Ultralight wheelchair part failures are associated with sensor-monitored road shocks: a pilot study

Abstract

Wheelchair part failures and repairs have significantly increased over the last decade, leading to severe consequences for wheelchair users. Servicing these devices by wheelchair repair technicians has reduced part failures. However, no tools or technologies have been developed to support servicing in practice. To inform servicing events, risk factors affecting wheelchair quality and reliability need to be identified. This pilot study tracks wheelchair usage for a week in the community for eight ultralight manual wheelchair users and assesses the relationship between usage variables and user-reported part failures over 20 months. The participants’ preferences for using smart technology for wheelchair servicing were evaluated. At least 73 wheelchair part failures and two adverse consequences were reported. Data analysis indicated associations between part failure frequency, usage variable of road shocks, wheelchair maintenance frequency, and the user’s demographic characteristics of training status and transportation. Six participants favored using smart technology for wheelchair servicing. This study’s findings encourage the development of usage monitoring technology and failure prediction models to support technician-led servicing and prevent wheelchair failures and user consequences.

Introduction

“The wheelchair is my leg, my chair and my everything,” quotes Sammy from Kenya, aged 32, in the Global Report on Assistive Technology published by the World Health Organization (WHO) and United Nations International Children’s Emergency Fund (UNICEF) (World Health Organization, 2022). For people with mobility limitations like Sammy, wheelchair devices help develop a personal identity, commute, perform activities of daily living, and integrate into society (Borg & Khasnabis, 2008). Unfortunately, wheelchair parts break frequently, risking the safety of 85 million wheelchair users globally (World Health Organization, 2023).

Wheelchair evaluation studies conducted worldwide have demonstrated wheelchair part failures within 2–3 months of wheelchair use (D’Innocenzo et al., 2021; Mhatre et al., 2020; Reese & Rispin, 2015; Rispin et al., 2018). Cross-sectional study evidence collected by the Model System Center on Spinal Cord Injury in the United States shows that wheelchair part failures have gone up from 45% in 2009 to 88% in 2020 (Henderson et al., 2020; McClure et al., 2009; M. Toro et al., 2016; L. Worobey et al., 2012, 2014). Among wheelchair users, Veterans are the hardest-hit population as they have witnessed the highest failure rate; about 88% of Veterans experience at least one part failure in six months (Henderson et al., 2020). A Veteran Affairs (VA) audit reported that Veterans miss medical appointments and suffer physical hardships due to wheelchair repair times extending more than two months (Veterans Health Administration, 2018). Frustration with repair times has led to the enactment of a consumer right-to-repair law in Colorado state of United States, raising the risk of safety hazards due to inappropriate wheelchair servicing (Consumer Right To Repair Powered Wheelchairs, 2022). Failure of critical parts like casters, brakes and rear wheels results in wheelchair tipping and the user falling out of the chair (Gaal et al., 1997; Xiang et al., 2006a), and can lead to wheelchair breakdown. About 40–65% of tips result in injuries or bruises (Gaal et al., 1997; Xiang et al., 2006a), leading to a downward spiral of economic and health outcomes. Users miss workdays while out-of-pocket repair bills range from $6 to $4000 (L. A. Worobey et al., 2022). Also, public health expenditures are doubling (Xiang et al., 2006b). Wheelchair failure constrains users to their beds (M. Toro et al., 2016; Veterans Health Administration, 2018) or an inadequate loaner chair (Boccardi Alyssa et al., 2022; Hogaboom et al., 2018), so failures are associated with pressure injuries (bed sores), depression and rehospitalization (Hogaboom et al., 2018). Hence, the WHO recommends safe and effective wheelchairs for the equitable right to mobility and to prevent the marginalization of people with disabilities in society (World Health Organization, 2022).

