Electric scooter–related orthopaedic injuries in Wellington

Electric scooters (e-scooters) have emerged as a novel and accessible mode of transport. Worldwide, patterns of e-scooter–associated injuries are emerging, where minor contusions, abrasions and lacerations are the most common. This is followed by fractures or other orthopaedic injuries, particularly among young male riders.^1–4 Drug and alcohol use, in conjunction with lack of helmet use, has also been shown to increase injury severity, hospital length of stay and need for surgical intervention.⁵

In New Zealand, e-scooters are classified as low-powered vehicles.⁶ E-scooter riders do not require a driver’s licence and helmets are not legally required.⁶ Commercially operated e-scooters can travel up to 25 kilometres per hour.⁷ Since their introduction in New Zealand, Brownson et al. investigated the epidemiology of these injuries in the Auckland Region, with 41.7% of their cohort sustaining fractures.¹ Of those requiring orthopaedic operative management, the total economic cost was estimated to be NZ$19,282 per person.⁸

On 18 June 2019, 800 commercial e-scooters were introduced to Wellington City as part of an 18-month trial.⁹ Not only known for its wind, Wellington City also has a unique geography. Being surrounded by hills, this densely populated city is filled with narrow and tortuous roads. By design of its small wheels, e-scooters may be susceptible to instability in such an environment.^8,10 The primary aim of this study is to compare e-scooter–related orthopaedic injuries presenting to Wellington Regional Hospital with known data from other regions of New Zealand. Secondary aims are to compare presentations in the 2 years preceding and following the introduction of commercial e-scooters, and to assess resources required for investigation and management.

Method

A retrospective chart review of all e-scooter–related injuries presenting to the Wellington Emergency Department over a 48-month period (June 2017 to June 2021) was performed. This period included 24 months prior to the introduction of commercial e-scooters (June 2017 to June 2019) and the 24 months after its introduction. Ethics approval was granted by the New Zealand Health and Disability Ethics Committees (Reference 2024 EXP 20944).

The Wellington Hospital Accident Compensation Corporation (ACC) Department provided a list of presentations to the Wellington Emergency Department with claims related to “scooter”. The list was then screened manually to confirm relevance to e-scooters. E-scooters were defined as scooters with a foot board, steering handle, two to three wheels and an electric motor. Mopeds or non-motorised scooters were excluded.

Using the screened list, we generated a list of National Health Indexes (NHIs) appropriate for review. These NHIs were then searched on the local electronic patient record system, known as the Medical App Portal (MAP). MAP allows access to patient data across the Wellington Region hospitals, which includes the Wellington Regional, Hutt and Wairarapa hospitals. Information was collected on patient demographics, including age, gender and ethnicity. Data on time and method of presentation, as well as circumstances of injury, considering mechanism, presence of personal protective equipment and intoxication were also collected. Alcohol involvement is a compulsory question prior to discharge from the electronic system of the Wellington Emergency Department. In instances where details were not documented, due to acuity or other reasons, these were noted as “no record”.

Injuries were classified as non-orthopaedic fractures, orthopaedic fractures and orthopaedic soft tissue injuries. Fractures involving the skull, facial bones, ribs and teeth were not considered orthopaedic fractures for the purpose of this review. To note, a fracture diagnosis could only be made following physical examination and imaging. Relevant imaging of all modalities was retrospectively reviewed by a single investigator. Data on acute admission, type of operative management, clinic appointments, readmission and complications were also collected.

Statistical analysis was performed using frequency and percentage for categorical variables and mean with standard deviation for continuous variables.

Results

Between 17 June 2017 and 17 June 2019, 14 e-scooter–related presentations were identified. There was a total of two orthopaedic fractures in a single patient. These were managed non-operatively and the patient was discharged from fracture clinic with no complications.

