A convenience sample of 19 newly discharged, manual wheelchair users with spinal cord injury was tested by measuring propulsion forces across 2 different terrains: flat tile (linoleum) and 5° slope. The participants were tested in their current wheelchair and a control wheelchair (Quickie GPV), optimized for posture and stability by experienced seating specialists. Data was collected determining wheel speed, acceleration, stroke angles and cadence using the Smartwheel™. The propulsion moment about the rear wheel hub was measured and the resultant force calculated. Propulsion performance showed that the users pushed with significantly lower frequency but higher propulsion peak forces in their own wheelchair compared to the optimized wheelchair on flat tile. However, there appeared to be little difference between own chair and control chair in propulsion parameters on the ramp.
manual wheelchair; propulsion; mobility; push-rim mechanics
In a study of participants discharged from all spinal injury units in UK conducted in 1997 (1,2) more than 40% had abandoned the wheelchair supplied at discharge in favor of one that was lighter and more "pushable". Although it is tempting to criticize the quality of wheelchair provision at discharge, an alternative interpretation of this data could be that significant changes in user needs occur after a few months of experience in their first wheelchair outside the hospital environment. This study measured a group of newly injured people with spinal cord injury to determine whether their wheelchair propulsion performance and stability requirements are optimized with their current wheelchair.
A convenience group of 19 manual wheelchair users with spinal cord injury was recruited from two spinal injury units within the UK. Their injury levels were C6(n=1), T1-T6 (n=8), T7-T12 (n=8), L (n=2); ASIA A-C and mean age was 33yrs ± s.d. 10yrs. All participants had been discharged from a spinal unit within the last year.
Each participant was tested by measuring propulsion forces across 2 different terrains using their current wheelchair, compared to a control wheelchair optimized for posture and stability by experienced seating specialists. The terrains were: flat tile (linoleum) and 5° slope. Propulsion kinetics were measured using an instrumented handrim (SmartWheel™ , Three Rivers Holdings LLC, AZ). The propulsion moment about the rear wheel hub was measured and the resultant force calculated. The wheel speed, acceleration, stroke angles and cadence were also measured by the Smartwheel™. Distance traveled was 10m and results determined for steady state pushing.
The users own wheelchair type ranged from minimally adjustable folding frame (n = 6) to multi adjustable, rigid ultra lightweight (n=13). Chair weight was 15.5kg ± s.d. 2 kg; and combined user/chair weight 88.0kg ± s.d. 13.7 kg. Own wheelchair rearward "tippiness" with hands on lap was 10.3°± s.d. 3.1° and hands on pushrim 8.7° ± s.d. 3.3°.
The control wheelchair was a rigid adjustable lightweight wheelchair (Quickie GPV, Sunrise Medical) . Chair weight was 13.8kg ± s.d. 1.2 kg. and combined user/chair weight 87.0kg ± s.d. 13.7 kg. Control wheelchair rearward "tippiness" with hands on lap was 8.4° ± s.d. 2.1° and on pushrim 6.7° ± s.d. 1.7° .
The rearward tippiness of the control chair was significantly less both with hands on lap and on the rim in push position (p<0.01). Users pushed at similar self selected speeds both on the flat (1.5 ± 0.3 m/s) and on the slope (0.7± 0.3 m/s). The users appeared to push at a higher frequency (0.94 ± 0.2 strokes/s) in the control chair than their own (0.88 ± 0.2 strokes/s), (p=0.06). However, there was no difference in the distance travelled per stroke (own chair 0.46 ± 0.06 m/stroke; control chair 0.45 ± 0.05 m/stroke) nor push angle (own chair 88 ± 11.7 degrees; control chair 85 ± 10.2 degrees) but there was less coasting in the control chair than their own chair (p=0.08). The similar velocity was thus explained by a difference in push forces and application through the stroke phase. The peak moment was greater in the users own chair (own chair 13.7 ± 4.3 Nm; control chair 11.7 ± 4.0 Nm), (p=0.07) and both peak resultant push force per stroke (own chair 69.3 ± 20.8 N; control chair 58.4 ± 20.0 N) and mean resultant push force (own chair 42.3 ± 11.4 N; control chair 35.7 ± 11.4 N) were significantly greater (own chair 13.7 ± 4.3 Nm; control chair 11.7 ± 4.0 Nm), (p=0.04 and 0.03 respectively). For this study group there was no significant difference in power per stroke in pushing the chair over the flat tile (own chair 40.2 ± 16.7 W; control chair 34.1 ± 15.4 W/stroke).
Although there was no significant difference in pushing up the slope between the control chair and the users' own chair there were, as expected, highly significant differences between pushing up the slope and on the flat tile. For this group of users the steady speed and cadence were halved on the slope, the push forces were three times greater, and the energy per stroke and average work done in pushing the chair up the slope were three times greater. However, there was no significant change in push frequency or push angle.
As the chair type varied between non-adjustable and folding frame to rigid types there were individual cases that demonstrated considerable difference between the control chair and their own chair. Some individuals pushed at greater speed with less energy per stroke in the control chair, whereas others with ultra lightweight rigid chairs would push faster at higher speed in their own chair but with similar stroke frequency but adopting larger push forces to enable longer coasting periods.
These results indicate that, in general, wheelchair provision has improved since our original survey. This is confirmed by the observation that most participants had rigid framed wheelchairs, whereas this was relatively rare in the survey in 1997. A repeat survey is being undertaken to study in greater depth the types of wheelchair now being provided in the UK. The differences in propulsion parameters between own wheelchair and the control were not dramatic between types of chairs and may not be the most important factor when choosing a wheelchair with this particular study group. The users reported that factors influencing their selection of wheelchair were based more on complex use situations such as transferring into cars and accessing the built environment. It was not clear whether higher repetition with lower force is any more beneficial than low frequency and higher push forces. The higher push frequency and lower forces in the control chair may be a result of unfamiliarity with the chair affecting stroke pattern. Propulsion performance showed that they pushed with significantly lower frequency but high propulsion peak forces in their own wheelchair compared to the optimized wheelchair on flat tile, however there appeared to be little difference in propulsion parameters on the ramp. Obviously the ramp proved to be an activity requiring much greater demands on propulsion, requiring 3 times as much power and nearly 5 times as much stroke impulse (area under the force-time curve) compared to the flat tile, regardless of the wheelchair used. Even though there was no coasting on the ramp, the distance traveled per stroke and active range of motion was the same as flat tile.
Funding for this project was provided in part by ASPIRE and the Big Lottery Fund.
Professor Martin Ferguson-Pell
ASPIRE Centre for Disability Sciences
Institute of Orthopaedics and Musculoskeletal Science
Stanmore HA7 4LP, UK
+44(1) 208 954 3188
This should be in the right column.