This cross-sectional study was performed using data from the Emergency Department-based Injury In-depth Surveillance (EDIIS) database sponsored by the Korea Disease Control and Prevention Agency (KDCA). The EDIIS is a prospective nationwide injury database that stores data on patients with injuries who present to the emergency rooms (ERs) of specific representative medical institutions in Korea.

In total, 17 (in 2014), 20 (in 2015), and 23 (from 2016 to 2018) tertiary academic hospitals shared data on patients with injuries who presented to ERs using the EDIIS database. The EDIIS database includes demographic, injury prevention–related, and epidemiologic information as well as pre-hospital procedures, initial clinical findings from the ER visit, diagnoses (coded using the International Classification of Diseases, 10th Revision [ICD-10]), treatments, dispositions, and patient outcomes after admission. Individual incidents were described in written Korean. Primary information was obtained by physicians from each medical institution during clinical practice and by trained coordinators from the EDIIS project assigned to each hospital. The coordinators collected the data from a standardized registry. The data from each ER were registered on a web-based database hosted by the KDCA, and a quality improvement program was conducted regularly [12].

Among all injuries reported in the EDIIS database between January 2014 and December 2018, only patients who were injured while operating an ePMD were included in this study. To identify ePMD-related injuries, we searched the database with ePMD-related keywords, including “electric scooter”, “e-scooter”, “electric kickboard”, “e-kickboard”, “hoverboard(s)”, “electric unicycle”, “e-unicycle”, “Segway”, “Lime scooter”, and “Ninebot”, and two researchers reviewed the incident descriptions. After reviewing each entry, data were excluded if 1) the injury was caused by an ePMD used as an aid for a disabled person, such as an electric wheelchair or mobility scooter, or 2) the accident was not related to the operation of the ePMD.

This study collected information for each injury on age, sex, injury date, device type, injury mechanism, injury location, the rider’s state of intoxication, helmet use, diagnosis (ICD-10 code), and disposition. Age was classified as 0 to 14, 15 to 24, 25 to 34, 35 to 44, 45 to 54, and 55 years or older, based on previous studies [13]. An “e-scooter” was defined as a device powered by an electric motor with wheels and handlebars that is designed to be stood upon by the operator. An “electric unicycle” and “hoverboard” were defined as narrow, horizontal boards with one or two wheels, respectively, that move when the rider leans forward. Mechanisms of injury were categorized as falloff, collision with a motor vehicle, collision with a human, collision with another ePMD, or other. Possible injury locations were indoors, public property (a car-free public facility), street (a thoroughfare for cars), sidewalk (a foot traffic-only pathway), bike-way (a bicycle-only pathway), driveway/parking lot, and alley (a road without sidewalks). Anatomical injury sites were categorized as head and face, neck, torso (including the thorax, abdomen, back, pelvis, and genitals), and upper and lower extremities (including the shoulders, upper arms, elbows, forearms, wrists, hips, thighs, knees, lower legs, ankles, and feet) according to the ICD-10 codes for the injury mortality diagnosis matrix by the Centers for Disease Control and Prevention of the United States [14].

To assess injury severity, the excess mortality ratio-adjusted Injury Severity Score (EMR-ISS) was calculated using ICD-10 codes. The EMR-ISS classifications used in this study were mild (scores 1–8), moderate (scores 9–24), severe (scores 25–74), or critical (scores ≥75 or death), as in a previous study [15].

The primary outcome of the study was the incidence of severe injury. The secondary outcome was the incidence of acute traumatic brain injury, defined by ICD-10 codes of S02.0xx, S02.1, S06.2, and S06.3x [16]. The tertiary outcome was the rate of intensive care unit (ICU) admission.

All statistical analyses were performed using STATA version 14.2 (StataCorp LP, College Station, TX, USA). Continuous variables were presented as medians with interquartile ranges (IQR), and categorical variables were presented as frequencies with percentages. To identify statistically significant differences between the outcome groups, we used the Wilcoxon rank-sum test for continuous variables and the chi-square test or Fisher exact test for categorical variables. Odds ratios (ORs) with 95% confidence intervals (CI) were calculated using multivariate logistic regression analysis to evaluate the factors associated with the outcomes. The level of statistical significance was defined as a P-value of ≤0.05.

Of the 1,391,980 patients injured between 2014 and 2018, 1,472 patients (0.11%) had records that included ePMD-related keywords in their descriptions. After excluding ineligible cases, 684 cases (0.05%) were ultimately used in the final analysis. Among them, 505 cases (73.8%) were related to e-scooters and 179 cases (26.2%) were related to electric hoverboards (Fig. 1).

Fig. 2 shows the trends in ePMD-related injuries among the study population. The rate of ePMD-related injuries due to e-scooters or hoverboards increased from 3.1 ePMD-related ER visits per 100,000 population in 2014 to 100.3 per 100,000 population in 2018.

Table 1 shows demographic characteristics and clinical features according to ePMD type. The median age was 28 years old (IQR, 19–38 years old) and a higher proportion of men (68.0%) sustained injuries related to ePMDs. Patients who were injured while operating electric hoverboards tended to be younger than those who were injured while operating e-scooters. The distribution of the time of injury was different between the two groups. The rate of helmet use was low in both groups, at 3.4% for the e-scooter group and 2.2% for the hoverboard group (P=0.210), and operation while intoxicated was more common in the e-scooter group than in the hoverboard group. Fall-off injuries were the most common injury mechanism, at 69.9% in the e-scooter group and 83.8% in the hoverboard group; however, collisions with vehicles or stationary objects were more frequent in the e-scooter group than in the hoverboard group. A higher proportion of injuries in the e-scooter group had EMR-ISS classifications of moderate or severe, and the e-scooter group also had a higher ICU admission rate than the hoverboard group. One death was recorded in the e-scooter group.

Table 2 shows the clinical results and demographic findings by age group. Injuries related to e-scooters were more common than injuries related to hoverboards across all age groups. The rate of emergency medical service utilization was highest among those aged >55 years old. Those aged 0 to 14 years old experienced the most injuries on public property, while the majority of injuries occurred on the street or sidewalk for the other age groups.

The results of the multivariate logistic regression models are shown in Table 3. The occurrence of severe injury, traumatic brain injury, and ICU admission was higher among those aged 55 and older. There were no cases of traumatic brain injury or ICU admission when the rider wore a helmet. The likelihood of a severe injury (i.e., with an EMR-ISS of ≥25) was higher for men and for patients who had consumed alcohol before riding (adjusted OR [AOR], 2.11; 95% CI, 1.14–3.90 for men and AOR, 2.78; 95% CI, 1.52–5.08 for those with alcohol consumption). Moreover, patients who sustained injuries on roads or streets (AOR, 2.68; 95% CI, 1.11–6.45) were more likely to be admitted to the ICU than patients injured elsewhere.