Among many factors related to wheelchair regulations, services, training and quality assurance, one area that is key to reducing failures and improving the reliability of wheelchairs, but often overlooked, is servicing. Servicing includes maintenance, repair and replacement of wheelchair parts according to policy (U.S. Centers for Medicare & Medicaid Services, 2022). Maintenance includes upkeep of the parts to stay functional and last longer before they fail and need repair. Repair is needed after a part partially fails to function as intended; repair activities include inspection followed by adjusting, cleaning, or greasing parts. Replacement is necessary when the part can no longer serve its intended function or cannot be repaired. When servicing activities are performed by repair technicians, part failures and consequences are reduced (Hansen et al., 2004a; Mhatre et al., 2021a). A model study in the wheelchair maintenance literature is a randomized controlled trial conducted by Hansen et al. with n=216 manual wheelchair users in Sweden. Repair technicians performed regular maintenance and part replacement as needed (intervention), which eliminated wheelchair failures and user consequences. In the control group, however, failures and consequences remained unchanged (Hansen et al., 2004a). Servicing by technicians was further proven effective in a secondary data analysis study (Mhatre et al., 2021a) of n=6470 caster failures reported in a U.S. wheelchair repair registry. The registry contains maintenance and replacement instances performed for the casters during repair visits for other broken wheelchair parts (A. James et al., 2021). Servicing significantly reduced caster failures that could cause user injury or damage to other wheelchair parts. Unfortunately, despite such research evidence, no tools, technologies or standards exist for wheelchair servicing by repair technicians.

Wheelchair servicing evidence calls for practices that exist for other mobility technologies like cars. For example, when the engine light sensor detects overheating and starts flashing on the car’s dashboard, servicing by an authorized shop technician is needed. Also, in some places, policy mandates regular car emission inspections by authorized technicians. The lack of such technologies and practices for wheelchairs could be attributed to policy. The Center for Medicare and Medicaid Services (CMS) in the United States only reimburses service providers for repairs and replacement by their technicians after a wheelchair part fails and if there is continued medical necessity for the user (U.S. Centers for Medicare & Medicaid Services, 2022). This reactive servicing approach leaves wheelchair users at constant risk of encountering part failures in the community and suffering adverse consequences until the technicians come for an inspection visit followed by a servicing visit(s). As the replacement of parts effectively reduces failures (Hansen et al., 2004a; Mhatre et al., 2021a), can the wheelchair industry leverage the latest tools and technologies to get informed of impending part failures and replacement events that are reimbursable. To build such technology, risk factors contributing to wheelchair fatigue need to be identified and modeled toward diagnosing faulty part behavior, and ultimately, failure. Cross-sectional studies indicate that failure measures and adverse consequences are associated with self-reported variables – demographic factors (user age, working status, years since injury) (L. Worobey et al., 2014), wheelchair type (manual or power), presence of seat functions (Mhatre et al., 2021a; L. Worobey et al., 2012, 2014), type of insurance (L. Worobey et al., 2012), travel distance (Henderson et al., 2020) and type of terrain (Mhatre et al., 2020; L. A. Worobey et al., 2021). In laboratory testing studies, wheelchair shocks and environmental humidity are the factors that degrade wheelchair condition (Mhatre et al., 2020; Wang et al., 2010). Since community evaluation studies have been cross-sectional in design, the longitudinal progression of risk factors, as well as failure measures (e.g., time to critical failure modes, effect or cost of related consequences), have neither been recorded nor modeled toward predicting impending failures.

Objectives

This pilot study’s aim is to test the feasibility of the usage data-driven wheelchair servicing concept with active wheelchair users in the community. The primary objective is to evaluate the relationship between wheelchair usage measures like travel distance, speed and shocks, and failure measures like failure frequency. The study’s central hypothesis is that wheelchair usage measures are correlated with failure measures. If the usage-failure association exists, we need to implement it through a servicing technology. A secondary objective of the study was to assess participant’s preferences for a new technology for wheelchair servicing including digital applications and sensors.

Methods

Study Design and Ethics

A descriptive, cross-sectional study design was implemented for 20 months to address the study objectives. Data collection on user demographic characteristics, wheelchair design characteristics, and participant preferences on servicing technology was conducted using surveys. Wheelchair usage was monitored for a week using a sensor. Study procedures occurred at the University of Pittsburgh and data analysis was conducted at The Ohio State University. The STUDY20100451 was approved and overseen by the Institutional Review Board of the University of Pittsburgh. Informed consent was obtained from all participants.

Study Participation

The study’s participation criteria are shown in Table 1. Ultralight wheelchair users were prioritized for the study since they are highly active in the community and experience a higher frequency of part failures. Recruitment was pursued through the Pitt+Me recruitment portal, the Human Engineering Research Laboratories (HERL) wheelchair user registry, posting study flyers at affiliated clinical facilities, and word of mouth for users participating in ongoing studies at the institute. Wheelchair users who expressed interest were contacted by phone and screened against the inclusion and exclusion criteria.

Table 1.

Study selection criteria.

