National Academies Press: OpenBook

E-Scooter Safety: Issues and Solutions (2022)

Chapter:Chapter 4 E-Scooter Injuries and Crash Context

« Previous: Chapter 3 E-Scooter Context and Safety Issues
Page 7
Suggested Citation:"Chapter 4 E-Scooter Injuries and Crash Context." National Academies of Sciences, Engineering, and Medicine. 2022. E-Scooter Safety: Issues and Solutions. Washington, DC: The National Academies Press. doi: 10.17226/26756.
Page 8
Suggested Citation:"Chapter 4 E-Scooter Injuries and Crash Context." National Academies of Sciences, Engineering, and Medicine. 2022. E-Scooter Safety: Issues and Solutions. Washington, DC: The National Academies Press. doi: 10.17226/26756.
Page 9
Suggested Citation:"Chapter 4 E-Scooter Injuries and Crash Context." National Academies of Sciences, Engineering, and Medicine. 2022. E-Scooter Safety: Issues and Solutions. Washington, DC: The National Academies Press. doi: 10.17226/26756.
Page 10
Suggested Citation:"Chapter 4 E-Scooter Injuries and Crash Context." National Academies of Sciences, Engineering, and Medicine. 2022. E-Scooter Safety: Issues and Solutions. Washington, DC: The National Academies Press. doi: 10.17226/26756.
Page 11
Suggested Citation:"Chapter 4 E-Scooter Injuries and Crash Context." National Academies of Sciences, Engineering, and Medicine. 2022. E-Scooter Safety: Issues and Solutions. Washington, DC: The National Academies Press. doi: 10.17226/26756.
Page 12
Suggested Citation:"Chapter 4 E-Scooter Injuries and Crash Context." National Academies of Sciences, Engineering, and Medicine. 2022. E-Scooter Safety: Issues and Solutions. Washington, DC: The National Academies Press. doi: 10.17226/26756.

Below is the uncorrected machine-read text of this chapter, intended to provide our own search engines and external engines with highly rich, chapter-representative searchable text of each book. Because it is UNCORRECTED material, please consider the following text as a useful but insufficient proxy for the authoritative book pages.

