The impact of electric scooters in Melbourne: data from a major trauma service
Abstract
Background
The proliferation of electric scooters globally has been associated with an increase in related injuries and consequent economic burden. This study aims to assess the injury patterns and the economic impact associated with electric scooter use in Melbourne, Australia.
Methods
A retrospective cohort study was conducted using hospital and registry data from January 2022 to January 2023. Data collected included demographic details, alcohol and helmet use, injury type and severity, operative treatment provided, and direct medical costs. The economic impact (in AUD) of the patient's emergency presentation and hospital admission was calculated.
Results
During the study period, 256 electric scooter related injuries were recorded, comprising 247 riders and nine pedestrians. The majority of patients were males (69%) with a median age of 29.5 (15–78). Alcohol use was reported by 34% and helmet use by 33%. Injuries most commonly affected the upper limb (53%) and head (50%), with abrasions (75%) and fractures (48%) being the most common type of injury sustained. The total hospital cost was $1 911 062, and the median cost was $1321.66 per patient (IQR: $479.37–$5096.65).
Conclusion
Electric scooter usage, as observed through patient presentations to the Royal Melbourne Hospital, is associated with a considerable number of injuries, primarily among young males, and an ensuing substantial economic burden. The findings underscore the urgent need for improved safety measures to minimize electric scooter-related injuries and their clinical and economic repercussions.
Introduction
Electric scooters (e-scooters) have become an increasingly popular mode of transport worldwide, offering an efficient, convenient, and affordable transport option for many individuals. With the recent introduction of e-scooter rental companies in major cities globally, the use of these scooters has dramatically increased.1-3 Expectedly, as more individuals utilize these devices, the number of injuries sustained from their use has also increased, raising concerns regarding their impact on public healthcare systems.2, 4, 5
It has been shown that presentations to the emergency department due to e-scooter injuries may increase by up to six times following the introduction of e-scooter rental services.3 The rate of injuries has been estimated to be 60 per 100 000 rides in one study.6 Common injuries include superficial wounds and fractures, however, serious head and facial trauma resulting in brain injury or even death have been described.7-9 The use of helmets is low, less than 5% in some studies, and the prevalence of rider alcohol usage is high.10
Despite the potential environmental benefits, e-scooter usage presents a growing public health concern with an increasing financial burden to local healthcare systems. Studies have attempted to quantify this cost, however, these estimates vary between countries and studies have been heterogenous in their method of cost determination.2, 6, 11-20
In Australia, rental e-scooters have recently become widely available in six out of seven states. In Victoria, the Government launched a trial of rental e-scooters on the 1st of February 2022.21 One of Melbourne's major trauma services, the Royal Melbourne Hospital (RMH), sits at the centre of this trial location and has seen a concerning influx of e-scooter related injuries. To date, there has been little research evaluating the injuries and cost associated with e-scooter use in Melbourne. This study aims to determine the financial cost of treating e-scooter injuries in Melbourne and characterize the number and type of injuries sustained. Ultimately, we hope to provide information to facilitate evidence-based decision-making and ensure the responsible integration of e-scooters into the urban transportation landscape.
Methods
This study was approved by the Melbourne Health Ethical Review Board (Reference Number: QA2023037). We conducted a retrospective cohort study of all patients who presented to the RMH Emergency Department due to an e-scooter crash between 1st January 2022 and 1st January 2023. During the study period on the 1st of February 2022, two ride sharing companies launched a total of 1500 scooters in the local government areas of Melbourne, Port Phillip and Yarra, cumulatively encompassing a population of approximately 356 000 people.21, 22 The scooters feature speed limitations, geofencing, and designated pathways for riding. Furthermore, riders are required to wear a helmet at all times while riding and not exceed a maximum speed of 20 km/h. In addition, private personally-owned e-scooters, which lack speed limitations, are also commonly utilized, despite being illegal to operate on public land.