Criteria	Inclusion	Exclusion
Population	Age>18 years, Wheelchair as a primary means of mobility	Cognitive impairment not allowing full participation in study tasks such as concentrating or understanding and responding to survey questions appropriately
Wheelchair Type	Ultralight manual wheelchairs

Study Procedures

Participants went through three study visits for data collection. Visits were scheduled in a private clinical room at the University of Pittsburgh’s Department of Rehabilitation Science and Technology.

During the first visit, survey administration, weight measurement, wheelchair inspection, and sensor installation were conducted. Surveys included demographic information about the participant, wheelchair manufacturer/model, part failures, maintenance, and repairs observed over the past year. Weight measurements with and without the wheelchair were measured on a wheelchair scale. Weight on the casters with the participant in the wheelchair was recorded. Wheelchair inspection of frames, casters, brakes, drive wheels, and axles was conducted using instructions noted in the validated Wheelchair Maintenance Assessment Tool (W-MAT) (M. L. Toro et al., 2017). Tire pressure and wheelchair part dimensions were measured. The sensor was mounted on the wheelchair’s rear axle, and wheelchair usage parameters were recorded. Participants were shown how to turn off and remove the sensor at the end of the week and return it via mail. If they could not perform this activity, a second visit for sensor retrieval would be completed. After eight months, a third visit was required to collect information on part failures and participants’ preferences for wheelchair servicing using new technology.

Surveys

Demographic variables included age, sex, racial identity, education, presence of familial or caregiver support, employment status, difficulty walking 100m, occupation, annual income, medical care expenses, number of years spent in the current wheelchair, training status on skills and maintenance, typical travel distance/day, and transportation use. Wheelchair manufacturer, model, maintenance activity routine, and failures in the past year were recorded. When failures were reported, participants were asked whether the following adverse consequence(s) occurred: no consequences occurred, stranded, injured, missing work/school, or missing medical appointment (M. Toro et al., 2016). In the second survey, at t=8 months, maintenance preferences using technology, in addition to failure and consequences since the first visit were recorded. Participants’ willingness to buy and use a new servicing technology, and other payer suggestions were noted. Both study surveys took 15–20 minutes on average to complete. Surveys are included in Supplemental Materials.

Sensor Instrumentation and Data Collection

Previous research on wheelchair caster quality testing standards demonstrated that outdoor shocks can be recorded and replicated onto laboratory-based testing equipment to reproduce caster failures seen in the community (Mhatre et al., 2020). This study extends this validated approach of data collection and employs refined instrumentation. The Arduino Nano BLE Sense replaced the off-the-shelf sensor from the previous study. The D-size alkaline batteries were replaced with a 3.7V, 6000mAh Lithium-polycarbonate battery. A bubble level on the sensor package ensured the installation was parallel to the ground. Wheelchair usage data from the sensor was collected on the SD card. The Arduino incorporates a 9-degree-of-freedom inertial measurement unit. Instantaneous accelerations that occur when the wheelchair wheel strikes an obstacle were recorded by the accelerometer at 100Hz. The magnetometer monitors a magnet on the rear wheel and computes travel distance and speed. The sensor’s battery capability allows seven days of data collection. Fig 1 shows the sensor and sensor installation on an ultralightweight wheelchair. A foam was placed between the sensor and the axle, and the sensor was zip tied to the axle such that it stays on top of the axle. The bubble level on the sensor was checked to ensure the orientation of the install.

Figure 1.

Sensor (left). Rearview of the ultralight wheelchair showing the sensor installed on the rear axle (right)

Laboratory trials with a wheelchair user and two able-bodied users were conducted to demonstrate the sensor’s feasibility of collecting mechanical usage parameters that are associated with wheelchair breakdowns according to experimental and cross-sectional studies (Henderson et al., 2020; Mhatre et al., 2020; M. Toro et al., 2016; L. Worobey et al., 2012, 2014; L. A. Worobey et al., 2022). Fig 2 shows these parameters for in-lab trials performed on three types of typical travel surfaces.

Figure 2.

Computing usage factors of travel distance, speed and shocks

Data Analysis

Descriptive statistics (mean ± standard deviation) were performed in Microsoft Excel on user characteristics, wheelchair-related variables, failures, consequences and participant preferences for new technology (Microsoft Corporation, 2023). Part failures were classified by risk as reported in a previous study (Mhatre et al., 2019). High-risk failures included those that could cause the user to get injured or other wheelchair parts to be damaged. Others were grouped as low-risk.