Between 18 June 2019 and 18 June 2021, there were 295 e-scooter related presentations. The monthly number of presentations is presented in Figure 1. As shown in Table 1, 177 of these patients identified as male. The most common age group was 20–29 years of age. The first quarter of the day had a significantly lower number of presentations compared with the remaining quarters (Chi-squared test, p=0.006). Alcohol was involved in 30.5% of male presentations and 15.3% of female presentations. Falling from the e-scooter was the most common mechanism of injury. No deaths were observed at 90 days post-presentation.

View Figure 1, Table 1–2.

Of the 295 presentations since 18 June 2019, 54 patients sustained minor contusions, abrasions and lacerations. Forty-seven patients were diagnosed with a head injury and 28 patients sustained dental trauma. There were 11 patients with fractures of the skull or facial bones, and seven patients with rib fractures.

Regarding orthopaedic injuries, 13 patients sustained soft tissue injuries, such as patella tendon rupture. Two of these patients required surgery. A total of 145 orthopaedic fractures were diagnosed in 117 patients. Nine patients had pure joint dislocations without fracture.

Of the 107 upper limb fractures, fractures of the proximal radius/ulna were the most common (30.8%), followed by distal radius/ulna fractures (21.5%). Fifty-four point eight percent of radial head/neck fractures and 58.8% of distal radius fractures were intra-articular. Eighty-two point two percent of upper limb fractures were managed non-operatively, as demonstrated in Table 2. One patient had planned removal of metalware, and three patients had surgical complications. One readmitted for pain, one readmitted for wound infection and one referred to private orthopaedic surgeons with avascular necrosis. Further details on patients requiring readmission are included in the Appendix.

Of the 36 lower limb fractures, malleolar fractures were the most common (27.8%), followed by fractures of the tibia (25%). Fifty-eight point three percent of lower limb fractures required operative management. All femur and tibia fractures resulted in operative management, with two tibia fractures requiring an external fixator prior to open reduction internal fixation. Two patients had planned removal of metalware in the public hospital and two patients were referred to private orthopaedic surgeons for the same. One patient developed septic arthritis following open reduction internal fixation of their tibial plateau fracture, resulting in two reoperations followed by a stitch abscess.

Assessing the effect of alcohol on the severity of long bone fractures, there were 14 long bone fractures sustained while intoxicated and 81 fractures sustained with no/unknown alcohol intake. Of the fractures sustained while intoxicated, 50% of these were intra-articular and 28.5% were open fractures. In the no/unknown alcohol group, 55.6% of fractures were intra-articular and 2.5% were open. The difference in open fractures between the two groups were statistically significant (Fisher’s exact test, p=0.003).

Of the 61 advanced imagings performed acutely within the Wellington Region hospitals, 38 were performed to evaluate orthopaedic injuries. Of these, 37 were computerised tomography (CT) scans and one was an magnetic resonance imaging (MRI) scan to evaluate for soft tissue knee injury. In the follow-up period, five scans were performed to assess for complications, e.g., ultrasound for deep vein thrombosis to CT to assess for union.

There were 40 admissions to Wellington Regional Hospital, with one to neurosurgery, one to general surgery and two to urology. The remaining 36 presentations resulted in acute (72.2%) or planned-acute (27.8%) admissions under the orthopaedic service. For this group, range of stay was 0–27 days, with a median of 2.5 days. Eight patients required readmission. The total acute surgical time at Wellington Hospital was 3,393 minutes from surgical incision to closure, with a further 126 surgical minutes for planned removal of metalware or further surgeries related to complications. Ninety-three patients were seen at the Wellington Fracture Clinic with a total of 253 clinic appointments.

Discussion

E-scooters are an accessible and convenient mode of transport. However, as illustrated by our study, orthopaedic injuries are common. In our cohort, 39.7% were diagnosed with orthopaedic fractures. While this is similar to the 37.2% described in the 2019 Auckland study by Brownson et al., 29.0% of orthopaedic fractures in our study required operative management, compared to the 15.0% found in the Auckland study.¹ Our study, as well as identifying 4.1% of total fractures as open fractures, also assessed further markers of high-energy trauma, such as fracture–dislocations and intra-articular involvement.