7 likely to perceive e-scooter use as safe as compared with men and significantly (p ≤ 0.001) more likely to indicate safety concerns such as worries about hitting someone or being hit and worries about feel- ing unsteady or falling, irrespective of past e-scooter use (Sanders et al. 2020). In Provo, men were significantly (p < 0.001) more likely to report having ridden on the street for their most recent e-scooter trip than women, who were significantly (p = 0.08) more likely to have used the sidewalk for riding (as opposed to both the sidewalk and street) (Glenn et al. 2020). College-aged respondents were signifi- cantly (p < 0.001) less likely to have ridden on the street than non- college-aged respondents. Seventy-four percent of respondents requested bike lane additions and improvements to encourage them to ride on the street. Focus groups were held in Portland to investigate safety-related e-scooter issues for residents of Black communities and East Portland. Results indicated that these communities are most concerned about racial profiling and harassment, not about having a helmet, lacking safe spaces to learn how to ride, and lacking safe bicycle infra- structure (Portland Bureau of Transportation and Alta Planning & Design 2020). Existing research on barriers to bicycling for Black and Latinx men and women indicates that these groups may perceive and experience safety-related issues and barriers for e-scooter use that are not widely represented in other research and need to be explicitly studied (Brown 2016). For people with disabilities, concerns related to safety included fears of being hit by vehicles, troubles with balance and vision, and encountering parked or moving e-scooters on sidewalks (Portland Bureau of Transportation and Alta Planning & Design 2020). In Chicago, almost half of respondents with vision, ambulatory, or hear- ing disabilities reported that e-scooters on the sidewalk highly incon- venienced them (City of Chicago 2020a). San Francisco’s “lock-to” solution requires e-scooters to use a locking or tethering mechanism. This seems to have addressed concerns about sidewalk parking more completely than efforts in other cities (SFMTA 2019), but more work is needed to understand the spectrum of challenges and to find solu- tions for the disabled community in relation to e-scooters. Helmet Use Helmets are known to protect riders from injury, but usage is con- sistently low across studies. The lack of advanced planning for many scooter trips does not allow for helmet use. Many riders may not own helmets for a variety of reasons. Twenty percent of the respondents to Portland’s survey reported usually or always wearing a helmet, and only 10% reported sometimes wearing a helmet (Portland Bureau of Transportation and Alta Planning & Design 2020). Approximately 80% of e-scooter users in Baltimore’s survey reported not wearing a helmet (Young et al. 2019). During Santa Monica’s pilot, 61% of the citations were given to people under age 16 not wearing a helmet (City of Santa Monica 2019). Barriers to helmet use while riding e-scooters need further study and could be explored in future research. CHAPTER 4 E-SCOOTER INJURIES AND CRASH CONTEXT This section examines the characteristics of e-scooter crashes and injuries to provide insights into trends in human behavior and health outcomes. The following topics are addressed: definitions and data sources used to classify e-scooter injuries, findings from lanes, or on the sidewalk) has also been studied. E-scooter users have consistently requested bicycle infrastructure to feel safe riding e-scooters in the street (Glenn et al. 2020; Portland Bureau of Trans- portation and Alta Planning & Design 2020; Denver Public Works 2019; Chang et al. 2019; Bird Rides, Inc. 2019; Young et al. 2019). For example, 53% of e-scooter users surveyed in Alexandria, Virginia, said that they prefer to ride in bike lanes over trails, sidewalks, and streets (NACTO 2020). In Hoboken, New Jersey, 88% of scooter users surveyed reported that they would feel safer riding in the street if protected bike lanes were present (NACTO 2020). In Kansas City, Missouri, nearly 50% of respondents to a survey of e-scooter users (n = unknown) reported that they would try riding an e-scooter if they felt safer biking on the street (City of Kansas City, Missouri 2020). In a survey of residents from Idaho, Washington, and Oregon cities with micromobility programs, more than 70% of respondents identified the lack of bike and e-scooter infrastructure as a significant deter- rent to multiple forms of micromobility (Pimentel and Lowry 2020). Respondents to Bird Rides’ 2019 user survey (n = 2,045) reported they wanted protected bike lanes (61%) and wider bike lanes (42%) for the purposes of improving perceived safety (Bird Rides, Inc. 2019). The City of Austin also found a clear preference for protected bike lanes, followed by paved urban trails and painted bike lanes, rather than sidewalks or roadways without bike facilities (City of Austin 2019). E-scooter users’ preference to ride in bicycle lanes aligns with pedestrians’ desire that e-scooter users not ride on sidewalks. For example, 34% of respondents to Denver’s e-scooter survey reported having been hit or almost hit by an e-scooter while walking (com- pared with 19% while driving and 8% while bicycling) (Denver Public Works 2019). This high percentage, along with community complaints, led the City of Denver to change its ordinance to allow e-scooter users to use bike lanes during its pilot. Additionally, 28% of Denver’s survey respondents requested more designated places, like bike lanes, to ride if the city wanted to encourage e-scooter use. In Chicago, 20% of survey respondents (riders and nonriders) who had a negative experience with the e-scooter pilot named safety as the reason, while 18% named riding on sidewalks (City of Chicago 2020a). In Austin, there were more comments about sidewalks than any other item in its community survey (City of Austin 2019). Com- munity feedback in San Francisco resulted in requests for increased protected bike lanes and bike racks to improve safety, as well as increased education and incentives for riders (SFMTA 2019). Revealed preference data from Portland suggest that e-scooter users ride in bike lanes when they are present (Portland Bureau of Transportation 2019). Thirty-two percent of e-scooter usage occurred on the bike network in Portland—a large portion, given how bicycle infrastructure and e-scooters are dispersed separately throughout the city (Portland Bureau of Transportation and Alta Planning & Design 2020). Portland’s 2019 pilot also presents evidence that providing comfortable alternatives for e-scooter users can change behavior. E-scooter use decreased substantially along Waterfront Park, a river- side multiuse pathway where e-scooter use was occurring despite a prohibition, after a two-way separated bicycle lane was installed on the street bordering the park (Portland Bureau of Transportation and Alta Planning & Design 2020). Perceived Safety Within Various Groups Perceptions of e-scooter safety appear to correlate with specific user characteristics. Women in Tempe were significantly (p ≤ 0.05) less