Data collection
Case identification was via the Royal Melbourne Hospital Trauma Registry (RMH-TR). Currently there is no ‘code’ in any injury dataset for e-scooter incidents. Therefore, the RMH-TR used the Victorian Emergency Minimum Dataset codes to enable identification of these cases. For Injury Cause Classification Code 8: Other Transport related circumstance was chosen, and the word ‘electric scooter’ used in the free text event details. The RMH-TR expanded its usual inclusion criteria to include all incidents, regardless of length of stay, completing a full trauma registry dataset on all e-scooter cases, to give as complete a hospital burden as possible. Furthermore, patient records were consulted for any additional data. Costing data for admission episodes are routinely generated by the RMH Health Intelligence Unit and reported to government. These costings were retrieved from the hospital's enterprise data warehouse.
Data variables
Collected variables included demographic details, time/date of injury and presentation to ED, Australasian Triage Scale (ATS) helmet and alcohol usage.23 Alcohol usage was considered as a binary variable and considered positive if there was blood alcohol level >0.00% or documentation of alcohol use. Alcohol intoxication was considered when blood alcohol level was >0.05% (>11 mmol/L). Similarly, recreational drug use was considered positive where biochemical evidence was available or based on documentation. Injury characteristics included type of injury, anatomical location, severity (based on Abbreviated Injury Scale (AIS)24 and Injury Severity Score (ISS)), and mechanism of injury (e.g., collision with a vehicle/object and fall from the e-scooter). The Victoria Trauma System Definition was used to define major trauma.25 Mild traumatic brain injury (TBI) was defined as a head-strike (or acceleration/deceleration injury) with initial Glasgow Coma Scale of 13–15 along with loss of consciousness <30 min, confusion/disorientation, amnesia, or post-concussive symptoms. Details of management collected included imaging performed and surgical interventions. Admission and discharge details were also collected. Disposition from ED was categorized as ED discharge home or admission into hospital.
Costing analysis
The cost (in AUD) was determined for all treatment pertaining to the patient's initial emergency department presentation and their subsequent admission if applicable. This included the cost of staff, radiology, laboratory testing, the cost of procedures or operations performed, and overheads. Cost analysis and determination was conducted by the RMH Health Intelligence Unit. This costing data is routinely calculated and provided to the Victorian Department of Health and Human Services as part of an activity-based funding method.26 For the 2021/2022 financial year, existing costing data were extracted and utilized. For the 2022/2023 financial year, costs were modelled by applying cost weightings from the 2021/2022 financial year to actual service/utilization rates in the 2022/2023 financial year.
Results
A total of 256 unique e-scooter related injuries were identified, including, 247 riders and nine pedestrians. Overall median age was 29.5 years (range:15–78). Males constituted 68.8% of patients, 31.3% were female.
At presentation, most patients had an ATS of 3 (43.0%) or 4 (37.5%). Helmets were reportedly utilized by 32.8%, not used by 26.3% and incompletely documented in 40.9% of cases. Alcohol use was found in 33.6% of patients, while 13.0% patients reported no alcohol use, and no documentation was available for 53.4%. Of the 16 patients with recorded blood alcohol levels >0.00%, 15 (93.8%) were intoxicated, with levels >0.05%, and 26.7% of these intoxicated patients were classified as major trauma cases. Drugs (other than alcohol) were used by 11.3% of riders. Just over half (53.1%) of the cohort were discharged home from the ED. Of the 120 admitted to hospital, the majority were discharged home (93.3%), while four (3.3%) patients were sent to a rehabilitation centre and three (2.5%) were transferred to another hospital. Lastly, there was one death due to complications of a traumatic brain injury.
Further demographic characteristics are presented in Table 1.