The sensor data was retrieved from the SD card and loaded into data analysis software. First, travel distance and speed were computed from magnetometer data on wheel rotations. Referring to methods from wheelchair standards development (Mhatre et al., 2020), vertical acceleration values greater than five times the RMS were computed and designated as fatigue-inducing shocks. Prior to correlation testing, outliers and normality using Shapiro-Wilk test were assessed. Scatter plots using the two variables were evaluated to understand the nature of relationship. If the plot showed a linear relationship, correlations between the continuous usage and failure data were assessed using Pearson’s correlation coefficient. Correlation coefficient values range from −1 to +1, with 0.00–0.10 interpreted as negligible correlation, 0.10–0.39 as weak correlation, 0.40–0.69 as moderate correlation, and above 0.70 as strong correlation.(Schober et al., 2018)

Descriptive, shock and correlation analyses were performed in Microsoft Excel, MATLAB R2022a and SPSS 28.0 respectively (IBM Corp., 2021; Microsoft Corporation, 2023; The MathWorks Inc., 2022). Since this was a feasibility testing study with a smaller sample size, a less conservative α-level of 0.1 was selected.

W-MAT condition ratings were assessed. Maintenance patterns were computed. The frequency of maintenance activity reported by participants was ordered with no/not possible = 0, daily = 5, weekly = 4, monthly = 3, quarterly = 2, and yearly = 1, and a cumulative total maintenance score for each participant was computed. Correlations between the ordinal maintenance score and training status were tested using point-biserial correlation, and maintenance score and failure measures were tested using Spearman’s rho.

Results

Recruitment Results

Twenty-three participants expressed interest in the study, seventeen were screened and fifteen were eligible for the study. Eight participants attended the first study visit between April and August 2021.Seven of them completed the full study procedures in April 2022 as one participant was injured during the study while performing daily life activities and could not return the sensor. Table 2 shows the participants’ demographic information.

Table 2.

Study participant characteristics.

Participant Characteristics	n (%) or mean ± SD
Age	42.57 ± 8.638 years
Weight	162.01 ± 38.41 lbs
Weight on casters	48.9 ± 25.19 lbs
Percentage of user weight on casters	27.2 ± 9.4%
Gender
Female	2 (25.0)
Male	6 (75.0)
Racial or ethnic identification
Black/African American	1 (12.5)
White	7 (87.5)
Education
Advanced Degree (Master, PhD)	4 (57.1)
College or University Degree	3 (42.9)
Employment status
Full-time	3 (42.9)
Part-time	1 (14.3)
Homemaker/Full-time parent	1 (14.3)
Unemployed	2 (25.0)
Difficulty walking a long distance (100 meters/10 yards/about a football field)
Severe difficulty/Cannot walk 100 meters	6 (85.7)
Some difficulty	1 (14.3)
Hours out of bed	13.57 ± 4.685
Hours per week spent working in a paying occupation	1.25 ± 21.665
Living situation
With spouse	2 (28.6)
Alone	2 (28.6)
With relatives	3 (42.9)
Hours per week spent in recreational activities	12.50 ± 29.517
Income
15,000-20,000	1 (14.3)
35,000-50,000	3 (42.9)
50,000-75,000	1 (14.3)
75,000 or more	2 (28.6)
Estimated amount paid last year for medical care expenses
1,000–2500	3 (42.9)
Less than 1,000	4 (57.1)
Years using the current wheelchair
1 year	2 (28.6)
2 – 3 years	4 (57.1)
5 years+	1 (14.3)
Manufacturer – model of current wheelchair (anonymized)
Manufacturer 1 – Ultralight model 1	5 (57.1)
Manufacturer 2 – Ultralight model 2	3 (42.9)
Trained in wheelchair maintenance	4 (57.1)
Trained in wheelchair skills	5 (71.4)
Satisfaction with wheelchair
Quite satisfied	2 (28.6)
Very satisfied	5 (71.4)
Distance travelled each day in wheelchair
1–5 miles	3 (42.9)
Up to 1 mile	4 (57.1)
Use of public or private transportation
2–3 times a week	3 (42.9)
Never	4 (57.1)
Transportation used
Car	6 (85.7)
Bus	1 (14.3)
Person who performs wheelchair maintenance activities
Professional/service provider	2 (28.6)
User	3 (42.9)
User with assistance from a caregiver	2 (28.6)

Part Failure Outcomes

At least 73 wheelchair part failures were reported over 20 months. Participants experienced 2.5±2.0 high-risk failures and 5.3±5.6 low-risk failures. High-risk failures observed were broken axle lock, caster wheel, caster fork, axle bearings, brakes, and push rims, and loose brakes. Low-risk failures observed were worn-out tires, seat, cushions, foot supports, leg supports, lateral supports, trunk supports, and sagging upholstery. Table 3 shows the failures encountered during the study. Tires and brakes saw more than 2 failures for 4 participants. No frame fractures, cushion punctures, broken or loose fasteners, caster bolt fractures, or axle-bearing fractures were reported.

