Identifying community safety and policy-based injury prevention opportunities to reduce golf cart injuries

Abstract

BACKGROUND:

In small US communities, golf cart utilization has become increasingly more common. In the past 3 years, the incidence and severity of pediatric golf cart–related trauma evaluated at our trauma center have noticeably increased. Thus, the aim of this study was to analyze trends, identify risk and protective factors, and provide community-level recommendations to improve golf cart safety for children in a coastal community.

METHODS:

A retrospective cross-sectional study of our institutional trauma registry was performed. The registry was queried for golf cart injuries between 2012 and 2022. Demographics, accident details, hospital course, and outcomes were reviewed. Data analysis involved quantitative statistics. Incident locations were mapped, including additional data from the County emergency medical service. In addition, customer education at four prominent golf rental shops was observed.

RESULTS:

Annual golf cart–related traumas doubled starting in 2020. Of 235 total patients, 105 (46%) were children. Median age was 11.5 years (range, 2–17 years). Fifty-five percent were female, and 67% were non-Hispanic White. Eighty percent were out-of-county residents. The most common injury location was extremity (56%). The median Injury Severity Score was 4, and 3% died. Only 10% of children were restrained. Forty-one percent were ejected, and most (84%) were front-facing passengers. Ejection was associated with more severe injury (odds ratio, 4.13; p = 0.01). Most injuries occurred during 5 to 10 PM (47%), weekends, and summertime. Nighttime injuries were more severe than daytime (p = 0.04). A hotspot of crashes was identified in a zone where golf carts were restricted. Rental stores provided education on seat belt use, car seat use for infants, and off-limit zones. However, rules were not enforced.

CONCLUSION:

Our results inform the following golf cart injury prevention opportunities: raising awareness of injury risks to children in high-tourist areas, partnering with rental stores to enforce rules, improving signage, adding protected lanes, and adopting a no nighttime operation policy.

LEVEL OF EVIDENCE:

Prognostic and Epidemiological; Level IV.

There has been a significant increase in golf cart popularity, beyond the golf course, as they offer a more convenient alternative to traditional cars.^1,2 The number of golf carts sold in the United States is estimated to increase at an annual rate of 3.2% or approximately 76,685.³ However, the increased utilization is accompanied by an increasing incidence of golf cart–related injuries, with an estimated 6,500 children injured each year.^4–6 This issue is further compounded by the lack of standardized safety features and laws for operation on public roads.

The Centers for Disease Control and Prevention has created a systematic framework for addressing injury prevention.⁷ This process consists of (1) defining the problem, (2) identifying and risk and protective factors, (3) developing and testing prevention strategies, and (4) assuring widespread adoption. Few studies have defined the incidence of golf cart–related trauma and identified risk factors associated with increased crash likelihood and injury severity.^8–13 Because of the heterogeneity of community factors and laws regarding golf carts, data that are specific to local communities are most useful for addressing the current gap in knowledge.^11–18

In the past 3 years, the incidence and severity of pediatric golf cart–related trauma evaluated at our trauma center have noticeably increased. Within our coastal community, there are an estimated 1,700 golf cart owners and renters.¹⁴ Although there are local restrictions prohibiting the use of golf carts on certain streets, the rise in golf cart crashes leading to injuries and deaths of children have gained media attention and an identified need for injury prevention.^14–17 The aim of this study was to analyze recent trends in pediatric golf cart injuries, identify risk and protective factors, and provide community-level recommendations to improve golf cart safety for children.