While blood alcohol level is not routinely measured, the Wellington Hospital Emergency Department electronic documentation system requires the discharging clinician to comment on alcohol involvement for every patient, whether trauma-related or not. In our cohort, alcohol was involved in approximately one quarter of presentations. This is again comparable to Brownson’s study in Auckland.¹ Across Australasia, intoxication rates as high as 46% have also been observed.¹¹ With our study showing a higher rate of open long bone fractures in intoxicated riders, this calls for further regulations or education to prevent e-scooter use while intoxicated.

Our study observed radial head/neck fractures to be the most common upper limb fracture. A fall onto an outstretched hand can produce such an injury, and this is in keeping with our cohort’s young mean age and most common mechanism of injury of rider fall. Only 8.1% of our cohort was wearing personal protective equipment at the time of injury. The low numbers are similar to that observed in other trauma centres, and are unsurprising given helmet use is not legally enforced in New Zealand.¹²

Our study found a stark difference in the number of e-scooter–related injuries since the introduction of commercially operated e-scooters in Wellington City. For fractures alone, there were two prior to their introduction, compared with the 145 noted in the 2 years following their introduction. This difference is likely dulled by the social effects of COVID-19 between 2020 and 2021, as supported by the low number of presentations around the lockdown periods. In our study, riders presenting in the post-introduction period were not classified into private and commercial e-scooter groups. While this is a limitation due the study’s retrospective nature, it is known that private and commercial e-scooter riders have different user profiles, with an Australian study demonstrating a higher prevalence of illegal riding in commercial e-scooter users.¹³ The large increase in the number of injuries seen is therefore likely from users of commercial e-scooters.

In the 2018 Census, 74.1% of the Wellington City population identified as European, 8.6% as Māori, 5.1% as Pacific peoples and 18.3% as Asian.¹⁴ Our study reflects a similar proportion of European and Māori patients (74.2% and 7.8% respectively), but found a lower percentage of patients identifying as Pacific peoples or Asian. While this may represent a true difference in the number of injuries, this may also indicate differences in injury severity or preference for accessing primary care. A 2020 study in Auckland found 68.1% of patients with e-scooter injuries presented to primary care.¹⁵ This indicates that the total number of e-scooter injuries and associated socio-economic cost are much higher than shown by our study.

Our study has redemonstrated the resource-intensive nature of e-scooter injury management. In our cohort, 36 patients required hospital admission and 42 fractures were managed operatively. With the average study participant being 33 years of age, such injuries often result in time off work or reduced duties, both of which have financial and productivity implications as demonstrated by Campbell et al.⁸ While our study focussed on orthopaedic injuries alone, it is important to acknowledge that head and facial injuries, as well as minor contusions, abrasions and lacerations are also common in e-scooter–related presentations, and can be a source of morbidity.^2,16–18

This study has a number of limitations. It is a retrospective review, whereby accuracy of our data is dependent upon the quality of prior documentation. In addition, data collection was carried out by a single unblinded investigator. While this promotes consistent data interpretation, it is also a source for potential bias. Furthermore, ACC data on mechanism of injury are reliant on details recorded by the patient and clinician. This, together with the high number of e-scooter injuries presenting to primary care, suggests the true burden of these injuries is much higher than that found in our study.

E-scooters are an increasingly popular transport option. Their unique user profile and accessibility create a new mechanism and pattern of orthopaedic injuries. The mountainous landscape of Wellington City poses new challenges in maintaining rider safety, where orthopaedic fractures in Wellington are almost twice as likely to require operative management as those in Auckland. Orthopaedic soft tissue injuries further add to the burden. Assessment and management of these injuries are also resource intensive. Future studies can further assess the economic cost of e-scooter injuries in Wellington, as well as the relationship between e-scooter models and risk of fall. The establishment of regulations and education around use of personal protective equipment, use while intoxicated and speed limits would also be beneficial.

View Appendix.