8 epidemiological studies identified in the literature review, and the results from an original analysis of data from North Carolina emer- gency departments. How Are Researchers Identifying E-Scooter Injuries? Three main types of epidemiological studies emerged during the liter- ature review of e-scooter injuries. The most common type examined electronic medical records (EMRs) and interviewed patients that met the researchers’ inclusion criteria. Although these studies permitted a comprehensive investigation of patient characteristics and clin- ical outcomes, many had small sample sizes and a low frequency of events because of time and resource constraints. The second type of study utilized electronic health data collected for syndromic and public health surveillance purposes. These types of studies collected less-detailed patient and encounter information but typi- cally required fewer resources and were able to monitor injury trends among e-scooter users over time and within specific populations. The third type of study used the U.S. Consumer Product Safety Commission’s National Electronic Injury Surveillance System (NEISS) injury data, a nationally representative probability sample of 100 U.S. hospitals. The strengths of studies based on NEISS injury data are that they permit national estimates and contain large sample sizes. The limitations include a lack of detailed information about patients and the circumstances of their injuries and the likelihood of injury misclassification. Most studies reviewed EMRs retrospectively and used keyword- based definitions to identify relevant injuries during a defined time frame. Researchers had to rely on text searches and manual reviews because of the lack of pertinent ICD-10-CM (International Classifica- tion of Diseases, Tenth Revision, Clinical Modification) codes, which are used to classify injuries and other diagnoses for the purposes of hospital administration, billing, and clinical research (National Center for Health Statistics 2022). New, e-scooter–specific ICD-10-CM codes were introduced on October 1, 2020, and so were not avail- able to researchers prior to that date. Keywords varied by study, but typical keywords included “scoot,” “scooter,” “electric scooter,” “bird,” and “lime” as well as possible typos and misspellings of key phrases. While medical records provide a high level of detail about each patient, certain information of interest to e-scooter researchers may not be recorded. In an attempt to capture that data, several studies supplemented their chart reviews with patient interviews. In one study, researchers prospectively recruited patients who pre- sented at an emergency department with suspected e-scooter inju- ries; eligible patients who elected to participate in the study were then interviewed about the circumstances of the crash and personal characteristics (Cicchino et al. 2020b). Only a few studies used public health surveillance or syndromic surveillance systems as a data source. The original syndromic sur- veillance systems were developed in the late 1990s and early 2000s to enable early detection of disease outbreaks caused by bioterror- ism by capturing and interpreting health data about a population in almost real time (Henning 2004). Although syndromic surveillance data contain much more limited information than that captured in EMRs, these data are valuable in that they are timely, are used extensively by local and state public health departments, and are population-based. Like EMR data, ICD-10-CM codes are typically used to classify injury mechanisms; however, because these codes were not available until October 2020, keywords were used to identify e-scooter–related injuries. Some studies used NEISS data to describe e-scooter injuries at the national level. NEISS Product Code 5042 (“standup scooter/ skateboard, powered”) is associated with potential e-scooter injuries, but records related to other modes (e.g., skateboards or hover- boards) must be excluded through manual or automated review, along with any potentially miscoded records. NEISS data are less detailed than medical records and lack some information (e.g., hel- met use), but can be used to estimate the incidence of injuries and changes over time. While NEISS is a nationally representative sam- ple, some researchers have noted that it might not reflect the true incidence of e-scooter injuries across the country because e-scooters are concentrated in certain cities, not equally distributed among the sampled areas (Farley et al. 2020). Given that most police-reported crash data require involvement of a motor vehicle and property damage, police-reported crashes fail to capture most e-scooter injuries—the preponderance of which involve falling from the e-scooter or colliding with a person or object other than a motor vehicle (Austin Public Health 2019). Some state depart- ments of transportation (DOTs) have started to record e-scooter– related crashes in separate categories. For example, Tennessee categorizes e-scooter crashes as “Non-Motorized Personal Convey- ance” in its statewide crash database. Overall, the published literature relies on data originating from hospitals. This could lead to biased results since hospital and emer- gency department data tend to overrepresent moderate to severe injuries because less-serious injuries may be treated at outpatient centers or may not receive any formal medical care. Therefore, less is known about minor e-scooter injuries as well as noninjury events (i.e., “near misses”). Who Is Getting Injured on E-Scooters? Recent research shows that males are more likely to ride e-scooters than females and that e-scooters are popular among riders under 40 years old. Injury demographics appear consistent with the rider- ship data. Researchers compared the age and sex of e-scooter injury patients in 26 epidemiological studies. Although there was variation in the studies’ methodologies and inclusion criteria, the data indi- cated that the average age of the patients was in the thirties. Most studies reported that a greater proportion of the injured patients was male; the percentage of male patients ranged from 38% to 100%, while the percentage of female patients ranged from zero to 57%. The injury demographic findings are also consistent with available data about e-scooter fatalities. Among 53 e-scooter traffic fatalities reported globally between March 2016 and March 2021, the mean age was 33 and the median age was 28; nearly 90% of the fatalities were male (Dwyer et al. 2021). Many studies excluded injuries involving children (typically those under the age of 18). Jurisdictions have imposed minimum age limits on the use of shared rental e-scooters—usually 16 or 18 years old— but enforcement mechanisms are unclear. A study in California found that the ages of 27 injured riders (10.8%) fell between 8 and 17, even though there was a state law in effect during the study period that required riders to be at least 16 and rental company agreements required riders to be at least 18 (Trivedi et al. 2019). In a study of