PATIENTS AND METHODS

Study Design

This study was approved by our institutional review board. Informed consent was not required because of the retrospective nature of this study. Our level 1 trauma center’s institutional trauma registry was queried for patients of all ages involved in golf cart crashes between January 2012 to December 2022. To minimize missed cases, the county emergency medical services (EMS) records were queried for all golf cart crashes between January 2017 and December 2022. Emergency medical services records from prior years were unavailable. Retrospective electronic medical record (EMR) review was performed on the combined list of patients. Individuals not evaluated at our institution were excluded from the characteristic and prediction analysis, but these individuals were included for the geospatial analysis. Data elements collected included age, sex, race, ethnicity, county of residence, time and date of accident, day of week, address of crash, seatbelt use, cause of crash, speed, intoxication status, Injury Severity Score (ISS), injuries sustained, disposition, and length of stay. Patients were excluded from specific analyses if data elements were missing. Demographic information, accident date and time, and patient position were available for 100% (235 of 235) of patients. Restraint status (92.7%, 218 of 235), ejection (97.8%, 230 of 236), unique causes per crash (35.7%, 60 of 169), and reported speed (25.9%, 61 of 235) were limited by reporting.

Cause of crash was classified as follows: (1) impact with pedestrian, (2) impact with car, (3) impact with golf cart, (4) impact with golf cart or all-terrain vehicle, (5) stationary object (e.g., light pole), and (6) hard cornering (turning too fast causing loss of control). Categories were determined based on documented crash descriptions listed in the history and physical. Passenger position was categorized as follows: (1) driver, (2) front-facing passenger, and (3) rear-facing passenger. Seatbelt status was therefore categorized as follows: (1) restrained or (2) unrestrained with the assumption that most seatbelts would be in the 2-point configuration.

The STROBE guideline was used to ensure proper reporting of methods, results, and discussion (Supplemental Digital Content, Supplementary Data 1, http://links.lww.com/TA/D616).

Statistical Analysis

Descriptive statistics were used to describe patient demographics, accident details, and seatbelt usage. Temporality and frequency of injury by anatomic site were analyzed using continuous frequency distributions. Logistic and linear regressions were used to identify associations. χ² and Fisher’s exact test were used to analyze associations for discrete variables. Results were deemed significant at a p value of <0.05. All statistical analyses were performed using Prism 10 (GraphPad, Boston, MA).

Geospatial Analysis

Addresses for all individuals involved in golf cart crashes evaluated at our level 1 trauma center or recorded into the county EMS records were compiled. Multipassenger accidents were aggregated into a single address. Addresses were then converted to latitude and longitude coordinates using SAS software (SAS Institute, Cary, NC). For crash sites that failed to match because of incomplete address information or recorded as an intersection of two streets, the coordinates were manually obtained using Google Maps. SAS was able to geocode 68% (n = 114) of the sites, and Google Maps, another 26% (n = 44) with the remaining 5% (n = 9) lacking sufficient information to identify the location.

Customer Education and Safety Assessment at Rental Shops

Customer education and the level to which golf carts were equipped with city mandated safety features was evaluated. Researchers visited the four largest golf cart rental shops on the island and assessed the rental process for elements of the safety briefing and safety equipment installed on the vehicles (Supplemental Digital Content, Supplemental Table 1, http://links.lww.com/TA/D617).

RESULTS

From January 2012 to December 2022, a total of 235 patients involved in golf cart crashes were evaluated at our trauma center. This translated to 168 unique crashes when multipassenger incidents were aggregated. Of these, 105 patients (46%) were children 0 to 17 years old. A total of 138 (58%) were female. The mean ± SD age was 26.2 ± 17.6 years (children, 11.7 ± 4.06 years; adults, 38.1 ± 15.5 years). Most patients were White non-Hispanic (162, 69%), followed by White Hispanic (38, 16%), and Black non-Hispanic (31, 13%). Visitors from outside counties comprised 173 patients (77%). The incidence of golf cart crashes steadily increased each year with the sharpest rise in 2020 by 110%, and this rise has remained consistent since then (Supplemental Digital Content, Supplementary Fig. 1, http://links.lww.com/TA/D617). Crashes were more frequent on the weekends, during the summer, and after 4 PM (Fig. 1A–C). Table 1 reports the demographic and outcome differences between pediatric and adult cases.

Figure 1.

Temporality of golf-cart related traumas. (A) Distribution of cases across days of the week. (B) Distribution of cases based on month of the year. (C) Distribution of cases across time of day. Of note, sunset occurs between the hours of 1800–2000.