| Discharged from ED (n = 136) | Admitted to Hospital (n = 120) | Total (n = 256) | |
|---|---|---|---|
| Age (median (range)) (years) | 26 (16–62) | 31 (15–78) | 29.5 (15–78) |
| Sex (n (%)) | |||
| Male | 86 (63.2%) | 90 (75%) | 176 (68.8%) |
| Female | 50 (36.8%) | 30 (25%) | 80 (31.3%) |
| Patient type | |||
| Rider | 131 (96.3%) | 116 (96.7%) | 247 (96.5%) |
| Pedestrian | 5 (3.7%) | 4 (3.3%) | 9 (3.5%) |
| ATS (n (%)) | |||
| 1 | 0 (0.0%) | 8 (6.7%) | 8 (3.1%) |
| 2 | 1 (0.7%) | 33 (27.5%) | 34 (13.3%) |
| 3 | 57 (41.9%) | 53 (44.2%) | 110 (43.0%) |
| 4 | 71 (52.2%) | 25 (20.8%) | 96 (37.5%) |
| 5 | 7 (5.1%) | 1 (0.8%) | 8 (3.1%) |
| Helmet use (n (%))† | |||
| Yes | 39 (29.8%) | 42 (36.2%) | 81 (32.8%) |
| No | 23 (17.6%) | 42 (36.2%) | 65 (26.3%) |
| Undocumented | 69 (52.7%) | 32 (27.6%) | 101 (40.9%) |
| Alcohol use (n (%))† | |||
| Yes | 36 (27.5%) | 47 (40.5%) | 83 (33.6%) |
| No | 7 (5.3%) | 25 (21.6%) | 32 (13.0%) |
| Undocumented | 88 (67.2%) | 44 (37.9%) | 132 (53.4%) |
| Recreational drug use (n (%))† | |||
| Yes | 5 (3.8%) | 23 (19.8%) | 28 (11.3%) |
| No | 7 (5.3%) | 29 (26.7%) | 38 (15.4%) |
| Undocumented | 119 (90.8%) | 62 (53.4%) | 181 (73.3%) |
| Imaging performed (n (%))‡ | |||
| None | 42 (30.9%) | 4 (3.3%) | 46 (18.0%) |
| X-ray | 84 (61.8%) | 95 (79.2%) | 179 (69.9%) |
| US | 2 (1.5%) | 23 (19.2%) | 25 (9.8%) |
| CT | 24 (17.6%) | 98 (81.7%) | 122 (47.7%) |
| MRI | 0 (0.0%) | 6 (5.0%) | 6 (2.3%) |
| NM Scan | 0 (0.0%) | 1 (0.8%) | 1 (0.4%) |
- Abbreviations: ATS, Australasian Triage Scale; ED, Emergency department.
- † Calculation of totals/percentages for these variables excluded the nine pedestrian patients.
- ‡ Imaging conducted throughout entire emergency presentation ± inpatient admission.
Injury characteristics
Table 2 displays the data on the injuries sustained. 8.2% of cases were classified as major trauma. Overall, most injuries resulted from individuals falling off e-scooters (78.9%), yet pedestrians were mostly injured due to riders of e-scooters colliding with them (88.9%). The upper limb was the most frequently injured site (53.1%). This was closely followed by injuries to the head, face, or neck (50.0%) and lower limb (37.1%). Abrasions, lacerations, and contusions comprised the greatest number of injuries observed (75.0%). The most common location for these injuries was the head, face, or neck (46.1%). Fractures were next most common (47.7%). The most commonly fractured body region was the upper limb (26.2%), followed by facial fractures (14.5%). 16.4% of patients sustained head injuries of some kind; most were mild traumatic brain injuries (10.9%), whereas 5.5% of patients sustained intracranial bleeds.
| Discharged from ED (n = 136) | Admitted to Hospital (n = 120) | Total (n = 256) | |
|---|---|---|---|
| Severity (n (%)) | |||
| Minor trauma | 134 (98.5%) | 101 (84.2%) | 235 (91.8%) |
| Major trauma | 2 (1.5%) | 19 (15.8%) | 21 (8.2%) |