9 Regardless of the measurement method, most e-scooter injuries were low severity, with more severe injuries comprising a small portion of the total reported. The ISS is a widely used anatomical scoring system for patients with multiple injuries. Medical personnel calculate a patient’s ISS by calculating the sum of the squares of the three highest AIS scores (range: “0, no injury” to “6, unsurvivable”) across six body regions: (1) head or neck, (2) face, (3) chest, (4) abdomen or pelvic contents, (5) extremities or pelvic girdle, and (6) external. The resulting ISS pro- vides an estimate of the severity of the patient’s trauma. ISS values are categorized as follows: a score of 1 to 8 is considered “mild,” a score of 9 to 15 “moderate,” a score of 16 to 24 “severe,” and a score of greater than or equal to 25 “critical.” However, if any single AIS body region receives a score of 6, the ISS is recorded as 75, indicating that the injury is unlikely to be survivable. Among five studies that used the ISS to assess the severity of injuries of patients admitted to hospi- tals and trauma centers for e-scooter incidents, the majority of cases were rated as not serious (Campbell et al. 2019; English et al. 2020; Kobayashi et al. 2019; Puzio et al. 2020; Siman-Tov et al. 2017). The median ISS was 1, 5, and 5.5 (mild injury) for three studies, each with sample sizes of roughly 100 (Puzio et al. 2020; English et al. 2020; Kobayashi et al. 2019). Campbell et al. (2019) found a higher median ISS of 9 (moderate injury) among a small sample of 21 patients who required surgery. Siman-Tov et al. (2017) did not report a median ISS but found that 69.8% of patients had a mild ISS between 1 and 8. That figure is slightly higher than the finding of 58% of patients with a mild ISS in Kobayashi et al. (2019). Hospital admission is another way to quantify injury severity, since injuries must meet certain severity thresholds before patients are admitted. One study reported that 26.9% of patients with observed e-scooter injuries were admitted to the hospital for their injuries (Moftakhar et al. 2020). However, a large national study using NEISS data estimated only 9.4% of e-scooter injury cases resulted in hos- pital admission in 2018 (Namiri et al. 2020). Injury severity recorded in police crash reports can also indicate the rate of severe crashes. Among 52 police-reported e-scooter crashes in Nashville, one e-scooter rider died and four were injured (Shah et al. 2021). Although the incidence of medically attended e-scooter injuries varied across studies, the frequency seems to be increasing over time as e-scooters gain popularity. Looking at the U.S. population overall, Farley et  al. found a striking increase in the incidence of patients receiving treatment in an emergency department due to e-scooter injuries between 2014 and 2019—a period that includes the intro- duction of shared rental e-scooter systems in many U.S. cities. In 2014, the incidence of patients admitted or transferred for e-scooter injuries was just 1.53 cases per 100,000 population; in 2019, the incidence was roughly 9.22 per 100,000 population, a percentage change of more than 500% (Farley et al. 2020). The mode by which patients arrive at an emergency department (i.e., private transport or ambulance) can also be used to approxi- mate injury severity. At a crash scene, emergency medical services (EMS) responders assess injury severity for all victims and priori- tize ambulance transportation for those with more severe injuries. Three recent studies that recorded patient transport mode found 24% to 51.6% of patients with e-scooter injuries arrived via ambu- lance (Badeau et al. 2019; Trivedi et al. 2019; English et al. 2020). In a review of 435 manual scooter and e-scooter rider injuries, a Copenhagen, Denmark, study found that a significantly larger 52 e-scooter and motor vehicle crashes in Nashville, Tennessee, 13% of e-scooter riders were younger than 18 years old (Shah et al. 2021). Several studies also reported race and ethnicity data relevant to U.S. researchers. Using a nationally representative NEISS sample of 32,400 weighted injuries, Aizpuru et al. (2019) found that 54.8% of patients were White, 10.9% Black, 0.9% Asian, 0.5% American Indian, and 33.0% unknown; though the NEISS coding manual includes His- panic ethnicity, this information was not reported. Kobayashi et al. (2019) studied e-scooter injuries at three Level I trauma centers in the United States and found 66% of cases were non-Hispanic White, 21% were Hispanic, 8% Asian, and 7% Black. What Characteristics Are Associated with E-Scooter Injuries? Injury Characteristics Much of the reviewed literature contained information regarding the type and severity of injuries associated with e-scooters. Across the literature, head and upper extremity injuries, especially fractures, were prevalent e-scooter injuries, although the frequency varied. Aizpuru et al. (2019) used NEISS data between 2013 and 2017 to study e-scooter injuries and found that the most commonly injured part of the body was the head, accounting for 27.6% of all inju- ries observed. Fractures and dislocations were the most common injury diagnoses (25.9% of injuries) and the most common fracture sites were the lower arm and wrist (Aizpuru et al. 2019). Similarly, Kobayashi et al. (2019) and Trivedi et al. (2019) studied patient medical records (n = 103 and n = 249, respectively) and observed that fractures (extremities and facial) and head injuries were the most common injuries. A study of 175 e-scooter patients in Vienna, Austria, found that, compared with other common types of injuries, head injuries were most likely to be severe (Moftakhar et al. 2020). In observing trends over time, Farley et al. (2020) noted that statistically significant increases in e-scooter–related traumatic brain injuries and fractures (any body part) were seen in U.S. emergency departments between 2014 and 2019, a time period encompassing the introduc- tion of shared rental e-scooters (Farley et al. 2020). Several studies found other types of injuries to be more preva- lent than fractures, although they occurred in similar locations of the body (the head and upper extremities). DiMaggio et al. used NEISS data to examine e-scooter injuries from 2000 to 2017 and concluded that soft tissue injuries comprised 53.6% of injuries, followed by frac- tures at 26.5% (DiMaggio et al. 2019). In the Vienna study, among serious head injuries, the most frequent injuries were not fractures, but soft tissue injuries (including abrasions, lacerations, or hemato- mas) and contusions, which occurred in 61.7% and 46.8% of patients, respectively. It is worth noting that most injury categories are not mutually exclusive, and many patients present with more than one injury type or location. Injury Severity There are many methods of measuring injury severity, and studies have differed in their approaches. Some used standardized scales such as the Injury Severity Score (ISS) and the Abbreviated Injury Scale (AIS), in which a patient’s injuries correspond to a number on a scale of least to most severe or urgent. Other studies used proxies for severity (e.g., hospital admission, ambulance transport, length of stay, medical cost) or original categories to describe severity.

