Clinical Rehabilitation 2010; 24: 867–886

A systematic review of wheelchair skills tests for manual

wheelchair users with a spinal cord injury: towards a
standardized outcome measure
Osnat Fliess-Douer, Yves C Vanlandewijck Department of Rehabilitation Sciences, Faculty of Kinesiology and Rehabilitation
Sciences, Katholieke Universiteit Leuven, Belgium, Galia Lubel Manor Zinman College of Physical Education and Sport
Science at the Wingate Institute, Netanya, Israel and Lucas HV Van Der Woude Center for Human Movement Sciences,
Center for Rehabilitation, University Medical Centre Groningen, University of Groningen, The Netherlands

Received 8th June 2009; returned for revisions 27th February 2010; revised manuscript accepted 27th February 2010.

Objective: To review, analyse, evaluate and critically appraise available wheelchair

skill tests in the international literature and to determine the need for a standard-
ized measurement tool of manual wheeled mobility in those with spinal cord injury.
Data sources: A systematic review of literature (databases PubMed, Web of
Science and Cochrane Library (1970–December 2009).
Subjects: Hand rim wheelchair users, mainly those with spinal cord injury.
Review methods: Studies’ content and methodology were analysed qualitatively.
Study quality was assessed using the scale of Gardner and Altman.
Results: Thirteen studies fell within the inclusion criteria and were critically
reviewed. The 13 studies covered 11 tests, which involved 14 different skills.
These 14 skills were categorized into: wheelchair manoeuvring and basic daily living
skills; obstacle-negotiating skills; wheelie tasks; and transfers. The Wheelchair Skills
Test version 2.4 (WST-2.4) and Wheelchair Circuit tests scored best on the Gardner
and Altman scale, the Obstacle Course Assessment of Wheelchair User
Performances (OCAWUP) test was found to be the most relevant for daily needs in
a wheelchair. The different tests used different measurement scales, varying from
binary to ordinal and continuous. Comparison of outcomes between tests was
not possible because of differences in skills assessed, measurement scales,
environment and equipment selected for each test. A lack of information regarding
protocols as well as differences in terminology was also detected.
Conclusion: This systematic review revealed large inconsistencies among the
current available wheelchair skill tests. This makes it difficult to compare study
results and to create norms and standards for wheelchair skill performance.

Introduction

Manual wheelchair mobility is important for a

Address for correspondence: Osnat Fliess-Douer, Katholieke large number of people, particularly those with
Universiteit Leuven, Faculty of Kinesiology and spinal cord injury. The majority of people with a
Rehabilitation Sciences, Department of Rehabilitation
Sciences, Tervuursevest 101, 3001 Leuven, Belgium. spinal cord injury (approximately 80%) are depen-
e-mail: douer@netvision.net.il dent on a wheelchair for their mobility for the rest
ß The Author(s), 2010.
Reprints and permissions: http://www.sagepub.co.uk/journalsPermissions.nav 10.1177/0269215510367981
868 O Fliess-Douer et al.

of their lives.1 To function independently, manual in the context of assessment. The current study,
wheelchair users must possess a variety of wheel- however, focuses mainly on wheeled mobility in
chair skills to be able to deal with the physical daily activities and social roles. Therefore, the
barriers they will encounter in various environ- objectives, analysis and outcomes of the two
ments in daily life.2 Manual wheelchair skill reviews are very much complementary.
performance of people with spinal cord injury The main objective of this review is to system-
is positively associated with activities and atically review, document, analyse and critically
participation.3 appraise the performance-based wheelchair
In this context, ‘wheelchair skill performance’ is skills tests for manual wheelchair users, especially
defined as: ‘The ability to move around and over- those with a spinal cord injury, currently available
come obstacles encountered when carrying out in the international literature. An added value of
daily activities or social roles in a self-propelled this study will be that it facilitates selection of
wheelchair’.4 the most suitable components from the existing
A ‘wheelchair skills test’ consists of various tasks wheelchair skills tests, in order to develop a stan-
to be performed by the candidate under standard- dardized test.
ized conditions. A validated and reliable wheelchair
skills test is necessary as a guiding instrument in the
rehabilitation process of people with spinal cord Methods
injury and those with lower limb impairments.
Such a tool can assist in making the appropriate
choice of skills to be trained in rehabilitation as A systematic review of the international literature
well as in the evaluation of training interventions. was performed. The search aimed for actual
Furthermore, a standardized and accepted wheel- performance-based hand-rim wheelchair skills
chair skills test could be used to develop standards tests. The databases used for selection of peer-
of wheelchair skills performance for individuals reviewed articles were PubMed, Web of Science
with different levels of impairment. and Cochrane Library (from 1970 to December
A review by Kilkens et al.5 mainly focused on a 2009). The database search and study appraisal
broad description and comparison of manual were conducted by the first and third author
wheelchair skills tests reported in the literature together in a systematic way. Only studies
between 1966 and 2001. The main conclusion of reported in English were selected.
that study was that there is no standard test to
measure wheelchair skills performance; most of
the tests have only been used in one or two studies, Search strategy
‘a fact that makes it impossible to compare study Peer-reviewed articles were selected using the
results’.5 The current study updates and completes keywords wheelchair(s) and measurement com-
Kilkens’ review, focusing on self-propelled wheel- bined with assessment. A second search using the
chair users with a spinal cord injury. keywords wheelchair(s) and rehabilitation, com-
Another more recent review by Mortenson et al.6 bined with mobility was preformed. To assure
identified and evaluated wheelchair-specific out- that all relevant literature was included, a final
come instruments, using the International Classifi- search was performed, using the keywords mobility
cation of Functioning, Disability and Health and wheelchair, while alternately adding the words
(ICF) definitions as a framework.7,8 This review skill, task, measurement, test, ADL, functional,
focused mainly on statistical properties of the instrument, performance, spinal cord injury, valid-
tests, which ranged from questionnaires to actual ity, reliability, pathology, behavior, activity, disabil-
performance tests, and it incorporated powered ity, assessment and quality of life. Initial study
wheelchairs as well. selection was based on title and on abstract when
A broad framework for mobility performance the title was not sufficiently detailed. References
assessment of wheelchair users was presented by given in relevant reviews and relevant publications
Routhier et al.4 According to Routhier et al.,4 the were also checked and examined. The outcome of
review by Mortenson et al. focused on mobility this search strategy is described in detail in Figure 1.
 Wheelchair skills tests for manual wheelchair users 869

Search 1: Wheelchair AND Search 2: Wheelchair AND Search 3: Wheelchair AND Mobility
Measurement Rehabilitation 595 studies
179 studies 2006 studies
+ skill (9 hits) OR, task (36), measurement
(28), test (224), ADL (15),
functional (122), instrument (19),
performance (91), SCI (52), validity (24),
Wheelchair AND Measurement Wheelchair AND Rehabilitation reliability (22), pathology (21), behavior (45),
AND AND activity (74), disability (99), assessment (98),
Assessment Mobility quality of life (60)
39 studies 428 studies 1039 studies (all together)

Checking based on title

1506 studies
Checking based on abstract &
Checking for double references

45 studies
43 studies From reference list of selected
studies

Checking for double

references

43 studies
In-depth reading, including only
studies testing wheelchair skills

40 studies
Including wheelchair skills Analysing including/excluding
criteria: only actual
performance-based WM skills
test
Included: 13 studies representing 11
different wheelchair skills tests Performing quality
assessment analysis11

6 studies without sufficient QAS 5 studies with sufficient QAS

Figure 1 Search strategy and the result of the search process. (QAS - quality assurance system).

