptimal power mobility driving performance. Asthe number of seniors who require power mobilitydevices continues to grow, it is imperative that re-searchers and clinicians together strive to estab-lish evidence-based training programs that willensure that individuals of all ages are providedwith a safe and appropriate means of independentmobility.

Acknowledgments: This research was conduct-ed to fulfill course requirements for the principalinvestigator (K.H.) while she was a student in theDepartment of Occupational Therapy, Universityof Toronto. The authors thank the residents of bothfacilities who participated in the study. Also,thanks are extended to occupational therapists JillDoyle, Tamara Giles, Melissa Lennox, Lydia Okta-ba, Barbara Skulko, Jennifer Suh, and occupation-al therapy assistants Abraham Azure, MarshaBerry, and Norma Daley of the Long-Term andVeterans Directorate at Sunnybrook and Women’sCollege Health Science Centre. In addition, K.H.




thanks her colleagues at the Assistive TechnologyClinic at Sunnybrook and Women’s College HealthCentre for their input and support. Statistics wereprovided by the Ontario Ministry of Health andLong-Term Care.

REFERENCES

Calder, C. J., & Kirby, R. L. (1990). Fatal wheelchair-relatedaccidents in the United States. American Journal of Phys-ical Medicine & Rehabilitation, 69, 184–190.

Chiodo, L. K., Gerety, M. B., Mulrow, C. D., Rhodes, M. C., &Tuley, M. R. (1992). The impact of physical therapy on nurs-ing home patient outcomes. Physical Therapy, 72(3), 168–175.

Dawson, D., Chan, R. A., & Kaiserman, E. (1994). Developmentof the power-mobility indoor driving assessment for resi-dents of long-term care facilities: A preliminary report. Ca-nadian Journal of Occupational Therapy, 61(5), 269–276.

Doyon, J., & Ungerlieder, L. G. (2002). Functional anatomy ofmotor skill learning. In L. R. Squire & D. L. Schacter (Eds.),Neuropsychology of memory (3rd ed., pp. 225–238). NewYork: Guilford Press.

Falconer, J. A., Naughton, B. J., Stasser, D. C., & Sinacore, J.M. (1994). Stroke inpatient rehabilitation: A comparisonacross age groups. Journal of the American Geriatric So-ciety, 42, 39–44.

Ferber, S., & Karnath, H. O. (2001). How to assess spatial ne-glect: Line bisection or cancellation tasks. Journal of Clin-ical and Experimental Neuropsychology, 23(5), 599–607.

Field, D. (1999). Powered mobility: A literature review illus-trating the importance of a multifaceted approach. AssistiveTechnology, 11, 20–33.

Furumasu, J., Guerette, P., & Tefft, D. (1996). The develop-ment of a powered wheelchair mobility program for youngchildren. Technology and Disability, 5, 41–48.

Gauthier, L., Dehaut, F., & Joanette, Y. (1989). The Bells test:A quantitative and qualitative test for visual neglect. Inter-national Journal of Clinical Neuropsychology, 11, 49–54.

Keith, R. A. (1997). Treatment strength in rehabilitation. Ar-chives of Physical Medicine and Rehabilitation, 78, 1298–1304.

Kirby, R. L., & Ackroyd-Stolarz, S. A. (1995). Wheelchair safe-ty—Adverse reports to the United States Food and Drug Ad-ministration. American Journal of Physical Medicine & Re-habilitation, 74, 308–312

Kirby, R. L., Coughlan, S. G., & Christie, M. (1995). Couldchanges in the wheelchair delivery system improve safety?Canadian Medical Association Journal, 153, 1585–1591.

Kwakkel, G., Wagenaar, R. C., Koelman, T. W., Lankhorst, G.J., & Koetsier, J. C. (1997). Effects of intensity of rehabili-tation after stroke: A research synthesis. Stroke, 28, 1550–1556.