10 proportion of injured e-scooter riders (67.9%) than manual scooter riders (40.1%) were transported by ambulance (Blomberg et al. 2019). Few studies reported the length of hospital stay—another indicator of injury severity—for admitted patients. In a study of 175 e-scooter injuries in Austria, Moftakhar et al. (2020) reported a mean length of stay of 6.5 days and a range of 1 to 115 days). Liew et al. (2019) reported a median length of stay of 2 days in a study of 36 e-scooter injuries in Singapore. Many low-income and minority individuals may not seek help or may only seek help when injuries are severe. Some concerns include cost, mistrust, fear of deportation, and fear of calling 911 and inter- acting with law enforcement. For these reasons, underreporting is likely, especially among historically marginalized populations, as suggested by previous studies that have examined this issue among pedestrians and bicyclists (Doggett et al. 2018). Although most e-scooter injuries are not severe, the increase in the overall incidence of e-scooter injuries (coinciding with the popularity of shared rental e-scooter systems) means that a greater number of severe and nonsevere injuries are occurring. More U.S. studies are needed for comparison to gain a better understanding of injury severity in the United States, as some metrics for severity crite- ria may differ across countries. This increase in injuries merits further attention from researchers and policymakers. What E-Scooter Crash Characteristics, Types, and Scenarios Are Most Common? Because epidemiological studies have largely been at the fore- front of the research on e-scooter crashes, there has been very little research on e-scooter–related crash types and scenarios to date. However, existing research can bring to light the factors associated with e-scooter incidents and help guide future research. Location of E-Scooter Incidents It is important to identify contextual trends in e-scooter crashes, such as where crashes typically occur in the cross section of the roadway. Some studies have explored this crash factor, including an Austin study of e-scooter–related injuries that found the percentage of incidents that occurred in the street and on the sidewalk to be 55% and 33%, respectively (Austin Public Health 2019). Similarly, results from San Francisco’s e-scooter pilot program showed that collisions occurred most frequently on roadways (83%), with a small portion occurring on sidewalks (10%) and in bike lanes (7%) (Vision Zero San Francisco Injury Prevention Research Collaborative 2019). Stronger evidence of sidewalk incidents was found in Washington, DC, where the proportion of e-scooter–related injuries that occurred on side- walks, roads, off-road locations, and bike lanes was 58%, 23%, 10%, and 8%, respectively (Cicchino et al. 2020b). It has frequently been asserted that the behavior of sidewalk riding and resulting crashes may indicate that the street, where adequate bicycle infra- structure may be lacking, is not a place where e-scooter riders feel safe (Santacreu et al. 2020; Portland Bureau of Transportation and Alta Planning & Design 2020; NACTO 2020). After PeopleForBikes conducted a bicycle network analysis (BNA) in more than 500 cities in the United States and Canada, Bird compared e-scooter crash rates for its vehicles with the BNA scores and found that cities with higher BNA scores showed evidence of being safer for e-scooter use (Santacreu et al. 2020). Few studies have evaluated the location of e-scooter crashes with respect to roadway segments. In one Nashville study, more than 65% of e-scooter crashes involving a motor vehicle occurred at intersections, followed by 17% of those crashes at driveway-to roadway junctions and 13% at a nonjunction segment of the road (Shah et al. 2021). E-scooter riders riding on the sidewalk collided with a motor vehicle at the driveway or crosswalk in most of these crashes. Roadway Conditions at the Site of E-Scooter Incidents The physical attributes of the roadway are significant factors in e-scooter–related incidents. In Austin, 10% of reported e-scooter incidents involved curbs and an additional 7% involved generally stationary objects (e.g., light poles or manhole covers) (Austin Public Health 2019). A San Francisco trauma center had similar findings, where 11% of e-scooter–related injuries that required trauma team activation were the result of riders colliding with stationary objects (Vision Zero San Francisco Injury Prevention Research Collaborative 2019). Roadway grade is infrequently identified as a factor contrib- uting to e-scooter incidents, but the aforementioned Austin study noted that most e-scooter crashes occurred on a level surface (65%) or downhill grade (24%), with a small portion occurring on an uphill grade (6%) (Austin Public Health 2019). These findings suggest that grade was not a significant factor in e-scooter incidents; however, this possibility could be explored in future research. The conditions of the roadway surface have been commonly attributed as crash factors for those involved and injured in e-scooter incidents. E-scooters are more susceptible to irregularities on the roadway than bicycles (Santacreu et al. 2020). One study specifically found that half of the riders interviewed claimed that the roadway surface conditions were a factor in their injury (Austin Public Health 2019). These claims are corroborated by a study of e-scooter injuries in Washington, DC, that found 25% of e-scooter injuries were a result of poor roadway surface conditions (e.g., potholes or uneven sur- faces). Even after the various locations on the roadway cross section where incidents took place were taken into account, adverse road conditions remained a consistent factor in crashes on roads (33%), sidewalks (25%), and bike lanes (25%). Nearly two-thirds of all inci- dents in the study were caused by adverse roadway features when accounts of additional adverse infrastructure features were included (e.g., driveway lips) (Cicchino et al. 2020b). Common E-Scooter Crash Scenarios One crash scenario that has been shown to constitute a significant, though varying, portion of e-scooter–related incidents is the single- vehicle incident, or falling (Santacreu et al. 2020). This is consistent with similar patterns in bicycle crashes. As previously indicated, approximately two-thirds of e-scooter–related incidents in a study of Washington, DC, were caused by physical features of the road- way, as opposed to collisions with or attempts to avoid vehicles or other road users (Cicchino et al. 2020b). In addition, San Francisco pilot program companies reported that single-vehicle crashes constituted 38% of all e-scooter incidents (Vision Zero San Francisco Injury Prevention Research Collaborative 2019). In Portland, 83% of e-scooter–related emergency department visits were attributed to falling (no collision) (Portland Bureau of Transportation 2019). Compared with Portland, San Francisco reported a lower percent- age of injuries involving falls from e-scooters (33%) (Portland Bureau of Transportation 2019, Vision Zero San Francisco Injury Prevention