Selection of criteria environments); (b) it was constructed for people

Since the focus of this review is on performance- using hand-rim wheelchairs and intended to
based wheelchair skills tests, especially for hand- assess wheelchair skills performances; (c) the
rim wheelchair users with spinal cord injury, the study sample also involved people with spinal
following selection criteria were used: cord injury; (d) available statistical data regarding
A test was selected if: (a) it was an observa- reliability and validity was presented. A test was
tional test (assessments in real or controlled excluded if: (a) it was constructed for persons
870 O Fliess-Douer et al.

using powered wheelchairs; (b) tests were per- were defined and assessed. For each aspect, the first
formed in a virtual environment; (c) assessment author and the third author had to agree whether it
was based on questionnaires or interviews was ‘strength’ or ‘weakness’. In case of disagreement,
(where subjective and retrospective characteristics the other authors were consulted. A test was con-
might strongly influence the outcome9), (d) tests sidered ‘stronger’ when it met more of the following
were focused on ‘body functions and structures’ aspects: (a) it tests necessary daily wheelchair skills
(measuring specific physiological and/or biome- in real life settings, rather than ‘lab’-based settings
chanical variables which do not comply with the (e.g. using own manual wheelchair rather than a
terms of ‘activity’ or ‘participation’ domains as standardized wheelchair); (b) it tests wheelchair
defined in the ICF7,8). skills performances of people with spinal cord
injury; (c) it assesses the differences in level of
wheelchair skills performances and/or tests the out-
come of an wheelchair skills intervention in a sensitive
Quality assessment and precise manner; (d) it presents quantitative
The quality of the selected papers (quality assur- data which facilitate comparison with other tests
ance system, QAS) was evaluated independently by results or previous trials; (e) it presents qualitative
the first and third authors according to a checklist data analysis, generated by experts, enabling trai-
for statistical review of general papers adapted ners to improve their teaching skills and can be also
by van Velzen et al.10 from Gardner and used to enhance test sensitivity; (f) it lasts a reason-
Altman.11 In case of disagreement of an item able amount of time (preferably no longer than 60
score, consensus was achieved through discussion. minutes); and (g) it presents a final test score.
For each selected wheelchair skills test, the fol-
lowing aspects were identified, described and
analysed:
Results
 The assessed wheelchair skills: How clear were
their specifications and the reason for their In total, 1506 articles were screened on basis of the
inclusion in the test? title (Figure 1). Checking for double references and
 The scaling and outcome parameters used to abstract relevancy rejected 97% of the findings.
assess the wheelchair skills performance A total of 43 articles were identified for further
 The feasibility and complexity of the test (dura- consideration. Forty studies involved different
tion, number of skills included, environment and forms of wheelchair skills examination, but not
wheelchair type, availability of test protocols) necessarily actual performance-based wheelchair
 The psychometric properties: the reliability as skills tests. Of those 40 studies, only 13 met the
well as validity information (if presented) was inclusion criteria.13–25 Most of the excluded studies
reviewed. Intra-class correlation coefficients were found to be ‘a list of tasks’, aiming to evaluate
(ICCs) 40.75 or k-values 40.75 are defined as an intervention or to detect differences between
very good reliability.12 Sensitivity to change, groups, rather than actual performance-based
ceiling and/or floor effects were also reviewed wheelchair skills tests. Two tests, which were
(if presented). included in previous literature reviews, were also
excluded: ‘Functional Evaluation in a Wheelchair’
Finally, a critical evaluation of strengths and (versions 1 and 2) instrument26,27 was excluded
weaknesses of the wheelchair skills tests was con- because it was designed as a questionnaire to be
ducted. Only studies with sufficient quality (QAS administered over time to consumers of seating-
60%10,11) were included in this phase of our study. mobility technology and not as an actual wheel-
The critical evaluation was based on the authors’ chair skills performance test. ‘Wheelchair Obstacle
opinion regarding the most suitable skills, scale, Course’28 was excluded because it was developed
environment and equipment that should be for people with a cerebral vascular accident only.
included in a standardized wheelchair skills test The remaining 13 studies represented 11
for person with spinal cord injury. Different aspects different wheelchair skills tests. In two cases,
 Wheelchair skills tests for manual wheelchair users 871

Kilkens et al.17,20 and Routhier et al.,14,15 the two Irregular surface was tested six times.14–16,18,19,22
studies of each author were considered as one since Overall, the tasks under this category lack sufficient
they presented the same wheelchair skills test in explanation.
both studies. Of these 11 tests, 9 were developed
in the last 10 years (1998–2007). Study designs
included cohort studies (8), cross-sectional studies Wheelie
(2) and a descriptive study (1). Out of these Only three tests measured wheelie skill16,19,22 as a
11 studies, 5 scored QAS above 6.5 separate task. The wheelie tasks included station-
points,13,14,16,17,22 having sufficient methodologi- ary, forward and backward and turn while wheelie.
cal quality following Gardner et al.11 None of the wheelie tasks were clearly explained.

Wheelchair skills assessment Making transfer

The 11 tests covered 14 different skills, which This was a repeated skill in many tests and it
could be categorized to wheelchair manoeuvring was subcategorized into ‘vertical transfer’ (up and
and basic daily living skills; obstacles negotiating down), ‘horizontal transfer’ and ‘other transfers’.
skills; wheelie and making transfers (Table 1). Transferring from the floor to the wheelchair is
more frequently tested than transferring to the
floor (5/3). In horizontal transfer, only one test
Wheelchair manoeuvring and basic daily living skill included a transfer into and out of a car.22
Level propulsion forward and backward is the
most frequent task tested. In three tests, the partic-
ipants had to cover the same distance when going Outcome and scaling
forward and backward.13,16,19,23 A variety of dis- Six out of 11 tests used a quality assessment
tances and methods were used but the reason for scale13,16,18,19,22,24 (i.e. level of independency
choosing the specific distance or method was not scale, ‘pass/fail’ scale, etc.), two used quantitative
mentioned in any of the tests. Manoeuvring was scales21,25 and only three tests combined both
tested four times14,17,19,22; chair turning was tested quality and quantity parameters.14,17,23 Four
in three tests16,19,22; Parking16,19 (parallel parking) tests used a final test score expression, but each
and sprint17,20 were tested once (in two different used a different type15,16,18,19 (Table 2).
studies).