Lange, M. L. (2000, October 9). Mobility training for powerwheelchairs. OT Practice, 21–22.

Law, M., Cooper, B. A., Strong, S., Stewart, D., Rigby, P., &

Letts, L. (1996). The Person-Environment-Occupation Mod-el: A transactional approach to occupational performance.Canadian Journal of Occupational Therapy, 63(1), 9–23.

Letts, L., Dawson, D., & Kaiserman-Goldstein, E. (1998). De-velopment of the Power-Mobility Community Driving As-sessment. Canadian Journal of Rehabilitation, 11(3), 123–129.

Miles-Tapping, C., & MacDonald, L. J. (1994). Lifestyle impli-cations of power mobility. Physical and Occupational Ther-apy in Geriatrics, 12(4), 31–49.

Molloy, D. W., Alemayehu, E., & Roberts, R. (1991). Reliabilityof a standardized mini-mental state examination comparedwith the traditional mini-mental state examination. Amer-ican Journal of Psychiatry, 148(1), 102–105.

Nisbet, P., Craig, J., Odor, P., & Aiken, S. (1996). ‘‘Smart’’wheelchairs for mobility training. Technology and Disabil-ity, 5, 49–62.

Reed, R. L., Yochum, K., & Schloss, M. (1993). Platform mo-torized wheelchairs in congregate care centers: A survey ofusage and safety. Archives of Physical Medicine and Re-habilitation, 74, 101–103.

Rosenburg, M. W., & Moore, E. G. (1997). The health of Can-ada’s elderly population: Current status and future impli-cations. Canadian Medical Association Journal, 157(8),1025–1032.

Routhier, F., Vincent, C., Desrosiers, J., & Nadeau, S. (2001).A new assessment for wheelchair user performance. In C.Marincek, C. Buhler, H. Knops, & R. Andrich (Eds.), Assis-tive Technology—Added value to the quality of life (pp. 189–193). Amsterdam: IOS Press.

Scott-Taplin, C., Smith, M., McLaughlin, J., & Mathews, T.(1989). The wheelchair stimulator: Powered wheelchair as-sessment and training. Canadian Journal of Rehabilita-tion, 3(1), 49–51.

Spreen, O., & Strauss, E. (1998). A compendium of neuropsy-chological tests: Administration, norms, and commentary(2nd ed.). New York: Oxford University Press.

Tefft, D., Guerette, P., & Furumasa, J. (1999). Cognitive pre-dictors of young children’s readiness for powered mobility.Developmental Medicine & Child Neurology, 41, 665–670.

Ummat, S., & Kirby, R. L. (1994). Nonfatal wheelchair-relatedaccidents reported to the National Electronic Injury Sur-veillance System. American Journal of Physical Medicine& Rehabilitation, 73, 163–167.

Wells, J. L., Seabrook, J. A., Stolee, P., Borrie, M. J., & Knoefel,F. (2003). State of the art in geriatric rehabilitation. Part I:Review of frailty and comprehensive geriatric assessment.Archives of Physical Medicine and Rehabilitation, 84, 890–897.

Wilkins, K., & Park, E. (1996). Chronic conditions, physicallimitations and dependency among seniors living in the com-munity. Health Reports, 8(3), 7–15.

Zanetti, O., Zaniere, G., Di Giovanni, G., De Vreese, L. P., Pez-zini, A., Metitieri, T., & Trabucchi, M. (2001). Effectivenessof procedural memory stimulation in mild Alzheimer’s dis-ease patients: A controlled study. Neuropsychological Re-habilitation, 11, 263–272.




APPENDIX. Power-Mobility Indoor Driving Assessment: Score Sheet

Documents

Power Mobility Driving Training for Seniors: A Pilot Studyfhs.mcmaster.ca/powermobility/hall_wheelchair_training_article.pdf · POWER MOBILITY DRIVING TRAINING FOR SENIORS 49 protocols,