11 design, roadway characteristics, rider education, or other factors that could be altered to minimize such incidents. Motor Vehicle–E-Scooter Collisions Only a small proportion of nonfatal e-scooter incidents involves motor vehicles. In Portland, 12.5% of emergency department visits related to e-scooters were caused by collisions with cars, with an additional 1.1% caused by collisions with trucks (Portland Bureau of Transportation 2019). The distinction between cars and trucks has seldom been made in e-scooter incident studies, despite its poten- tial relationship to severity. An Austin study found that motorized vehicles were involved in 16% of the incidents, with 10% resulting in a collision. In instances that did not result in a collision, riders reported swerving, stopping, or jumping off e-scooters (Austin Public Health 2019). Similarly, moving vehicles caused 10% of e-scooter– related injuries in Washington, DC (Cicchino et al. 2020b). In the same study, an additional 4% of injuries were caused by riders attempting to avoid moving vehicles. In San Francisco, pilot program e-scooter companies reported that motor vehicles were the cause of 44% of e-scooter collisions. The same proportion of trauma center patients with e-scooter–related injuries in San Francisco had experi- enced collisions with motor vehicles, but this unusually high propor- tion was attributed to the small sample size and the severity of injury of patients with e-scooter–related trauma (Vision Zero San Francisco Injury Prevention Research Collaborative 2019). The literature has consistently shown evidence that motor vehi- cles are frequently involved in e-scooter–related fatalities. The Inter- national Transport Forum (ITF) found that motor vehicles such as cars and trucks are involved in more than 80% of e-scooter and bicycle accidents with rider fatalities (Santacreu et al. 2020). In a safety study conducted by an e-scooter company, 10 fatalities occurred in the almost 38 million trips that riders completed during the 1-year study period, nine of which involved a motor vehicle (Murray 2020). Pedestrian–E-Scooter Collisions On occasion, pedestrians are involved in e-scooter crashes. In San Francisco, only 11% (n = 1) of e-scooter–related injuries that merited trauma team activation were pedestrians (nonriders) who were hit by an e-scooter. Additionally, pedestrian and e-scooter collisions repre- sented just 12% of e-scooter crashes recorded and reported by the companies involved in San Francisco’s pilot program (Vision Zero San Francisco Injury Prevention Research Collaborative 2019). In Portland, an even smaller portion (1.7%) of emergency department visits asso- ciated with e-scooters were caused by a collision with a pedestrian (Portland Bureau of Transportation 2019). An investigative study of e-scooter–related emergency department visits in Washington, DC, reported that one out of 105 participants sustained injuries as a result of being struck by a moving e-scooter (Cicchino et al. 2020b). Pedes- trian involvement in e-scooter collisions may be tied to instances of sidewalk riding. The same Washington, DC, study found that 12% of injuries that occurred on sidewalks involved efforts to avoid hitting pedestrians. Although rare, e-scooter collisions do involve pedes- trians. ITF analyzed 38 media reports of standing e-scooter fatalities that occurred between September 2018 and the end of October 2019 and found that pedestrian fatalities represented less than one in 10 e-scooter–related fatalities (Santacreu et al. 2020). Some e-scooter incidents involve pedestrians and stationary or parked e-scooters that present tripping hazards, but this has not been Research Collaborative 2019). However, San Francisco’s study pop- ulation involved patients with injuries of sufficient severity to trigger activation of a full trauma team. Since falls are less likely to result in serious injuries than other mechanisms, injuries due to motor vehi- cle collisions were overrepresented in this study. In addition, the study performed by the San Francisco Vision Zero Collaborative consisted of only nine patients, which limited the conclusions that can be drawn from the research (2019). Recent epidemiological studies also show a large portion of injuries resulting from single-vehicle crashes and, in particular, falls. For example, in a study of adults with injuries incurred in using shared rental e-scooters in Austin between September and November 2018, English et  al. (2020) found that a fall from the scooter was the cause of injury in 105 out of 124 cases (84.7%). The other mechanisms of injury included the following: struck by an object (15.3%), struck by a moving vehicle (9.7%), struck a parked vehicle (4.0%), pedestrian struck by scooter (1.6%), unknown (0.8%), and multiple mechanisms (15.3%) (English et  al. 2020). Similarly, Bloom et al. (2020) studied standing scooter injuries in 2018 in Los Angeles and found that 121 of 248 patient records (49%) noted that a loss of balance led to the injury, compared with 35 incidents (14%) in which a collision with an automobile was the mechanism and 25 incidents (10%) where uneven pavement was the mechanism. In a study of manual and electric scooter injuries in Copenhagen, Blomberg et al. (2019) found that of 112 reported injuries to e-scooter riders, 97 (86.6%) resulted from a fall, 5 (4.5%) resulted from a collision with an object, and 10 (8.9%) resulted from a collision with a vehicle or moving object. In addition to stationary objects and adverse surface conditions, various other factors have been attributed to e-scooter injuries, some of which may lead to falling without a collision. For example, one study found that hardware failure or malfunction contributed to e-scooter–related injuries (Santacreu et  al. 2020). Mechanical issues, such as defective brakes, a sticky throttle, or faulty steering, were indicated as crash factors in 16% of e-scooter–related injuries in Washington, DC (Cicchino et al. 2020b). Another study in Austin found that 19% of injured e-scooter riders attributed their crash to vehicle issues, while 37% of riders attributed their crash to extremely high e-scooter speeds (Austin Public Health 2019). Contribution of Inexperience to E-Scooter Incidents There is also evidence that inexperience plays a role in the likeli- hood of e-scooter incidents. A safety study that explored a year of e-scooter crash data in the United States found that the proportion of incidents occurring during a rider’s first trip and within a rider’s first five trips was 23% and 36%, respectively (Murray 2020). Austin Public Health documented the experience level of each injured rider and found that 33% of injuries occurred in first-time e-scooter riders, compared with about 30% of injuries to riders with one to nine previous rides, 22% to riders with 10 to 29 rides, and 15% to those with more than 30 rides (Austin Public Health 2019). Similar results were found in Washington, DC, where 37% of riders reported that their injuries occurred during their first ride (Cicchino et al. 2020b). In response to these findings, one e-scooter company has introduced an e-scooter “comfort mode” that will enable new riders to select a lower speed limit for their vehicle (Murray 2020). Future research could focus on identifying the specific circumstances that lead to single-vehicle e-scooter crashes and falls and the aspects of vehicle