Feasibility
Obstacle-negotiating skills The feasibility variables are presented in detail in
The variety of tasks under this category is the Table 2, only major aspects will be highlighted.
largest. Ramp was tested five times,13,17,18,23,24 Eight studies gave information regarding the dura-
but with different slopes and distances. For the tion of administering the test. Most tests lasted
kerb/pavement (curb/sidewalk) test 10 different between 30 minutes and 1 hour. The longest test
kerbs were tested (heights range 2.5–17.8 cm). duration was 1–1.5 hours,18 while the shortest
Only four tests used two different kerb was 15 minutes.23 Total skills number for each
heights13–15,23 in order to differentiate functional test varied from a minimum of five13 up to 61,22
abilities. Incline/slope was tested eight times, and when focusing on wheeled mobility perfor-
using a variety of inclines, with no consistency. mance, the number of skills varied from a minimum
Only two tests measured two different slopes of three21 up to 30.16 All tests were conducted in a
within the same test to compare abilities13,17,20; rehabilitation centre. Only 5 out of 11 tests sup-
two slope tasks were tested on a treadmill17,20; plied information regarding the wheelchair type
only one test measured crossing a slope.16 that was used during the assessment.14–17,19,20,25
Doorstep was tested six times14–21; none of the In three tests participants used their own wheel-
tasks used the same height of door threshold. chair,14–16,19 while in two tests standardized
872 O Fliess-Douer et al.

Table 1 Skills review

Category/subcategory Number The task

of tests
out of 11

1) Wheelchair manoeuvring and basic daily living skills

Level propulsion forward 9 Direction: forward13a,16b,17a,20a,18c,19b,21a,23a,24c,25a;
and backward backward13a,16b,19b,23a,25a
Distance: forward 6–50 m, backward – the same distance as going
forward except for one25 (3 m)
Duration: 180 s–6 min17,20,21
Speed: 0.56–1.11 m/s17.20
Method: level propulsion13,16,19,23,24; covering the greatest distance
possible in 6 min21; wheelchair-adjusted treadmill17,20
Street crossing18
Chair turning 3 Direction: turn in place left and right16a,22c; turn moving left and right16b; turn
three points left and right16c,19c
Parking 2 Parallel parking16c,19c
Manoeuvring 4 Figure of eight shape/slalom17,20a,19c
Narrow corridor14a
Wheelchair manoeuvring indoors and outdoors22c
Sprint 1 Propelling the wheelchair as fast as possible for 15 m17,20a
2) Obstacles negotiating skills
Ramp 5 Slope: 1:20–1:1213a,17,20a,18c,23b,24c
Direction: up and down13,23; up only17,20; unspecified18,24; two different
slopes in the same test13
Kerb/pavement 7 Height: 2.5–17.8 cm13a,14a,16a,18c,19a,22c,23b
Direction: ascent and descent16a,19a,22c; ascent only13a,14a,23b;
unspecified18c; two different kerbs in the same test13,14,16,23
Incline/slope 6 Height: 1:16–1:814b,18b,16b,19b; 3% and 6%17,20a
Direction: going up and down14,16,19,22; going up only17,20; cross slope16
Time: 10 s17,20
Doorstep 6 Height: 2.5–7.5 cm14a,16c,17,20a,18c,21a
Two different doorsteps in the same test14; request opening a door
before passing18,21
Irregular surface 6 Surface: carpet14c; gravel14c,16c,19c; unspecified18,19,22; two different surfaces
in the same test14,16
3) Wheelie
3 Wheelie – stationary16c,19c,22c; wheelie forward16c; wheelie backward16c;
wheelie – turn in place and while moving16b
4) Making transfer
Vertical transfer – up 4 From the floor/mat to the chair13a,22a,23a,24a; from bathtub to wheelchair22a
Vertical transfer – down 3 From wheelchair/chair to floor18c,22a,24b; transfer from wheelchair to
bathtub22a
Horizontal transfer 6 Bed transfer/treatment table/plinth17,20a,18c,23a,25a; toilet transfer18c,22a;
chair transfer (to another chair)21a,25a
Car transfer22b
Other transfers 3 Transfer out/into a wheelchair16c,19c; transfer from wheelchair to shower/from
shower to wheelchair22c

Clarity of explanation: avery clear; bquite clear; cunclear.

wheelchairs were used.17,20,25 Description of using the wheelchair previous to the first test
quality of the wheelchair varied widely. Only trial.14,16,19 Only 5 out of 11 tests described the
three tests provided information regarding the test protocols sufficiently clearly (i.e. enabling
period of time that the participants had been reproduction of test procedures).13,15,20,21,23
Table 2 Test analysis wheelchair skills tests – general summary

Test Objective Study design QAS Study Test design Outcomes Duration
population
Way of No. WC Test WC info Parameters Scale Express final
choosing skills/no. environment test score
the skill skills

Middleton, 200613
FIM – 5 To better delineate Patients were 7 SCI: 11 tetra, 28 Previous test by 4/5 SCI and rehabili- No info Quality 1–7: 7 complete Shown only No info
additional important functional tested within para ASIA A–C Harvey, were tation 2 units independence, 6 graphically
items differences between 72 h of WM for (D excluded) modified to in Sydney, modified indepen-
5-AML groups with different the first time N¼39 ensure suit- Australia dence, 5 supervi-
levels of neurological since injury, 32M/7F ability for tetra sion, 4 minimal
impairment 1 month, Age: R 22–35, assistance, 3 mod-
2 months, mean 28 erate assistance,
3 months and 2 maximal assis-
6 months later. tance, 1 total assis-
COHORT tance. Every skill
has its own speci-
fic remarks for
each point
Routhier, 2004, 200514,15
OCAWUP To evaluate WC user 17 experienced WC 7.5 SCI: 6; neuro- Consulting 10/10 The Institute de Their own WC Time (s) þ DE: 3 ¼ total Global DE scores No info
performance in users using 3 muscular dis- experts and Readaptation that they have quality success; are added to
potentially difficult different propul- ease: 4; CVA: scientific litera- en Deficience for at least 6 2 ¼ success with give a GSE
environmental situa- sion methods 5; amp: 2 after ture (based on Physique de months. using difficulty; from 0 to 30
tion, for all propulsion were assessed rehab min 1.5 Routhier15) Quebec 3 different pro- 1 ¼ partial failure;
methods and all cli- twice on the 10 years (IRDPQ) pulsion 0 ¼ complete
ents group obstacle course. N¼17 methods failure.
COHORT 10M/7F
Age: R 24–64,
mean
50.9  12
16
Kirby, 2004
WST-2.4 To evaluate the effect Subjects were 10 Amp: 62; CVAþ Pilot study þ17 30/50 Rehab. centre Conventional: Quality 0 ¼ fail, Total WST 27  9.3 min
of WC skills training videotaped per- TBI: 52; MSK: WC users sug- 119, 1 ¼ pass, scores:
for WC users under- forming 50 skills 20; SCI: 34; gested the lightweight: NA ¼ not GAS ¼ total
going initial rehabilita- twice. The order able-bodied addition or 34 ultraligh: 15 applicable, goal
tion, to screen of skills was volunteers: modification of WC was in NG ¼ not a goal. attainment
subjects before according to dif- 129 obstacles the patient’s
wheelie training, to ficulty level and N¼298 use for 2 days
compare the perfor- the natural M140/F158 prior the test
mance of manual groupings. If the Age: R 17–88,
WC and push-rim first attempt mean 8  22.5
power-assisted WC, failed a second
to evaluate the effect try was allowed.
of WC skills training Subjects were
for occupational oriented to the
therapists student at test expecta-
second year tions. COHORT