| Mechanism of injury | |||
| Fall | 114 (83.8%) | 88 (73.3%) | 202 (78.9%) |
| Collision | 22 (16.2%) | 32 (26.7%) | 54 (21.1%) |
| Region injured (n (%))† | |||
| Head, face, or neck | 59 (43.4%) | 69 (57.5%) | 128 (50.0%) |
| Upper limb | 72 (52.9%) | 64 (53.3%) | 136 (53.1%) |
| Lower limb | 40 (29.4%) | 55 (45.8%) | 95 (37.1%) |
| Thorax | 9 (6.6%) | 25 (20.8%) | 34 (13.3%) |
| Abdomen | 3 (2.2%) | 7 (5.8%) | 10 (3.9%) |
| Spine | 1 (0.7%) | 7 (5.8%) | 8 (3.1%) |
| Type of injury (n (%))† | |||
| Abrasion/laceration/contusion | 99 (72.8%) | 93 (77.5%) | 192 (75.0%) |
| Head/face/neck | 53 (39.0%) | 66 (55.0%) | 118 (46.1%) |
| Trunk | 6 (4.4%) | 17 (14.2%) | 23 (9.0%) |
| Upper limb | 42 (30.9%) | 41 (34.2%) | 83 (32.4%) |
| Lower limb | 35 (25.7%) | 42 (35.0%) | 77 (30.1%) |
| Fracture | 42 (30.9%) | 80 (66.7%) | 122 (47.7%) |
| Skull | 3 (2.2%) | 12 (10.0%) | 15 (5.9%) |
| Face | 6 (4.4%) | 31 (25.8%) | 37 (14.5%) |
| Upper limb | 33 (24.3%) | 34 (28.3%) | 67 (26.2%) |
| Lower limb | 1 (0.7%) | 19 (15.8%) | 20 (7.8%) |
| Thorax | 2 (1.5%) | 15 (12.5%) | 17 (6.6%) |
| Spine | 1 (0.7%) | 7 (5.8%) | 8 (3.1%) |
| Dental injury | 9 (6.6%) | 9 (7.5%) | 18 (7.0%) |
| Head injury | 9 (6.6%) | 33 (27.5%) | 42 (16.4%) |
| Mild TBI | 9 (6.6%) | 19 (15.8%) | 28 (10.9%) |
| Intracranial bleed | 0 (0.0%) | 14 (11.7%) | 14 (5.5%) |
| Vascular injury | 0 (0.0%) | 2 (1.7%) | 2 (0.8%) |
| Thoracic organ injury | 0 (0.0%) | 3 (2.5%) | 3 (1.2%) |
| Abdominal injury | 2 (1.5%) | 6 (5.0%) | 8 (3.1%) |
| Abdominal wall | 1 (0.7%) | 3 (2.5%) | 4 (1.6%) |
| Abdominal organ | 1 (0.7%) | 5 (2.5%) | 6 (2.3%) |
| Sprains/strains/tears | 23 (16.9%) | 11 (9.2%) | 34 (13.3%) |
| Upper limb | 14 (10.3%) | 4 (3.3%) | 18 (7.0%) |
| Lower limb | 9 (6.6%) | 3 (2.5%) | 12 (4.7%) |
| Spine | 0 (0.0%) | 4 (3.3%) | 4 (1.6%) |
| Joint dislocation | 4 (2.9%) | 5 (4.2%) | 9 (3.5%) |
| Shoulder | 2 (1.5%) | 2 (1.7%) | 4 (1.6%) |
| Midfoot | 1 (0.7%) | 1 (0.8%) | 2 (0.8%) |
| Carpal joint | 0 (0.0%) | 1 (0.8%) | 1 (0.4%) |
| Ankle | 0 (0.0%) | 1 (0.8%) | 1 (0.4%) |
| Elbow | 1 (0.7%) | 0 (0.0%) | 1 (0.4%) |
- † A single patient may have sustained multiple injuries, to multiple sites.
Injuries increased in number after the introduction of the rental e-scooters in February 2022, peaking in December 2022 with 40 injuries (Fig. 1). Most injuries occurred on Friday, Saturday, or Sunday numbering 46, 44 and 49 injuries respectively, with a greater proportion of riders under the influence of alcohol on these days as well.
Operative characteristics
In total, 40.0% of patients received operative treatment (Table 3). The most common operations performed were wound washout/debridement/closure (14.2%) and upper limb fracture Open Reduction and Internal Fixations (ORIF) (11.7%). Orthopaedics performed the majority (55.6%) of procedures, followed by plastic surgery (20.4%) and oral & maxillofacial surgery (13.0%).