12 of helmets among injured e-scooter users is lower than that of all e-scooter riders, perhaps because riders wearing helmets are less likely to experience serious injuries because of the protective nature of the helmets. Helmet use among injured riders in 15 studies fell in the range of zero to 9% (excluding one finding of 25% in San Francisco, owing to a small sample size of n = 8). Although in many cases the use of protective equipment was not known or not recorded, confirmed helmet use was similar across studies, regardless of the percentage of unknown helmet use. One study in Los Angeles reviewed medical records and conducted an observational study of local e-scooter riders to compare helmet use between injured riders and the overall e-scooter rider populations. This study found that 11 out of 193 riders (6%) were wearing helmets, while the injury records showed only 4% helmet use among injured riders (Trivedi et al. 2019). Low helmet use among e-scooter riders is a troubling statistic with potential implications for injury severity, given the known protective benefits of helmets and the notable frequency of head injuries. In Washington, DC, a study comparing injuries to riders of personally owned bicycles with riders of shared e-scooters found that injured e-scooter riders wore helmets significantly less often than injured bicyclists (<2% of e-scooter riders vs. 66% of bicyclists) (Cicchino et al. 2020a). In that study, e-scooter riders were almost three times more likely than bicyclists to experience a concussion and lose consciousness (Cicchino et al. 2020a). What Role Does Reported Impairment Play in Incidents of E-Scooter Injury? The heterogeneity of study methods creates a challenge to com- paring findings across studies and estimating the prevalence of alcohol or drug impairment in e-scooter incidents. For example, Austin Public Health (2019) found that 29% of interviewed riders reported drinking an alcoholic beverage in the 12 hours preceding their injury. However, an affirmative answer does not necessarily mean that the rider was intoxicated when the injury occurred, nor does it indicate the level of intoxication. A few studies examined toxicology results, which are a more objective measure of impairment than riders’ self-reports, but have their own limitations. Notably, there are few clinical models for ascertaining the relationship between the level of substance observed and the level of impairment for drugs other than alcohol. Among studies that analyzed toxicology results, findings varied. In a study of e-scooter injury patients in the United States (N = 103), Kobayashi et al. (2019) found that relatively high numbers of these patients underwent alcohol and drug testing, with high proportions receiving positive results. Among the 79% of patients who were tested for alcohol at the time of hospital admission, 48% were legally intoxicated (blood alcohol level > 0.08%). Sixty percent of patients received a urine toxicology screen, and, of those, 52% were positive for various substances [e.g., tetrahydrocannabinol (THC), metham- phetamines, and amphetamines]. One study of injured e-scooter riders in Indianapolis (n = 92) found that 33% had consumed alco- hol, according to lab testing and physician assessment (Puzio et al. 2020). In contrast, Trivedi et  al. (2019) determined that just 12 of 249 patients (4.8%) treated for e-scooter–related injuries in Los Angeles exceeded a blood alcohol level of 0.05% or were deter- mined to be intoxicated by a physician. Neither Puzio et al. nor Trivedi et al. reported the proportion of patients screened for alcohol. extensively studied. Out of 18 injuries to nonriders in Copenhagen between 2016 and 2019 (representing 13.8% of all e-scooter injuries during that time), 10 patients were hit by e-scooters and eight were injured by tripping over a parked e-scooter (Blomberg et al. 2019). Similarly, a Los Angeles study of 249 e-scooter injuries reported 21 (8.4%) nonrider pedestrian injuries—11 pedestrians were involved in a collision with a scooter, five tripped over a parked scooter, and five were injured while attempting to lift or move a scooter (Trivedi et al. 2019). As noted previously, prior to October 1, 2020, health care data did not contain codes identifying injuries related to being struck by an e-scooter or tripping over an e-scooter, and, therefore, e-scooter–involved pedestrian injuries may be underreported in hos- pital or public health surveillance data. Frequency and Severity of E-Scooter Incidents by Time of Day and Day of Week It is unclear whether there are consistent trends in the time of day and week that e-scooter incidents occur, given the varied findings and the operational hours permitted by cities. Many of the early deployments were in cities with vibrant entertainment districts. For example, in Austin, the days of the week with the highest frequency of reported injuries were Saturday and Sunday. In addition, many e-scooter riders reported being injured during the overnight hours, with 39% of injuries occurring between the hours of 6:00 p.m. and 6:00 a.m. (Austin Public Health 2019). Similarly, a study of e-scooter crashes in Washington, DC, found that 33% of injuries took place on the weekend (Cicchino et  al. 2020b). However, reporting from the San Francisco Police Department showed that San Francisco e-scooter collisions took place most frequently between 3:00 and 8:00 p.m., whereas the hours of midnight to 7:00 a.m. saw no colli- sions. Companies that participated in San Francisco’s pilot program reported similar findings, with just 18% of collisions occurring in the evening hours (Vision Zero San Francisco Injury Prevention Research Collaborative 2019). The Washington, DC, study found that only 14% of e-scooter–related injuries happened between 9:00 p.m. and 6:00 a.m. (Cicchino et al. 2020b). Similarly, 69% of the 52 crashes involving e-scooters and motor vehicles in Nashville occurred during daylight conditions (Shah et al. 2021). Because some e-scooter com- panies only allow their devices to operate during certain hours (typ- ically daytime), it is likely that the reduced availability of scooters contributed to the lower frequency of nighttime crashes observed in these studies. While most studies have suggested that e-scooter incidents are more likely to occur during daytime hours, data on e-scooter fatalities for which information on time of death is available show that most of those crashes occurred during the evening and nighttime hours (Santacreu et  al. 2020; Dwyer et  al. 2021). These varied findings suggest that more research is needed to investigate the relationship between the frequency and severity of e-scooter injuries (fatal and nonfatal) and the time of day. What Is the Prevalence of Helmet Use Among Injured Riders? Helmets are known to reduce the risk of serious head injury, including traumatic brain injury, yet helmet use among e-scooter riders is low. Helmet mandates and enforcement also carry important equity implications. A variety of approaches exists to encourage helmet use without a formal mandate. Existing literature indicates that the use