(Continued )
Table 2 Continued

Test Objective Study design QAS Study Test design Outcomes Duration
population
Way of No. WC Test WC info Parameters Scale Express final
choosing skills/no. environment test score
the skill skills

Kilkens, 2002, 200417,20

Wheelchair To assess mobi- Longitudinal WC 10 SCI: para: 53 Tasks adapted 8/9 8 Rehab centres Hand-rim rigid frame Quality, time Ability score: yes/no, Scores are Each task
Circuit lity during and Circuit was (18 inco.); from the mobi- in Holland WC was available and HR þ3 items could be described in duration is
after clinical performed at tetra: 21 lity-related in 2 seat widths scored as partially table. described
rehabilitation start (T1) and (9 inco.) tasks used by able 0.5), task per- No general score but not
mainly for at the end (T3) N¼74 Harvey23 formance time (s) that combines the overall
research pur- of inpatient M51/F23 (2 skills), and % all 3 measure- duration
poses but also functional Age: R 18–65 HR (2 slops task) ments
for clinical set- rehabilitation.
tings (longitu- COHORT
dinal study)
Stanley, 200318
WUFA To develop a Subjects were 3 Not described A panel of 6 PT 9/13 The Baltimore No info Quality 1 (total dependence) - Overall indepen- 1.0–1.5 h
performance- videotaped clearly. experts in Veterans and 7 (completely inde- dence (similar
based func- once while Including: CP, rehab. þ Affairs Center pendent). A score to the FIM)
tional mea- performing 13 SCI, spina survey of 30 of 6–7 includes a
surement tool skills. bifida, CVA, manual WC specific time
that is easy to DESCRIPTIVE M, Amp and users requirement for
administer and more task completion
score, which N¼101
can be applied M83/F18
to the wide Age: male
range of indivi- 42.4  12,
duals who use female
manual WCs 45.7  13
Kirby, 200219
WST-1 Improvement for Subjects were 7.5 Amp:11; CVA: 4; Skills were based 13/33 Rehab. centre Their own WC; partici- Quality 0–2: 0¼fail, 1¼partial Overall % of 29 min
individual WC videotaped MSK: 3; SCI: on previous lit- pants were using completion, success
users, out- while perform- 3; neuromus- erature review the WC min 2 2¼successful
come measure ing 33 skills cular: 3 þ occupational days before the
for rehab. pro- twice. The N¼24 therapists’ test; time using
grammes, protocol is M16/F8 opinion WC min 3 weeks.
assist in test- similar to Age: R 18–85, only 10 use the
ing research, WST-2.4. mean 19  59 WC outside;
developing COHORT method of propul-
new sion (2 hands, one
technologies hand, one foot)

(Continued )
Table 2 Continued

Test Objective Study design QAS Study Test design Outcomes Duration
population
Way of No. WC Test WC info Parameters Scale Express final
choosing skills/no. environment test score
the skill skills

Cress, 200221
WC-PFP test To measure the Tasks were per- 3 MS: 4; CVA: 2; Adaptation of the 3/11 A rehabilitation No info Quantity scores Scores were trans- Reported mean 40 min
ability of formed serially polio: 1; CS-PFP facility (weight, dis- formed to a scale þ SD in a
people using and partici- arthritis: 1; tance and from 0 to 100 table, only
manual WC to pants were TBI: 2; SCI: 8 time) in 4 (¼highest perfor- general score
perform tasks asked to work N¼18 domains: mance in the test
which are at maximal M13/F5 Upper body population)
important for exertion yet to Age: R 18–67 strength, flex-
living inde- pace them- ibility, endur-
pendently selves to com- ance, balance
plete as many and
tasks as possi- coordination
ble. COHORT
Taricco, 200022
Measure- To evaluate the Scoring is done 9.5 SCI: para: 67;Input from the 23/61 8 Italian SCI No info Quality 1–5 scale: from Mean score for 30–50 min
ment impact of according to tetra: 33. 81 literature units inability to com- each domain
scale rehabilitative direct observa- had a trau- and specia- plete the task to at baseline
VFM interventions tion of patient matic lists. (more completing the and follow-up
on the func- performance, aetiology details were task without
tional status of excluding 1 N¼100 M77/F23 reported in difficulties
SCI patients task that is Age: R 14–76 previous
simulated and articles)
2 that are col-
lected by inter-
view (none of
those are
related to
mobility). VFM
was adminis-
tered at the
beginning and
the end of the
rehabilitation
programme.
COHORT

(Continued )
Table 2 Continued

Test Objective Study design QAS Study Test design Outcomes Duration
population
Way of No. WC Test WC info Parameters Scale Express final
choosing skills/no. environment test score
the skill skills

Harvey, 199823
Assessing To quantify Each subjects per- 4 SCI: para 20. Skills typically 5/6 No info (most No info Quality (based 1–6 scale: Level of Median and IQR 15 min
mobility in the mobility formed the test No info on learned in phy- likely rehabili- on also on independence, levels of mobi-
paraplegic of WC- twice at the same numbers siotherapy tation centre) time to time and task com- lity attained on
dependent day, assessed by Age: mean programmes complete plexity, according each task
paraplegics. 2 different asses- 45.6  16.8 the task) to each task
sors. COHORT
Gans, 198824
TAMP Providing com- Test was adminis- 4.5 Orthopaedic sur- No info 4/32 Rehab centre No info Quality Four different scales Mean kappa Less than
prehensive tered by PT/OT. gery 6; closed in 4 dimensions: 1h
clinical eva- Instructions and head injury 6; Assistance (5 scale
luation of demonstration muscle dis- points, from inde-
physical were given prior ease/atrophy pendence to total
function and to each item. 3; SCI 5; CP 6; dependence)
motor per- Breaks were pro- other miscella- Approach (2 scale
formance. vided as needed. neous 14 points: general
To examine Order of tasks: N¼40 technique used to
motor per- from fine to dres- M14/F26 complete the task.)
formance in sing, transfer and (20 adults, 20 Pattern (2 scale
sufficient mat mobility and children) points: clinically
detail to finally mobility. Age: mean important move-
assist with Tests were video- 25.6  19.5 ment patterns in
treatment taped and ana- selected fine
planning and lysed at a later motor activities
to ade- date. CROSS-
and in gate)
quately SECTIONAL
Proficiency (3 scale
describe
points: movement
and identify
control, accuracy
meaningful
of extremity place-
clinical
ment and aspects
changes in
of coordination)
motor
abilities.