Table 3.

Wheelchair part failures at the two visits. Failures reported during the second visit are specified separately.

Wheelchair part failures	n (%)
Worn-out tires or tubes
Once	3 (42.9)
Once (second visit)	4 (57.2)
Twice	3 (42.9)
More than 2 times	1 (14.3)
More than 2 times (second visit)	1 (14.3)
Broken spokes*
Once	1 (14.3)
Broken axel bearing*
Once	1 (14.3)
Broken wheel rim or push rim*
Once	1 (14.3)
Flat tire
Once	1 (14.3)
More than 2 times (second visit)	1 (14.3)
Broken axle lock*
Once	1 (14.3)
Twice	1 (14.3)
Worn out caster tires
Once	2 (28.6)
Broken caster wheel*
Once	1 (14.3)
Broken caster bearings*
Once	1 (14.3)
Twice	1 (14.3)
Once (second visit)	1 (14.3)
Broken caster bolts or nuts*
Once	1 (14.3)
Broken caster fork*
Once	1 (14.3)
Once (second visit)	1 (14.3)
Loose brakes
Once	2 (28.6)
Once (second visit)	3 (42.9)
Twice	2 (28.6)
More than 2 times	2 (28.6)
Broken brakes*
Once	2 (28.6)
Once (second visit)	2 (28.6)
More than 2 times	1 (14.3)
Loose foot supports, leg supports, arm supports, lateral supports, and/or trunk supports
Once	1 (14.3)
Twice	1 (14.3)
Worn-out foot supports, leg supports, lateral supports, and/or trunk supports
Once	1 (14.3)
Sagging seat upholstery
Once	1 (14.3)
Once (second visit)	3 (42.9)
Sagging back upholstery
Once	1 (14.3)
Worn-out seat
Once	3 (42.9)
Once (second visit)	2 (28.6)
Worn-out back cushion
Once	1 (14.3)

Adverse consequences were experienced from high-risk failures. The participant who reported a broken axle lock during the first visit missed work, school, or medical appointments. The participant, who reported a broken caster wheel and fork during the first visit, was left stranded. Another participant demonstrated high usage characteristics but encountered only one high-risk failure. This participant was trained, technically skilled, and compliant with maintenance. One participant did not encounter high-risk failure. Their usage characteristic was an outlier; they propelled using one foot and climbed obstacles and ramps using backward propulsion, i.e., rear wheels first. Hence, they were omitted from the usage-failure correlation analysis.

Usage-failure relationship

Table 4 displays the wheelchair usage measures recorded in the study. Between failure measures and the user’s demographic characteristics, maintenance training status was associated with brake failure frequency (r_pb=.79, n=7, p=.04). Public transportation use correlated negatively with total part failures (r_pb=−.82, n=7, p=.02) and tire failures (r_pb=−.8, n=7, p=.03).

Table 4.

Wheelchair usage recorded for n = 7 study participants.

Usage measures	Mean ± SD	Correlated Failure Measure(s)	Correlated User Demographic Characteristic(s)
Travel distance	8589 ± 5955.86 m	None	Maintenance training status (rpb=−.84, n=6, p=.03) Total user and wheelchair weight (r=−.87, n=6, p=.02)
Speed	0.71 ± 0.24 m/s	None	Skills training status (rpb=.74, n=6, p=.09)
Shocks	8.32 ± 2.58 g	Total Failures (r=.74, n=6, p=.09)	Load on front casters (r=.85, n=6, p=.03)

Participant Feedback on Wheelchair Servicing Technology

Six participants were interested in receiving regular notifications on servicing events via a smartphone app. Three of them expressed interest in buying the technology. Two of them reported they would pay $25 and $100 out-of-pocket, and the third noted that the purchase would depend on the technology features. Additionally, six participants reported that insurance should cover this technology and one noted that manufacturers should include this technology with the wheelchair.