TABLE 1.

Comparison of Pediatric and Adult Patients Who Were Injured in Golf Cart Crashes

Demographics
Variable	Overall N = 235	Adults n= 130 (55%)	Children n = 105 (46%)	p
Age, mean ± SD (range)	26.2 ± 17.6 (2–85)	38.1 ± 15.5 (18–85)	11.7 ±4.06 (2–17)	—
Sex, n (%)
Female	138 (58)	81 (62)	58 (55)	0.31
Male	97 (41)	50 (38)	47 (45)
Race/ethnicity, n (%)
White non-Hispanic	162 (69)	92 (70)	70 (67)	0.54
White Hispanic	38 (16)	18 (14)	20 (19)	0.36
Black non-Hispanic	31 (13)	16 (12)	15 (14)	0.59
Black Hispanic	1 (<1)	1 (<1%)	0 (0%)	1.00
Other non-Hispanic	3 (1)	2 (1)	1 (<1)	1.00
Residence, n (%)
County	51 (23%)	30 (25%)	21 (20%)	0.35
Out of county	173 (77%)	89 (75%)	84 (80%)
Injury Outcomes
Variable	Overall N = 235	Adults n = 130 (55%)	Children n = 105 (46%)	p
ISS, median (IQR)	5 (1,6)	5 (1,9)	4 (1,5)	0.07
LOS, mean ± SD	1.4 ± 1.8	1.9 ± 2.9	2.3 ±3.4	0.10
Disposition from emergency department, n (%)
Discharged	99 (42)	32 (32)	45 (45)	0.09
Floor	94 (40)	58 (25)	36 (13)
ICU	23 (10)	14 (61)	9 (4)
OR	9 (4)	5 (2)	4 (2)
Morgue	5 (2)	2 (1)	3 (1)
Against medical advice	5 (2)	4 (2)	1 (<1%)

ICU, intensive care unit.

Injury Details

A total of 1,009 injuries were recorded for the 235 patients (575 in adults, 434 in children). In children, the most common injuries were extremity soft tissue (225, 51.8%) followed by facial soft tissue (66, 15.2%), extremity fractures (63, 14.5%), intracranial injuries (33, 7.6%), and chest injuries (17, 3.9%) (Supplemental Digital Content, Supplementary Fig. 3, http://links.lww.com/TA/D617). There were no significant differences in injury locations in pediatric patients compared with adult (p > 0.05). The median ISS was 5 overall (Table 1).

The mean ± SD length of stay was 1.4 ± 1.8 days (children, 2.3 ± 3.4 days; adults, 1.9 ± 2.9 days). Most patients were admitted to the floor (93, 39.5%) followed by directly discharged from the emergency department (76, 32%). Twenty patients (8.5%) were admitted to the intensive care unit, and nine (3.8%) were taken to the operating room after the initial trauma evaluation. The remaining five patients (2.1%) were declared deceased after resuscitative attempts were made in the trauma bay. Three patients were children (case fatality rate, 2.8%).

Crash Characteristics

Of the unique 165 accidents identified, 60 had reported causes. Hard cornering (34, 56%) was the most common accident cause, followed by impact with an automobile (15, 25%) and impact with a stationary object (8, 13%) (Table 2). Approximately half of patients were ejected from golf carts during crashes (Table 2). Alcohol was involved in nine cases (3.8%), of which five were pediatric (56%). A total of 198 patients (84%), 114 adults (93%) and 84 children (88%), were not wearing seatbelts.

TABLE 2.