(Continued )
Table 2 Continued

Test Objective Study design QAS Study Test design Outcomes Duration
population
Way of No. WC Test WC info Parameters Scale Express final
choosing skills/no. environment test score
the skill skills

Jebsen, 197025
Time mea- To measure Subjects were 3.5 100 healthy No info 6/31 No info 16 and 18 inch seat Time (s) Mean time of 30 min for
surement various timed while per- 18 patients width, large each subtest, healthy
in a stan- aspects of forming each 10 hemiparesis; wheels in rear no general 60 min for
dardized patient task. They were low limb amp: (standard WC) score patients
test mobility by advised to per- 4; peripheral
measuring form each task neuropathy 1;
the time as quickly as hip fracture 1;
necessary possible. polio 1; SCI 1
for indepen- Healthy sub- N¼100
dent com- jects: short M50/F50
pletion of practice session Healthy: 20 from
each of the was allowed each age
group before testing group
subsets. started. Age: Patient:
Subjects with mean 49.7
disabilities:
were tested on
2 occasions, 1–
4 days apart.
COHORT

WC, wheelchair; QAS, quality assurance system; WC-PFP, Wheelchair Physical Functional Performance; SCI, spinal cord injury; VFM, Valutazione Funzionale Mielolesi;
TAMP, Tufts Assessment of Motor Performance; WM, wheeled mobility; WST-2.4, Wheelchair Skills Test version 2.4; DE, degree of ease; GSE, global score of ease;
CVA, cerebrovascular accident; TBI, traumatic brain injury; MSK, musculoskeletal; GAS, Goal Attainment Score; CP cerebral palsy; MS, multiple sclerosis; PT, physiothera-
pist; OT, occupational therapist; FIM, Functional Independence Measure; WUFA, Wheelchair Users Functional Assessment; CS-PFP, Continuous Scale Physical Functional
Performance Measure; tetra, tetraplegic; para, paraplegic; comp, complete; inco, incomplete; Amp, amputation; M, male; F, female; R, range;
878 O Fliess-Douer et al.

One test added photos to the article.15 Some studies their injury, and the rehabilitation process is now
addressed that a copy of the entire test and score more likely to be in the community, in day hospital
sheets are available upon request. or the home environment.29 It is therefore recom-
mended that skills associated with the daily needs
of this population will be included in wheel chair
Psychometric properties skills tests. Some of the reviewed skills do not
An analysis and summary of the psychometric comply with this recommendation. For example, in
properties of the 11 tests is presented in Table 3. the level propulsion skill, the participant has to cover
Seven out of 11 tests presented information on the same distance when going forward and back-
validity.13–19,22 For most tests, content validity ward, as if it has the same necessity in daily life. In
was based on the involvement of health profes- the Wheelchair Circuit,17,20 three skills out of eight
sionals in the development of the test and on lit- require the use of a treadmill. In contrast, the 10
erature studies. Three tests14,16,18 based the wheelchair skills of the Obstacle Course
content validity on wheelchair users’ suggestions; Assessment of Wheelchair User Performances
two tests15,17 based the content validity on related (OCAWUP)14,15 are most relevant for daily needs
instruments. Construct validity was determined in of wheelchair users. In addition, this test is the only
seven tests,13–17,19,25 correlating the new test with one that associates skill selection with the average
the Functional Independence Measure (FIM) height of pavements in Quebec City as well as corre-
instrument, and/or correlating the test result with lating the level of inclines to the National Building
age, gender, lesion, type of wheelchair or time Code of Canada. It is recommended that the skills to
using the wheelchair before the first test trial. be included in a test are chosen according to a large
Only tests with sufficient reliability were survey among experienced wheelchair users, sorting
included in this review. Test–retest reliability was out the most essential skills for daily life in a wheel-
assessed in five tests.14,16,19,21,25 Time between the chair. It is also advised that slopes, heights and other
two trials varied from 1 day up to 28 days. Inter- measurements selected for the test are associated
rater was evaluated in eight tests13–20,23,24 and with norms, standards and architectural accessibility
intra-rater was examined in four tests.16–20 codes, regional as well as universal.
Only three studies reported sensitivity to
change over time in wheelchair skills perfor-
mances.13,17,22 In these studies the tests were
administered at the beginning and the end of the Scaling and outcome parameters
rehabilitation programme. The above three studies, Scaling method is a crucial aspect and has a
which assessed changes in wheelchair skills perfor- direct impact on the responsiveness of the test.
mances over time, were the only studies that indi- Some of the reviewed tests use only qualitative
cated ceiling effect for the paraplegic group and scales. Specifically, the ‘pass/fail’ ranking, used
higher responsiveness (detecting functional in the Wheelchair Circuit,17,20 Valutazione Fun-
changes over time) in tetraplegic persons.13,17,20,22 zionale Mielolesi (VFM)22 and Wheelchair Skills
Test version 2.4 (WST-2,4)16 tests, cannot guaran-
tee a clear and sensitive distinction between levels
of performance in groups but especially in individ-
Strengths and weaknesses
The results of this evaluation are presented in uals. Some tests evaluate performance ability by
Table 4, and will be incorporated in the discussion assessing only the ‘degree of independency’. This
part of this article. may not reflect the actual performance level, since
a person may be able to perform a certain wheel-
chair skill independently, but if it requires an
unreasonable amount of time or high energy
Discussion cost, the person might choose not to perform this
skill in daily life. The most sensitive quality scales
Today, individuals with spinal cord injury spend were found to be the independency scale of the Five
less time in the hospital during the acute phase of Additional Mobility and Locomotor (5-AML)13
Table 3 Psychometric properties of the wheelchair skills tests

Test Validity Reliability Sensitivity to

change over time
Content/ Construct/concurrent Test–retest Intra-rater Inter-rater
face

Middleton, 200613
FIM 5 addi- 1 A No info No info k range Although written ‘high
tional items; 0.82–0.96 responsiveness to
5-AML change over time’ it
refers to the tetra-
plegic group, while a
ceiling effect is men-
tioned in the paraple-
gic group
Routhier, 2004, 200514,15
OCAWUP 1,2,3 §A ICC range 0.74–0.99 No info ICC 0.98 No info
B (FIM) r ¼ 0.84, P  0.05 (time) (time)
k range 0.09–1.00 (GSE) k range 0.82-
ICC 96 (overall score) 0.96 (GSE)
ICC
0.97(ove-
rall score)
Kirby, 200416
WST-2.4 2 þA r ¼ –0.434, P50.001. ICC ¼ 0.90 ICC ¼ 0.96 ICC ¼ 0.97 No info
A r. P ¼ 0.38. ^A
P ¼ 0.0112.*A
B (the global assessments of
WC users’ therapists and
admission and discharge
FIM scores) 0.394, 0.38,
0.31
Kilkens, 2002, 200417,20
WC circuit 3 Ability score (T1–T3):  A. No info ICC ¼ 0.98 ICC ¼ 0.97 All 3 scores showed
þ A (only at T3); B (FIM strong significant
mobility score, peak power improvement
output and Vo2 peak) between T1 and T3.
Performance time score  A The SRM was 0.6
(only at T3); þ B (FIM for ability score, 0.9
mobility score, peak power for performance
output, and peak oxygen time, and 0.8 for
uptake) physical strain score
Physical strain score:  A
(only at T3); B (peak
power output and peak
Wheelchair skills tests for manual wheelchair users

oxygen uptake inversely

related)
879

(Continued )
Table 3 Continued
880
Test Validity Reliability Sensitivity to
change over time
Content/ Construct/concurrent Test–retest Intra-rater Inter-rater
face

Stanley, 200318
WUFA 1, 2 No info No info ICC ¼ 0.78 ICC ¼ 0.96 No info
19
Kirby, 2002
WST-1 1 þ A negative r P50.05; r ¼ 0.65 r ¼ 0.96 r ¼ 0.95 No info
^A P ¼ 0.0085; B
P ¼ 0.008 (the OT global
O Fliess-Douer et al.