| Number (%) (n = 120) | |
|---|---|
| Operative treatment (n (%)) | 48 (40.0%) |
| Operation performed (n (%))† | |
| Upper limb fracture ORIF | 14 (11.7%) |
| Wound washout/debridement/closure | 17 (14.2%) |
| Lower limb fracture ORIF | 13 (10.8%) |
| Facial fracture ORIF | 8 (6.7%) |
| EVD/ICP insertion | 3 (2.5%) |
| External fixation | 3 (2.5%) |
| Craniectomy/craniotomy | 2 (1.7%) |
| Tendon repair | 1 (0.8%) |
| Arterial repair | 1 (0.8%) |
| Cranial Burr hole | 1 (0.8%) |
| Evacuation of intracranial haematoma | 1 (0.8%) |
| Laparoscopy/laparotomy | 1 (0.8%) |
| Stomach repair | 1 (0.8%) |
| Carpal tunnel release | 1 (0.8%) |
| Elbow ligament repair | 1 (0.8%) |
| VATS washout of haemothorax | 1 (0.8%) |
| Rib ORIF | 1 (0.8%) |
| Spine ACDF | 1 (0.8%) |
| Operating unit (n (%))† | |
| Orthopaedic Surgery | 30 (55.6%) |
| Plastic Surgery | 11 (20.4%) |
| Oral and Maxillofacial Surgery | 7 (13.0%) |
| Neurosurgery | 4 (7.4%) |
| Vascular Surgery | 1 (1.9%) |
| Thoracic Surgery | 1 (1.9%) |
- Abbreviations: ACDF, anterior cervical discectomy and fusion; EVD, external ventricular drain; ICP, intracranial pressure monitor; ORIF, open reduction internal fixation; VATS, video-assisted thoracic surgery.
- † A single patient may have had multiple operations by multiple units.
Costing analysis
The cost of treating these e-scooter injuries, when only considering the cost of the emergency presentation and initial inpatient admission, totalled to $1 911 061.59 (Table 4). This was comprised of $370 056.18 associated with the emergency presentation costs and $1 541 005.41 associated with inpatient admission costs. Costs were predominantly accrued due to staff costs, theatre costs and the cost of ICU admissions. The range of costs per patient was wide, ranging from a minimum cost of $79.97 to a maximum cost of $259 876.83. The median cost per patient was $1321.66 with an interquartile range of $479.37–$5096.65.
| Category | Cost (AUD) |
|---|---|
| Emergency presentation | |
| Staff costs | $ 193 224.91 |
| Pathology costs | $ 27 356.76 |
| Pharmacy costs | $ 11 182.99 |
| Radiology costs | $ 128 079.80 |
| Other costs† | $ 10 211.72 |
| Total emergency cost | $ 370 056.18 |
| Inpatient admission | |
| Staff costs | $ 605 267.87 |
| Pathology costs | $ 22 932.76 |
| Pharmacy costs | $ 22 550.20 |
| Radiology costs | $ 48 042.38 |
| Theatre costs‡ | $ 460 706.47 |
| ICU costs | $ 353 057.93 |
| Other costs | $ 28 447.79 |
| Total inpatient admission cost | $ 1 541 005.41 |
| Total cost | $1 911 061.59 |
- † Other costs include patient meals and consumables.
- ‡ Theatre costs include theatre consumables, theatre staff costs, central sterile supply department (CSSD) costs and prostheses costs.
Discussion
E-scooters have become increasingly popular globally, with described benefits such as lower cost and less carbon emissions. However, with this rise in popularity has come a dramatic increase in e-scooter related injuries. In the decade prior to 2020, the RMH Trauma Registry recorded less than five e-scooter admissions per annum. The present study reveals that, over the course of a single year, this number has substantially increased and that these injuries incurred a substantial economic burden. Ultimately, this study provides important information that should be considered by policymakers when making decisions about the continued presence of e-scooters as a mode of transport in Victoria.
The direct economic impact of e-scooter injuries is substantial. Emergency care and hospitalization costs are the most direct financial impacts, as victims may require immediate medical attention, hospital admission or operations for management. This study reveals that the State Government of Victoria is absorbing substantial cost, causing an increased burden on the healthcare system. In addition, costs associated with initial hospitalization are known to represent only a portion of the total economic impact of trauma patients.27 Currently, the Victorian government receives no funding from commercial e-scooter companies to base their scooters in Melbourne. Furthermore, no appropriate insurance policy exists, and patients are liable to incur the costs of their injuries. Considering the economic impact of these injuries, to offset this cost, a funding mechanism similar to the revenue generated by motor vehicles through the Transport Accident Commission (TAC) could be implemented. In this model motorists contribute a component of their vehicle registration cost to fund the TAC. Similarly, e-scooter companies could contribute funds to the Victorian government as a condition for operating their services in the state.