Accident Details for Both Children and Adults Involved in Golf Cart Accidents During the Study Period

Crash Details
Position	Overall N = 235	Adults n = 130 (55)	Children n = 105 (46)	p
Driver	52 (22)	43 (33)	9 (8)	0.15
Front-facing passenger	93 (40)	38 (29)	55 (52)	0.08
Rear-facing passenger	20 (8)	15 (11)	5 (4)	1.00
Seatbelt status	Overall N = 218	Adults n = 122 (55)	Children n = 96 (46)	p
Restrained	20 (9)	8 (6	12 (12)	0.50
Unrestrained	198 (90.8)	114 (93)	84 (87.5)
Ejection status	Overall N = 230	Adults n = 130 (55)	Children n = 105 (46)	p
Ejected	131 (56.9)	61 (47)	70 (53)	0.42
Not ejected	105 (45.7)	54 (41)	51 (48)

Risk Factors for Increased Injury Severity

When looking at children only, ejection had a slightly stronger association with increased ISS (odds ratio [OR], 1.114; confidence interval [CI], 1.018–1.258; p = 0.014). This association was not present when assessing only adults (OR, 1.03; CI, 0.966–1.106; p = 0.363).

Rollover crashes were not associated with more severe injuries for children (R² = 0.04, F_{1, 98} = 4.25; p = 0.40) and adults (R² = 0.06, F_{1, 116} = 7.46, p = 0.070). Rollover was more likely when there was hard-cornering (OR, 1.934; CI, 1.021–3.630; p = 0.041) and impacts with stationary objects (OR, 2.693; CI, 1.003–7.220; p = 0.049).

In comparison, ejection was associated with more severe injuries in children (OR, 1.114 for higher ISS; CI, 1.018–1.258; p = 0.014). This association was not present when assessing adults (OR, 1.03; CI, 0.966–1.106; p = 0.363). Thus, we explored risk and protective factors that specifically led to ejection.

Risk and Protective Factors for Ejection

In children, seatbelt use (OR, 0.149; CI, 0.031–0.647; p = 0.01) and traveling less than 10 miles per hour (mph) (OR, 0.046; CI, 0.011–0.217; p = 0.0002) were protective for ejection. Rear-facing passengers were at higher risk for ejection than front-facing passengers (OR, 7.548; CI, 1.956–33.77; p = 0.002) and drivers (OR, 6.800; CI, 1.665–30.81; p = 0.013). Hard cornering (OR, 4.286; CI, 1.094–13.36; p = 0.022) and rollovers (OR, 3.633; CI, 2.061–6.410; p < 0.0001) increased odds of ejection (Table 3).

TABLE 3.

Risk and Protective Factors for Ejection in Children Involved in Golf Cart–Related Trauma

Risk and Protective Factors for Ejection
Crash Type	OR (CI)	p
Hard cornering	4.29 (1.09–13.36)	0.02
Impact with automobile	0.12 (0.024–0.66)	0.01
Impact with stationary object	0.84 (0.21–3.38)	0.81
Rollover	3.63 (2.06–6.41)	<0.0001
Speed
>10 mph	6.68 (1.55–31.77)	0.01
<10 mph	0.15 (0.03 to 0.65)	0.01
Position
Driver	0.51 (0.17–1.83)	0.28
Front-facing passenger	1.00 (0.11–3.62)	0.69
Rear-facing passenger	7.548 (1.96–33.77)	0.002
Seatbelt use
Unrestrained	6.686 (1.55–31.77)	0.01
Restrained	0.1496 (0.03–0.65)	0.01

Geospatial Analysis

Most crashes occurred on the eastern end of the city between 3rd to 61st streets. Crashes tended to occur within the residential neighborhoods and along a high traffic street (Fig. 2). Two clusters of crashes were identified near the western boundary street. Clusters consisted of at least five crashes occurring within one block. Six of the 10 crashes between each cluster were caused by “bad driving” (taking a corner too fast, rolling the vehicle, or ejecting a rider) and collisions (impact with pedestrian, car, another golf cart or all-terrain vehicle, or a stationary object).

Figure 2.

All accidents occurring on Galveston Island during the study period. Each red dot represents one unique accident. Accident clusters encircled in blue.

A total of 77% patients evaluated were out of county visitors. Golf cart rental shops are the primary point of contact to transmit safety regulations to these patients. While renters are given a briefing on prohibited streets, no physical maps are given. Signs are posted indicating boundaries for golf cart operation; however, they are small and placed at the boundary roads rather than leading up to the boundary. At the intersection that is the western boundary of golf cart operation, we found two clusters of crashes. This is a high-traffic area, and thus, golf cart drivers may be forced to drive and turn faster to leave the intersection before the prohibited golf cart zone. In addition, the inadequate signage may not give golf cart drivers enough warning to turn around before entering the prohibited zone.