rating)
Cress, 200221
WC-PFP test 3 B (Sickness Impact Profile ICC 0.96 No info No info No info
scale) significant correla-
tion (r ¼ –0.45) only for
bathing and dressing
(self-related health and
upper body domain)
Taricco, 200022
Measurement 3 A *A No info No info Was evalu- Scores had significantly
scale VFM B (Barthel Index, QIF, ated in improved for tetra-
FIM) range 0.67–0.88 previous plegia and high-level
phase of paraplegia between
this project T1 and T3
(published
in Italian)
Harvey, 199823
Assessing mobil- 1 No info No info No info k range No info
ity in paraplegic 0.82–0.98
Gans, 198824
TAMP No info No info No info No info k range No info
0.65–0.83
Jebsen, 197025
Time measure- No info *A (healthy people vs WC r 0.85–0.99 (time) No info No info No info
ment in a stan- users)
dardized test

FIM, Functional Independence Measure; 5-AML, Five Additional Mobility and Locomotor test; OCAWUP, Obstacle Course Assessment of Wheelchair User
Performances; SRM, standardized response mean; GSE, global score of ease; WC, wheelchair; ICC, intraclass correlation coefficient; WST, Wheelchair Skills
Test; WUFA, Wheelchair Users Functional Assessment; WC-PFP, Wheelchair Physical Functional Performance; SCI, spinal cord injury; VFM, Valutazione
Funzionale Mielolesi; TAMP, Tufts Assessment of Motor Performance.
Validity legend:
Content/face: 1, involvement of health professional in development; 2, involvement of WC users in development; 3, based on related instrument.
Construct: A – significant factor.
Concurrent: B – correlation with a related instrument/physiological parameters. T1 – first test trial; T3 – last test trial (mostly at the end of the rehabilitation
programme).
Variables signs: , lesion level (tetraplegia/paraplegia); *, diagnostic category (SVA, CP, Amp, SCI, etc.); þ, age; ^, experience of WC use; §, propulsion
methods.
Table 4 The critical evaluation (only tests with QAS above 6.5)

Test Protocol þ design Summary of strengths Summary of weaknesses

13
Middleton, 2006
5-AML Patients were tested (b) Only SCI, male and female (a) Only 4 WM skills, suitable for rehabilitation use; No info
within 72 h of mobilizing in a reasonable ratio regarding the wheelchair
in a WC for the first (e) Quality assessment of 7 (c) No info about differences in ASIA; suitability of all tasks
time since injury, 1 levels for each skill to tetraplegia; ceiling effect on paraplegic in general
month, 2 months, 3 Popularity of the FIM (d) No info regarding the quantity data
months and 6 months (e) Quality scale based on level of independency
later. COHORT (f) No info about feasibility
(g) No expression of general score
Routhier, 2004, 200514,15
OCAWUP 17 experienced WC users (a) 10 WM skills, suitable for (a) Including motorized wheelchair; Skills are not including
using 3 different propul- daily use; Using their own transfers and wheelie backwards
sion methods were WC (b) Not only for SCI
assessed twice on the (d) Time scale (f) No info about feasibility
10 obstacles of the (e) Quality assessment of 4
OCAWUP. COHORT levels
(g) GSE – global scores are
added to give a GSE from
0 to 30
Kirby, 200416,19
WST-1 and - Subjects were videotaped (a) Using their own wheel- (a) Skills were not described in details; it is hard to base a
2.4 while performing 50 chair, many WM skills (i.e. future test on the articles
skills twice. The order left and right for the same (b) Not only for SCI
of skills was according skills) (c) No info regarding sensitivity for change, no report
to difficulty level (from (f) It takes about 30 min to regarding different level of WM in the group of SCI
easy) and according to complete the test (d) No quantity data
the natural groupings. If (g) Total WST scores – allow (e) Quality scale not sensitive enough (only pass/fail and NA
the first attempt failed a comparison with previous and not a goal
second try was allowed. trial
Subjects were oriented
to the test expectations.
COHORT
Kilkens, 2002, 200417,20
WC Circuit Longitudinal. WC Circuit (b) SCI only (big sample (a) Only 8 WM skills, 3 skills require treadmill, using transfer
was performed at start (c) Longitudinal, assessing to treatment table; not suitable for everyday needs,
(T1) and at the end (T3) changes over time rather lab examination
of inpatient functional (d) Quantity data based only on the sum of two skills; HR
rehabilitation. COHORT of 2 skills only
(e) Quality scale only of 3 level (pass/fail/partially of 3 skills
only
(f) Each task feasibility is described but not all test
Wheelchair skills tests for manual wheelchair users

feasibility
(g) No expression of a total score
881

(Continued )
 882

Table 4 Continued

Test Protocol þ design Summary of strengths Summary of weaknesses

22
O Fliess-Douer et al.

Taricco, 2000
VFM A direct observation of (a) Test can be used in differ- (a) 23 WM skills out of 61, not measuring only WC skill
patient performance, ent settings (home, performances
excluding 1 task that outpatient) (d) No quantity data
is simulated and 2 (b) Only SCI, male and female (e) Quality scale based on level of independency
that are collected by in a reasonable ratio, para- (f) Between 30 and 50 min (not clearly described)
interview (none of plegic and quadriplegic, big (g) No expression of a total score
those are related to sample
mobility). VFM was (e) 1–5 quality scale
administered at the
beginning and the
end of the rehabilita-
tion programme.
COHORT

5-AML, Five Additional Mobility and Locomotor; OCAWUP, Obstacle Course Assessment of Wheelchair User Performances; VFM, Valutazione Funzionale
Mielolesi; WM, wheeled mobility; WST, Wheelchair Skills Test; SCI, spinal cord injury; FIM, Functional Independence Measure; GSE, global score of ease
(a) Quantity data for comparison with other tests or previous trails.
(b) Quality analysis (on field and/or video-recorded performance, analysed by experts) enable trainers (PT, PE and peers models) to improve their teaching
skills and also for the enhancement of the sensitivity of the test.
(c) Feasibility – duration of the test: no longer than 30 minutes.
(d) Express final test score.
(e) Test everyday wheeled mobility skills rather than ‘lab’ examinations: using skills that are needed in everyday life, using their own manual wheelchair.
(f) WM test for person with SCI.
(g) Assessing the differences in level of wheelchair skills performances of SCI (optimal WM, good WM, moderate WM and poor WM) and testing the
outcome of an WM intervention (or learning phase during the rehabilitation process), sensitively and precisely.
 Wheelchair skills tests for manual wheelchair users 883