Wheelchair Condition Assessment Outcomes

During the first visit, the W-MAT inspections found complete and impending wheelchair part failures. One wheelchair had more than two broken spokes. Four wheelchairs had tires in inappropriate condition; two had deflated tires below recommended pressure, one had flat tires, and one had tires with bald tread. Flat tires resulted in the disengagement of the axle lock or brake. Two casters were rusted, creating rolling resistance. All caster stems except one experienced rolling resistance. One chair veered to the right when allowed to roll freely. One participant experienced caster shimmy. The part failures were not reported by the participants in the survey.

Maintenance-related Outcomes

Fig 3 shows the maintenance activity patterns practiced by six participants for 17 tasks. The cumulative score for maintenance patterns noted in Fig 3 was 37.14 ± 15.35 out of a total 85, which is possible if a participant did all 17 tasks daily. Participant training in maintenance was correlated with maintenance patterns (r_pb=.74, n=7, p=.056). Maintenance patterns were significantly correlated with total part failure frequency (r_s=.91, n=7, p=.003). Among maintenance conducted by participants, brake inspection/adjustment was the most frequently performed activity followed by tire pressure check/inflation, dirt removal from axles, and inspection of cushions and handrims.

Figure 3.

Regular maintenance activities conducted by study participants.

Discussion

Servicing is essential for wheelchairs to ensure user safety and higher reliability of wheelchair products. Studies investigating technician-led wheelchair servicing have demonstrated a reduction in part failures (Hansen et al., 2004b; Mhatre et al., 2021b). This evidence has prompted an investigation into data-driven wheelchair servicing that can inform stakeholders of servicing events and support this pilot study. Associations between risk factors and failure measures tested in this study show that community-based wheelchair shocks and part failure frequency are correlated. This feasibility demonstrated in this pilot with a small number of active wheelchair users should be tested further with a larger sample size at a suitable statistical confidence level. This correlation finding coincides with the findings observed in laboratory-based wheelchair quality standards research (Fried, 2022; Mhatre et al., 2020) and a cross-sectional survey study (L. A. Worobey et al., 2022). Such relationships support future model-based prediction of failure and servicing events based on mechanical usage risk factors measured longitudinally in the community and user and wheelchair characteristics.

Prediction models can form the backend of technologies and digital platforms to convey wheelchair condition or faulty behavior to stakeholders and alert them for just-in-time servicing intervention. The development of new servicing technologies was upvoted by study participants. Six participants favored using and buying smart technology for servicing, upvoting the development of servicing information technology. This outcome coincides with a recent qualitative study which showed that 75% of participants were interested in using maintenance technology incorporating a smartphone app that connects with repair personnel (Boccardi et al., 2022). Even providers called for active servicing interventions using the latest technologies in a recent mixed methods study (Ruffing et al., 2022). Such findings encourage the development of servicing technology along with research on servicing models and interventions.

As found in a previous survey study, travel distance did not correlate with failure measures (Henderson et al., 2020) while shocks did. This indicates that travel miles may include rough terrain where wheelchairs will experience more failures than wheelchairs with the same travel miles in indoor or institutional settings with smooth surfaces. Rough terrains are commonly witnessed in rural areas or less-resourced settings where the frequency of wheelchair failures is higher, requiring frequent servicing interventions.(Mhatre et al., 2017)

All participants performed routine maintenance, unlike a previous study with ultralight wheelchair users (Fitzgerald et al., 2005) in which only one-fourth of participants did. Undergoing maintenance training was associated with greater travel distance, frequent maintenance activity, and more brake failures. Frequent maintenance activity was also associated with more part failures. Since the authors did not study how maintenance was performed by users, it is difficult to know why maintenance did not reduce or prevent failures. At the same time, maintenance training has not had an effect on failure reduction in previous studies.(D’Innocenzo et al., 2021; Garcia-Mendez et al., 2024) Based on findings, the more active wheelchair users are, they prefer training in maintenance and witness a high failure rate proportional to usage or shocks. These findings highlight that user-led maintenance alone cannot prevent failures despite self-reported maintenance compliance, encouraging a new approach to servicing. Employing repair technicians to service wheelchairs through a data-driven, multi-pronged approach practiced in other industries can prove beneficial. For instance, passenger cars require user-led maintenance activities like washing the car, checking hazard lights, and ensuring appropriate tire pressure. Authorized repair technicians typically perform oil changes when the vehicle has covered a certain thousand kilometers, part inspection and repair when a warning light goes off on the dashboard and annual inspections in some places. Such a multi-pronged approach where stakeholders share the responsibility of wheelchair servicing can alleviate the servicing burden on users who lack the ability, support, and resources for maintenance.