Survey of Local Rental Shops

Of the four rental shops on the island, all provided a safety briefing upon golf cart rental (4 of 4). Seatbelt use during operation was mentioned during all safety briefings (4 of 4), and all mentioned which streets were prohibited to golf cart operation (4 of 4). Only one golf cart rental shop had a printed map of prohibited streets (1 of 4). No rental shops offered printed maps of prohibited streets.

Upon inspection of the golf carts available for rental, most golf carts met the minimum required safety equipment. The most commonly missing equipment were side reflectors and side rear view mirrors.

DISCUSSION

As of 2022, the annual number of golf cart related injuries evaluated at our institution has increased by more than 350% since 2012 (Supplemental Digital Content, Supplementary Fig. 1, http://links.lww.com/TA/D617) and has been dramatically increasing annually since 2020. Similar trends have been reported in other golf cart trauma related studies.^2,4,5 Of note, children accounted for 46% of cases and an increasing proportion of children were injured over the last 4 years. The majority (88%) of children in our study were unrestrained, and they were at increased odds of ejection. Furthermore, ejection was positively associated with increasing ISS. All five fatalities during the study period, including three children, were unrestrained and ejected.

According to the National Highway Traffic Safety Administration, the federal organization that sets regulations for minimum safety features on motorized vehicles, golf carts are considered “low-speed vehicles.”^18,19 Low-speed vehicles are four-wheeled vehicles that can maintain a speed of 20 mph but not exceed 25 mph.^18,19 However, passengers have been ejected from golf carts at low speeds of 11 mph.^20–22 Ejection is associated with higher risk of severe injury and mortality.^2,10–16 In multiple studies, the use of restraints has been the most significant factor in reducing risk of severe injury, disability, and mortality.^2,10–16 Texas and Galveston city law mandates seatbelts be installed in all golf carts operated on public roadways and worn by all occupants when the vehicle is moving.^9,19 Children less than 4 ft 9 in must be in a car seat or booster seat unless they are older than 8 years.^9,19 Violators can face up to a $200 fine. Despite local laws and safety briefings at rental shops, only 20 patients (8 adults, 12 children) in our 10-year study were wearing seatbelts. Studies on golf cart injury by Starnes et al.⁸ and Passaro et al.² have identified sparse seatbelt use in other communities in the United States.

Our data showed that hard cornering (making sharp turns) increased the odds of ejection for children, which is in line with previously published data. Biomechanical simulations of golf cart maneuvers have demonstrated that moderate turns at speeds of 13 mph can still result in ejection of restrained children.^22,23 The peak acceleration force during these turns occurs at 0.5 seconds, limiting the time children have to react before being ejected.^22,23 We further found that children involved in crashes occurring at speeds of <10 mph were at lower odds of ejection.

Compounding the issue is the ineffective design of restraint systems. Two-point hip restraints that secure over the waist are commonly found in golf carts. These restraints were designed around the frame of an adult sized male. However, past studies have shown that this restraint style is ineffective and may increase the injury severity of an ejection by acting as a “tripping point,” launching the occupant out of the vehicle headfirst.^23,24 Children are at higher risk of being ejected headfirst with these restraints given their smaller frames and larger head and truncal proportions.^24,25

Decades of epidemiologic research have shown that proper use of car seats and booster seats decreases mortality and injury severity in children involved in automobile crashes.^26,27 However, traditional car seats and booster seats were designed and tested for use in automobiles.^27,28 Golf carts and other recreational vehicles lack the regulated attachment systems and safety features of these vehicles.²⁸ The Juvenile Products Manufacturers Association is a national trade organization that certifies products built with safety in mind. As of 2020, Juvenile Products Manufacturers Association has made a statement against the use of traditional car seats and boosters in golf carts because of the lack of testing in recreational vehicles.²⁸ Other after-market shoulder restraints systems have been designed for children; however, these systems also lack testing. To date, there are no published studies examining the effectiveness of these devices in golf cart crashes. This remains an area of study that may benefit the pediatric population.