test and the performance’s quality scale of the instance, when performing slalom the participant
OCAWUP test.14,15 Some of the reviewed tests has to drive forward, and in this case the need to
use quantitative scales (e.g. time score, heart test level propulsion forward is questionable.
rate, etc.). Such scales allow simple and objective The chosen environment for all the tests reviewed
comparison with other trials and tests. For in the present study was a rehabilitation centre,
instance, in the Wheelchair Circuit test17,20 the where it is easy to recruit participants, equipment
highest statistical significant score was the ‘perfor- as well as examiners, but whether an accessible
mance time score’, providing evidence for rehabilitation centre is the most suitable environ-
improvement in wheelchair skills performances ment for evaluating wheelchair skills that are
over time. Nevertheless Routhier et al.14 men- required for daily life is debatable. Furthermore,
tioned that one should be careful with the use of patients in rehabilitation are rather immature in
time in short tasks (e.g. ‘gets over 2 cm threshold’). their wheelchair skills performances. Nevertheless,
If a wheelchair user takes 15 seconds instead of 10 a distinction should be made between the location
(an extra 50%) to achieve a task, it does not nec- in which the reviewed studies took place for their
essarily mean that the wheelchair user has difficul- validation and the ability to administer them out-
ties. To conclude, in a future standardized test it is side rehabilitation centres for clinical purposes.
suggested to combine both quantity measures and The type of wheelchair has an important impact
sensitive quality scales. on carrying out a task, especially when testing
level of performance (e.g. type and wheelchair
quality, tyre pressure, maintenance status, etc.)
Using one’s own wheelchair during the test
Feasibility and complexity would probably better reflect performance level
In general, the duration of a test is dictated by but threatens the stability of the test. Aiming the
the number and complexity of the skills included, wheelchair skills test for daily activities and social
but in some of the reviewed tests there is a contra- roles, the use of the participants’ own wheelchairs
diction to this concept. On the one hand, some is preferred, accompanied by a full description of
tests evaluate many skills in a relatively short the wheelchair used.
time (the VFM22 test assessed 61 skills within Accessibility to the test protocols is important in
30–50 minutes and the WST-2.416 assessed 50 order to understand the instructions and settings
skills within 27 minutes). An explanation to this of the tasks. Some of the reviewed studies lack a
observation is that some tests do not assess all the full description of the tasks, which may explain the
skills they contain, giving the observer the freedom multiplicity of tests. It is advised that each skill
to choose which skill to evaluate from the pro- should be well described in the article. It is also
posed list of skills. This involves a subjective recommended that the reader is given easy access
judgement of the observer, eliminating skills that to the full test protocols (e.g. through a website16).
are ‘not a goal’ for the specific participant, and This will invite rehabilitation professionals and
that therefore might reduce test sensitivity. other researchers to use the same test, and may
It also makes a comparison between different test- promote the development of a universal wheel-
ing difficult, if not impossible. On the other hand, chair skills test.
some tests contain fewer skills but last longer (the
Wheelchair Users Functional Assessment18 test
assessed 13 skills in 1–1.5 hours). Reducing the
number of selected skill shortens test duration Psychometric properties
but may subject the test to ceiling effect (e.g. When conducting research in a rehabilitation
5-AML21 test which has a ceiling effect on para- domain, setting the goals according to precise def-
plegics). The number of skills included in the test is inition of terms is critical as a guideline for devel-
a compromise between the need to cover many oping the tool and choosing the most appropriate
aspects of wheelchair skill performances and min- skills. A clear definition of terms, specifically of
imizing overlaps between the selected skills. More the outcome expressions (e.g. wheelchair user’s
complex skills inherently include ‘subskills’; for ‘function’ or ‘wheeled mobility’), is the foundation
884 O Fliess-Douer et al.

for ensuring a good content validity. Many of the Despite this comment, the same skills were used
reviewed tests lack a clear terms’ definition; for to evaluate paraplegic and tetraplegic clients.
instance, the Wheelchair Users Functional Spinal cord-injured wheelchair users may demon-
Assessment study18 aimed to ‘develop a perfor- strate the best wheelchair skill performance, and
mance-based functional measurement tool that could be the benchmark to reach in term of wheel-
needed to incorporate home and community chair skills. It is advisable to first develop sensitive
wheelchair skills’. One of the skills included in wheelchair skills test specifically for the spinal
the Wheelchair Users Functional Assessment is cord injury population, with adaptations for the
upper and lower dressing, but because of the two subgroups (tetraplegic and paraplegic). Such
absence of definition of terms, it is unclear why a test might reduce ceiling or floor effects, and
dressing is included as a wheelchair skill. In con- could be generalized later on to the entire wheel-
trast, a good example where the objective of the chair user population.
test as well as definition of terms are presented can Most of the reviewed articles aim to assess the
be found at OCAWUP15: ‘the purpose of the reliability and validity of the test, but do not derive
OCAWUP is to assess and document the mobility norms and standards for wheelchair skill perfor-
performance of manual and motorized wheelchair mances. Previous studies do not present data that
users in potentially difficult environmental situa- can be compared with results from other studies.
tions. Environmental situations are obstacles when As revealed from the current review, there is no
they limit the social participation of an individ- broadly accepted wheelchair skills test. Such an
ual’.15 Another example of unclear term definition ‘ideal’ instrument should measure the level of
selected aspects of wheelchair skill performance,
is a confusion between ‘ability’ (e.g. sprint task)
be methodologically strong and practically feasi-
and ‘motor skill’ (e.g. ‘transferring from the floor
ble. Once a standardized wheelchair skill test has
back into the wheelchair’), which is found in many
been developed, accepted worldwide and fre-
tests. It is important to understand the differences
quently used, it will enable norms and standards
between these two terms and to select tasks that
to be created for wheelchair skill performance.
clearly meet the definition of motor skill.
In a population of people with spinal cord Study limitations and strength
injury, an individual might function at minimal This review was limited to actual performance-
(C3) or maximal (L3) levels. Ceiling and floor based wheelchair skills tests published in peer-
effects are the result of lack of precision and the reviewed articles that were available in English
ability of an instrument to detect meaningful only. Since this review aimed to serve as a base
changes in level of performance at the upper or for developing a standardized manual wheelchair
lower ends of the scale. If individuals have reached skills test for those with spinal cord injury, assess-
the maximum rating in their first trial, more subtle ment tools that were constructed exclusively for
improvement will not be reflected. The idea is to other disabilities were not included, although
establish one test for all wheelchair users, similar some lessons may be also drawn from those instru-
to the Functional Independence Measure30 that ments and studies.
has been widely adopted by the rehabilitation The detailed skill-specific review along with gen-
eral analysis is a particular strength of the study,
community as a tool for use with diverse patient
since it points out the important aspects of each
populations. Tools that generalize wheelchair user
test and facilitates a selection of the most suitable
populations, however, may fail to differentiate
components, enabling the development of a stan-
between levels of performance. In the WST-2.4,16
dardized wheelchair skills test, relying on the
although ceiling effect was not reported, the use of strengths of existing measurement tools.
quality scale only could probably lead to ceiling
effects (e.g. assessing ‘50 meter level propulsion
forward’ with a pass/fail scale will definitely Conclusion
produce a ceiling effect for all paraplegic clients).
The VFM22 was the only test that addressed the As shown in this review, many wheelchair skills
need to subdivide the spinal cord injury group. tests are applied to the measurement of wheelchair
 Wheelchair skills tests for manual wheelchair users 885