The pattern of injuries emerging from Melbourne mirrors reported literature. Consistent with multiple other studies, young males constituted the majority of our patients, totalling 69%, and the median age 29.5 years old (range: 15–78).10, 13-16, 19, 20, 28-35 One systematic review by Singh et al. in 2022, analysed the pattern of e-scooter injuries among 34 studies.7 They found that of the 5702 injured patients, approximately 60% were male with a mean age of the total sample of 33.3 years.7 In many studies, risk-taking behaviour, associated with factors such as non-use of protective gear and alcohol use, have been seen to be more prevalent in males, a potential contribution to their higher representation among e-scooter injury cases.36-38
The use of alcohol in this study was concerningly prevalent, given the inherent need for a high level of balance and coordination when riding an e-scooter. In the present study, 34% of patients used alcohol at the time of their injury. Many studies worldwide have stressed this correlation, such that, in nearly all the conducted e-scooter injury evaluations, a high proportion of patients are alcohol affected at the time of the accident.7, 17, 20 The numbers in Melbourne appear to be on the higher end of this spectrum. Singh et al. in their review, demonstrated the rate of alcohol use to be approximately 25% (range: 3%–48%) among 24 included studies.7 Moreover, in the present series a large proportion (40.9%) of patients had incomplete documentation regarding alcohol use. Therefore, it is feasible that the rate of alcohol use is even higher. In addition, we found a higher rate of injuries on Fridays and the weekends, with higher rates of alcohol usage on these days. Given that individuals might choose not to drive cars when they plan to consume alcohol, the widespread availability and low barrier to entry for e-scooters (usually just a smartphone and a credit card) might tempt some to use them opportunistically when under the influence of alcohol. Recreational drugs were used less frequently (11.3%), yet 28.6% of the major trauma patients used recreational drugs prior to riding e-scooters, compared to just 9.1% among the minor trauma patients. It has long been known that the use of drugs and alcohol in transport related accidents is a risk factor for injury and has been shown to increase the risk of sustaining more serious injuries or death.39-41 Thus, the high prevalence of alcohol related accidents among e-scooter injuries raises serious concerns and suggests an urgent need for improved legislation and public education on the consequences of utilizing e-scooters whilst under the influence of alcohol.
Similarly, helmet usage, or rather the lack thereof, also stood out as a concerning issue. Despite the fact that helmets are mandatory in Victoria, only 33% of patients definitively reported wearing helmets, although this estimate is dependent upon the quality of the medical record. This mirrors other studies conducted in countries where helmets are compulsory, which report rates of helmet usage between 19% and 46%.17, 18, 42 Like alcohol use, the use of helmets in transport related accidents has long been studied and the protective benefits of wearing helmets are well known.43, 44 Helmets provide a dramatic reduction in head injuries and reduce the risk of serious brain injuries which can be fatal.45
This study demonstrated that 53% of patients sustained some form of upper limb injury and 50% of patients experienced head injuries. These patterns are again consistent with other studies in Australia and international data.7, 16-18, 20 Frequently, riders instinctively extend their arms to break a fall and the head may often strike against the ground or collide with an object. The high frequency of head injuries is of particular concern given the low utilization of helmets and their potential for acute and long-term morbidity. Superficial wounds such as abrasions, contusions or lacerations predominated, followed by fractures (particularly of the upper limb or face). In this context, most operations were performed by orthopaedics, plastic surgery and oral & maxillofacial surgery. This is consistent with other studies demonstrating a high reliance on orthopaedics, plastics or OMFS units for operative management.2, 7, 19, 20, 46-48
The findings presented herein are consistent with trends observed in other Australian and New Zealand cities which have an established history of electric scooter use.6, 17, 18, 35, 49 The Melbourne data indicates comparable rates of low helmet use, akin to those found in Western Australia and Tasmania, despite mandatory helmet regulations in these areas.35, 49 Incidences involving alcohol are notably prevalent across Australia and New Zealand, with rates ranging from 27% to 46%, and the most prevalent occurrences reported in the Northern Territory.6, 17, 18, 35, 49 Across all cities, upper limb and head injuries emerge as the predominant types of injury resulting from e-scooter accidents.6, 17, 18, 35, 49 In contrast, the data from Melbourne reveals a higher cost per patient compared to these other cities, although these cost comparisons are limited and potentially imprecise due to the varying costing methodologies employed.6, 17, 18, 35, 49 Collectively, the research from these regions paints a uniform picture of the considerable risk of injury associated with e-scooters, along with a consequential financial burden on local health services.