Furthermore, we found that patients in the passenger position were more likely to be ejected than drivers. Rear-facing passengers were more likely to be ejected compared with all other occupants. One study found that current 2-point hip restraints are ineffective in preventing rear-facing passengers from being ejected.²⁰ This applied to turns and rapid forward acceleration. The forces in normal forward acceleration push front-facing occupants into their seatbacks. However, rear-facing passengers are tipped forward based on their position, increasing their risk of ejection. Three-point restraints that secure over the chest and waist or shoulder restraints have demonstrated better effectiveness in preventing ejection across crash scenarios for children.⁴ Educating parents on the benefits of this restraint configuration may improve pediatric safety by influencing choice of golf carts for operation or after-market restraint systems. Furthermore, national and local regulations regarding golf cart restraints may be better informed by these findings.

The laws governing golf cart operation vary by city and state, which may cause confusion for drivers.^3–5 A majority of patients in our study were out of county residents who rely on local rental shops and public signage for directions on safety and local laws. We found that, while safety briefings were provided to renters, printed materials for later reference were lacking. Adding to the issue are the scarcity and low visibility of signage regarding golf cart operation in our community. Signs designating prohibited streets are difficult to read from a distance (Fig. 3A), and no advanced warning is provided to allow golf carts to safely turn around. Our study has highlighted potential areas that can improve signage around our community and directions given at rental shops.

Figure 3.

(A) Example of a golf cart sign used to indicate a prohibited street in our community. (B) Educational flyer created for community education and awareness initiative.

Injury severity over the last 10 years has been low with a mean ISS of 5.3 (children, 4.9; adults, 5.9). Most injuries involved the soft tissue of the extremities and face followed by fractures of the extremities and face. While most injuries were of lower severity, there is an increasing number of golf cart–related traumas at our institution annually. This may translate to increased resource utilization and financial burden over time. A similar study performed at a level 1 trauma facility in Tennessee estimated $1 million in hospital charges for 40 golf cart–related traumas treated over 8 years.¹⁰ The increasing trend of golf cart injuries nationally necessitates improving awareness of risks and creation of community-based strategies.

There are several limitations to this study. First, using EMR data limited the size of the initial study population. Combining EMR data with EMS department crash data expanded our sample size, but these data points lacked granularity in crash details. In addition, these data sources introduce sampling bias to those who sought medical care. Second, variables of interest were not present for all individuals, which reduced the power and generalizability of some analyses. A degree of reporting bias may have occurred with crash causes, speed reporting, and restraint status, as this information was obtained from chart narratives. Finally, local laws may limit the generalizability of our results. Of note, one of the busiest streets in our community had a speed limit decrease in 2017 to 35 mph. Golf carts may now operate along this road, which translated to an increase in the golf-cart trauma crashes observed at our institution per year. It is difficult to say if the increase in crashes is a product of golf-cart popularity or the recent law change.

Based on the findings of this study, we recommend the following to improve golf-cart safety and reduce crashes and injury severity in our community:

1. Raising awareness of the risks through posters at high-tourist area locations and rental shops (Fig. 3B);

2. Partnering with local police and rental shops to enforce the use of seatbelts in golf carts;

3. Prohibiting evening golf cart use;

4. Improving the number of signs and sign visibility indicating prohibited streets;

5. Lowering the speed limit for golf carts at intersections;

6. Creation of protected lanes for golf carts in high traffic areas.

CONCLUSION

Our study found that children comprised nearly half of the total golf cart–related injuries in the past decade. We identified several risk and protective factors leading to injury and identified opportunities to improve safety at the community level. Our next steps are to increase public awareness and partnering with city leadership to advocate for improved signs, protected lanes, and policy changes to improve golf cart safety in our community.