skill performance using a large variety of skills, References

equipment and outcome measures. The main con-
clusion is that there is no one broadly accepted 1 Post MWM, Asbeck van FW, Dijk van AJ,
wheelchair skills test allowing comparison Schrijvers AJ. Services for spinal cord injured:
of study results. It is hoped that this review availability and satisfaction. Spinal Cord 1997;
will appraise the need for agreement among 35: 109–15.
researchers as to which skills must be included as 2 Gouvier WD, Cottam G, Webster JS, Beissel GF,
well as standardization of the use of one measure- Wofford JD. Behavioral interventions with
ment instrument. Cooperation between profes- stroke patients for improving wheelchair
navigation. Int J Clin Neuropsychol 1984; 6:
sionals in the development of this ‘ideal’ test, as
186–90.
well as the promotion and endorsement of such a 3 Kilkens OJ, Post MW, Dallmeijer AJ, van
tool, is desirable in order to develop universal Asbeck FW, Van Der Woude LH. Relationship
norms and standards for wheelchair skill between manual wheelchair skill performance and
performances. participation of persons with spinal cord injuries
1 year after discharge from inpatient rehabilita-
tion. J Rehabil Res Dev 2005; 42(suppl 1): 65–73.
Competing interests 4 Routhier F, Vincent C, Desrosiers J, Nadeau S.
None declared. Mobility of wheelchair users: a proposed
performance assessment framework. Disabil
Rehabil 2003; 25: 19–34.
5 Kilkens OJ, Post MW, Dallmeijer AJ, Seelen HA,
Contributors Van Der Woude LH. Wheelchair skills tests: a
OFD initiated the study, designed the study, systematic review. Clin Rehabil 2003; 17: 418–30.
conducted the databases search, scored methodo- 6 Mortenson WB, Miller WC, Auger C. Issues for
logical quality, wrote the paper, and was guaran- the selection of wheelchair-specific activity and
tor. LvdW initiated the study, designed the study, participation outcome measures: a review. Arch
monitored progress and contributed to the discus- Phys Med Rehabil 2008; 89: 1177–84.
sion of the results. YCV initiated the study, 7 World Health Organization. Towards a common
designed the study, monitored progress and con- language of function, disability and health: ICF.
Geneva: WHO, 2002, 1–23.
tributed to the discussion of the results. GLM con- 8 World Health Organization. ICF: international
ducted database search, scored methodological classification of functioning, disability and health.
quality and contributed to the discussion of the Geneva: WHO, 2002.
results. 9 Washburn RA, Montoye HJ. The assessment of
physical activity by questionnaire. Am J Epidemiol
1986; 123: 563–76.
10 van Velzen JM, van Bennekom CA, Polomski W,
Slootman JR, van der Woude LH, Houdijk H.
Clinical messages Physical capacity and walking ability after lower
limb amputation: a systematic review. Clin
 Applying general wheelchair skill tests to a Rehabil 2006; 20: 999–1016.
spinal cord injury population leads to ceiling 11 Gardner MJ, Altman DG. Statistic with
confidence: confidence intervals and statistical
effect and lack of sensitivity to minor
guidelines. London: BMJ Books, 1989.
changes among the paraplegic group. 12 Rudner LM, Shafer WD. Reliability. ERIC
 The OCAWUP is the most relevant test for Digest. College Park, MD: ERIC Clearinghouse
wheelchair users daily needs, but stills lacks on Assessment and Evaluation, 2001.
some skills. 13 Middleton JW, Harvey LA, Batty J, Cameron I,
 Combining the best and most relevant com- Quirk R, Winstanley J. Five additional mobility
ponents of existing tests, as outlined in this and locomotor items to improve responsiveness
review, may lead to the development of a of the FIM in wheelchair-dependent individuals
superior test. with spinal cord injury. Spinal Cord 2006; 44:
495–504.
886 O Fliess-Douer et al.

14 Routhier F, Desrosiers J, Vincent C, Nadeau S. with spinal cord injury: a new standardized
Reliability and construct validity of an obstacle measurement scale. Gruppo Interdisciplinare
course assessment of wheelchair user performance. Valutazione Interventi Riabilitativi. Arch Phys
Int J Rehabil Res 2005; 28: 49–56. Med Rehabil 2000; 81: 1173–80.
15 Routhier F, Vincent C, Desrosiers J, Nadeau S, 23 Harvey LA, Batty J, Fahey A. Reliability
Guerette C. Development of an obstacle course of a tool for assessing mobility in wheelchair-
assessment of wheelchair user performances dependent paraplegics. Spinal Cord 1998;
(OCAWUP): a content validity study. Technol 36: 427–31.
Disabil 2004; 16: 19–31. 24 Gans BM, Haley SM, Hallenborg SC, Mann N,
16 Kirby RL, Dupuis DJ, MacPhee AH et al. Inacio CA, Faas RM. Description and interobser-
The Wheelchair Skills Test (version 2.4): measure- ver reliability of the Tufts Assessment of Motor
ment properties. Arch Phys Med Rehabil 2004; 85: Performance. Am J Phys Med Rehabil 1988; 67:
794–804. 202–10.
17 Kilkens O, Dallmeijer A, de Witte L, van der 25 Jebsen RH, Trieschmann RB, Mikulic MA,
Woude L, Post M. The Wheelchair Circuit: con- Hartley RB, McMillan JA, Snook ME.
struct validity and responsiveness of a test to Measurement of time in a standardized test of
assess manual wheelchair mobility in persons with patient mobility. Arch Phys Med Rehabil 1970;
spinal cord injury. Arch Phys Med Rehabil 2004; 51: 170–75.
85: 424–31. 26 Mills T, Holm MB, Trefler E, Schmeler M,
18 Stanley RK, Stafford DJ, Rasch E, Rodgers MM. Fitzgerald S, Boninger M. Development and
Development of a functional assessment measure consumer validation of the functional evaluation
for manual wheelchair users. J Rehabil Res Dev in a wheelchair (FEW) instrument. Disabil
2003; 40: 301–7. Rehabil 2002; 24: 38–46.
19 Kirby RL, Swuste J, Dupuis DJ, MacLeod DA, 27 Mills TL, Holm MB, Schmeler M. Test-retest
Monroe R. The Wheelchair Skills Test: a pilot reliability and cross validation of the functioning
study of a new outcome measure. Arch Phys Med everyday with a wheelchair instrument. Assist
Rehabil 2002; 83: 10–18. Technol 2007; 19: 61–77.
20 Kilkens O, Post M, van der Woude L, 28 Webster JS, Cottam G, Gouvier WD, Blanton P,
Dallmeijer A, van den Hauvel W. The wheelchair Beissel GF, Wofford J. Wheelchair obstacle
circuit: reliability of a test to assess mobility in course performance in right cerebral vascular
persons with spinal cord injuries. Arch Phys Med accident victims. J Clin Exp Neuropsychol 1989;
Rehabil 2002; 83: 1783–8. 11: 295–310.
21 Cress ME, Kinne S, Patrick DL, Maher E. 29 Ditunno JF, Formal CS. Chronic spinal cord
Physical functional performance in persons using injury. N Engl J Med 1994; 330: 550–6.
a manual wheelchair. J Orthop Phys Ther 2002; 30 Guide for the uniform data set for medical rehabili-
32: 104–13. tation (including the FIM instrument), version 5.1.
22 Taricco M, Apolone G, Colombo C, Filardo G, Buffalo, NY: State University of New York at
Telaro E, Liberati A. Functional status in patients Buffalo, 1997.