Next: Chapter 5 E-Scooter Program Safety Management Practices »
E-Scooter Safety: Issues and Solutions Get This Book
MyNAP members save 10% online.
Login or Register to save!
Download Free PDF

Many communities with electric‐scooter (e‐scooter) programs have observed social, health, and environmental benefits; enhanced multimodal connections; and positive economic impacts (such as those derived by delivery services and couriers using e‐scooters and the resultant jobs created). However, these effects are often accompanied by real and perceived safety challenges.

The TRB Behavioral Transportation Safety Cooperative Research Program's BTSCRP Research Results Digest 1: E-Scooter Safety: Issues and Solutions is an initial deliverable to a larger ongoing project, in the form of a literature review, that identifies emerging behavioral safety issues arising from the expanding use of e-scooters and summarizes how cities are working to prevent and mitigate injuries.

  1. ×

    Welcome to OpenBook!

    You're looking at OpenBook,'s online reading room since 1999. Based on feedback from you, our users, we've made some improvements that make it easier than ever to read thousands of publications on our website.

    Do you want to take a quick tour of the OpenBook's features?

    No Thanks Take a Tour »
  2. ×

    Show this book's table of contents, where you can jump to any chapter by name.

    « Back Next »
  3. ×

    ...or use these buttons to go back to the previous chapter or skip to the next one.

    « Back Next »
  4. ×

    Jump up to the previous page or down to the next one. Also, you can type in a page number and press Enter to go directly to that page in the book.

    « Back Next »
  5. ×

    To search the entire text of this book, type in your search term here and press Enter.

    « Back Next »
  6. ×

    Share a link to this book page on your preferred social network or via email.

    « Back Next »
  7. ×

    View our suggested citation for this chapter.

    « Back Next »
  8. ×

    Ready to take your reading offline? Click here to buy this book in print or download it as a free PDF, if available.

    « Back Next »
Stay Connected!