While all study participants confirmed they performed maintenance in Fig 3, some arrived at study visits with failures like no air in tires and shimmying casters. This indicates that either maintenance compliance is an issue, users need additional training or training needs further validation. For instance, matching maintenance activities to the user’s ability, motivation, health conditions, and caregiving support may improve compliance. Maintenance tasks can be specific based on product profiles and failure evidence, and such guidance can be emphasized in wheelchair user manuals through revised requirements in ISO 7176–15: Requirements for information disclosure, documentation, and labeling (International Organization for Standardization, 2014).

According to the literature, one-third of users experience adverse consequences (M. Toro et al., 2016; L. Worobey et al., 2012, 2014; L. A. Worobey et al., 2021). In this study, a similar consequence rate was observed with a broken axle lock and fractured casters, resulting in one participant missing work/school and another stranded. With an average of over two high-risk and five low-risk failures recorded over 20 months, the failure numbers reported in Table 2 are similar to the failure evidence reported in previous community studies (Henderson et al., 2020; McClure et al., 2009; M. Toro et al., 2016; L. Worobey et al., 2012, 2014) but higher than that reported with ultralight wheelchair users (Fitzgerald et al., 2005). Brakes and tires encountered a higher proportion of failures, a finding synonymous with a study (A. M. James et al., 2022) reporting failure proportions among manual wheelchair parts.

Certain unexpected outcomes were observed. First, participants indicated higher satisfaction with wheelchairs despite a high failure rate. This antithetical outcome may be due to insurance coverage of wheelchairs and participants’ fear of devices being taken away. Second, as seen in Table 2, a smaller number of failures was witnessed in the second survey, perhaps due to reduced travel during the pandemic’s Omicron variant, which emerged in November 2021. Third, when self-reported travel distance was compared to that measured by the sensor, five participants were found to underestimate their miles traveled. This finding further supports using technology to objectively monitor usage in the community and track miles and physical activity.

To promote the implementation of this research in the future and reduce wheelchair failures, the authors advocate for changes in the current servicing practice which are possible with policy change for servicing reimbursement. The authors anticipate building evidence on the cost-effectiveness of wheelchair servicing and improvements in wheelchair reliability, quality of life adjusted years, and social return on investment due to active servicing. Such evidence should strengthen the advocacy efforts and inform policymakers regarding including servicing and related reimbursement in service provision. Collaborative development efforts in this area with industry stakeholders are imperative toward policy change that incentivizes service providers for active wheelchair servicing, thereby safeguarding users from frequent part failures and severe consequences.

Limitations

Completing study procedures like data collection and recruitment was a significant challenge as the study period coincided with the peak pandemic and wheelchair users were a high-risk population. Lack of using cognitive screening tests during screening against exclusion criteria may have resulted in selection bias. The study findings are biased since there is no representation of users who did not perform maintenance and did not encounter failures. Also, the sample is homogenous, and the participant pool is small, which is suitable for the pilot stage of research but limits the reliability and generalizability of the findings. Recall bias with self-reporting and presence of the Hawthorne effect are potential limitations because the participants were seen in a academic setting affiliated with a medical center.

Future Work

Future work involves developing models and interventions as part of a multi-pronged servicing approach to demonstrate the efficacy of preventing part failures. Two studies are ongoing to further strengthen the evidence on associations between wheelchair usage, failures and consequences with other wheelchair models and evaluate the model-based capability of failure risk classification and prediction. Maintenance performance and compliance and its association with wheelchair degradation needs to be studied through prospective studies.

Conclusions

Research evidence demonstrates that wheelchair servicing by repair technicians is effective in reducing wheelchair part failures and consequences. This pilot study aims to develop a servicing tool for technicians and demonstrates a correlation between failure measures and risk factors of mechanical wheelchair usage and weight distribution, encouraging the development of servicing prediction models based on sensor-monitored usage. Ultralight wheelchair users supported the development of a user-facing smart technology to inform them about wheelchair servicing events. This study’s findings encourage the development of a usage monitoring technology to support technician-led servicing and prevent wheelchair failures and user consequences.