This study had several limitations. Firstly, it does not account for injuries that did not result in hospitalization or were treated at primary care clinics. Moreover, our modelling of costs for the 2022/2023 financial year may under-represent actual costs due to inflation. Furthermore, the cost of outpatient appointments, ambulance/paramedic costs, time off work, permanent illness or disability, and chronic pain were also not considered. Vasara et al. in 2022 performed a comprehensive costing analysis using their data from Helsinki, including many of these aforementioned costs. They found that of the total cost observed, the emergency presentation and inpatient stay constituted only 30% of the total costs, with a substantial expense attributed to time off work. Thus, the true economic burden of e-scooter injuries is likely much greater than what was determined in this study. However, conversely, this study did not consider the potential societal costs saved from e-scooter use, such as reduced traffic congestion, carbon pollution, and other benefits. Another limitation was the retrospective nature of this study which relied upon the accuracy and completeness of medical records. Furthermore, this study largely relied upon self-reported data for alcohol/drug use and helmet usage, which is likely to have introduced reporting bias. It was also not possible to differentiate between privately owned e-scooters and the rental e-scooters that are publicly available. Lastly, we were unable to measure the total number of e-scooter trips in Victoria during the timeframe of this study (as this data was unavailable) and therefore, were unable to provide an estimate of the risk of hospital presentation as a proportion of this total. While it has been previously reported by the Victorian Government that there were 3.7 million short trips taken as of 30th of March 2023,50 what was considered a ‘trip’ remains unclear. Furthermore, an estimate of the injury rate based on the statistics presented in this study alone would not capture presentations to other hospitals and would therefore be inaccurate. Strengths of this study include the thorough prospective case acquisition, plus the inclusion of all cases presenting to the RMH not just admissions nor major trauma, and the collection of a full dataset (usually reserved for major trauma) for all cases.
Conclusion
E-scooters have resulted in 256 injured persons presenting to the Royal Melbourne Hospital in 2022, the first year of the trial. Despite potential economic and climate benefits, these injuries have been costed at just under $2 Million AUD due to direct medical costs, but potential further costs such as productivity loss remain unmeasured. Alcohol & other drug usage is high and helmet usage reportedly low, suggesting opportunities for targeted prevention to increase safety. The information presented here provides crucial data to policymakers when considering the future presence of e-scooters in Melbourne and the safety of riders. We propose an additional levy imposed on e-scooter companies to recuperate some of these costs.
Author contributions
Jevan Cevik: Conceptualization; data curation; formal analysis; investigation; methodology; visualization; writing – original draft; writing – review and editing. David Read: Conceptualization; methodology; project administration; resources; software; supervision; validation; writing – review and editing. Mark Putland: Conceptualization; methodology; project administration; resources; software; supervision; validation; visualization; writing – review and editing. Timothy Fazio: Conceptualization; data curation; methodology; project administration; resources; software; supervision; validation; visualization; writing – review and editing. Kellie Gumm: Conceptualization; methodology; project administration; resources; software; supervision; validation; visualization; writing – review and editing. Amrita Varma: Data curation; methodology; project administration; resources; supervision; writing – review and editing. Roselyn Santos: Data curation; investigation; methodology; resources; software. Anand Ramakrishnan: Conceptualization; methodology; project administration; resources; software; supervision; validation; visualization; writing – review and editing.
Acknowledgements
We would like to acknowledge Elin Wee and Jeffrey Ong from the Clinical Costing Unit at the Royal Melbourne Hospital for their assistance in costing data extraction and modelling. Open access publishing facilitated by The University of Melbourne, as part of the Wiley - The University of Melbourne agreement via the Council of Australian University Librarians.
Conflict of interest
None declared.
