a b s t r a c t

Background: Horse-related injuries can cause severe morbidity and mortality. The objective of this study is to in- vestigate the epidemiological features of horse-related injuries treated in emergency departments (EDs) in the United States.

Methods: A retrospective analysis of horse-related injuries from 1990 through 2017 was conducted utilizing the National Electronic Injury surveillance System . Cases were identified using the NEISS code 1239 (Horse- back Riding: Activity, Apparel, or Equipment). Analyses performed included calculation of national injury esti- mates, relative risks (RRs) with 95% confidence intervals (CIs), and linear regression.

Results: From 1990 through 2017, an estimated 1,836,536 (95% CI: 1,494,788-2,178,284) individuals presented to United States EDs with horse-related injuries. The annual injury rate decreased by 30.8% from 1990 to 1996, in- creased by 33.0% from 1996 to 2000, and then decreased by 46.0% from 2000 to 2017. Among older adults N60 years of age, the annual injury rate increased by 139.6% during the study period. The number of concussion and closed head injury diagnoses increased by 337.2% from 1990 to 2009. The most common mounted mecha- nism of injury was a fall or being thrown (73.9%), while the most common unmounted mechanism was being kicked (42.1%). Mounted injuries were more likely to lead to hospitalization than unmounted injuries (RR, 2.10, 95% CI: 1.59-2.77).

Conclusions: Despite the recent decline in horse-related injuries, these injuries are still common. Clinicians should be aware of the spectrum of ED presentations of horse-related injuries. Prevention efforts focused on older adults and concussions and Closed head injuries warrant special attention.

(C) 2019

Introduction

Horses are utilized for a diverse array of work and recreational activ- ities. In the United States (US), there are an estimated 9.2 million horses [1], and there are millions of people who ride or care for horses each year [2,3]. The interactions between humans and horses are multiface- ted and complex [4,5]. Horses can become unpredictable and poten- tially dangerous animals to riders, handlers, or spectators. The average

Abbreviations: CHI, closed head injury; CI, confidence interval; CPSC, United States Consumer Product Safety Commission; ED, emergency department; NEISS, National Electronic Injury Surveillance System; RR, relative risk; TBI, traumatic brain injury; US, United States.

* Corresponding author at: Center for Injury Research and Policy, The Research Institute at Nationwide Children’s Hospital, 700 Children’s Drive, Columbus, OH 43205, United States of America.

E-mail address: [email protected] (G.A. Smith).

horse weighs 1500 pounds and can reach speeds of N30 miles per hour [6,7]. Therefore, they possess the ability to directly or indirectly cause high impact or high force injuries secondary to ejections, falls, trampling, kicks, or collisions [3,7-11].

Though fewer individuals are involved in recreational activities in- volving horses compared with sports such as football or soccer, the pro- portion of injuries among participants is relatively high; one survey noted nearly 28% of participants who rode at least six times in a year had been treated by a medical provider for horse-related injuries within the past two years [12]. Deaths from horses comprise a large proportion of non-venomous, animal-related fatalities in the US [13], and death is a recognized risk of participating in equestrian activities [14]. This poten- tial for morbidity or mortality from horse-related activities has long driven injury prevention efforts in the equestrian field [15,16].

The epidemiology of horse-related injuries has been difficult to

comprehensively characterize. The majority of the literature has

https://doi.org/10.1016/j.ajem.2019.158366

0735-6757/(C) 2019

Fig. 1. Annual number and rate of nonfatal horse-related injuries treated in United States emergency departments by sex and by year, NEISS 1990-2017.

been comprised of case series or the experiences of single institu- tions [11,17-21]. Several regional or national studies have been per- formed, but are international [22-25], limited to the pediatric population [26,27], or are limited in temporal scope [3]. Given the prior documented changes in the type and nature of horse-related injuries over time [28-30], an updated understanding of injury

patterns is important in developing timely, evidence-based, injury prevention recommendations. Horse-related injuries can be partic- ularly severe and often require initial triage and treatment in emer- gency departments (EDs). The objective of this research is to investigate the epidemiology of horse-related injuries treated in US hospital EDs from 1990 through 2017.

Fig. 2. Annual rate of nonfatal horse-related injuries by age group and by year, NEISS 1990-2017.

Fig. 3. Annual number of nonfatal horse-related injuries treated in United States emergency departments by diagnosis and by year, NEISS 1990-2017.

Methods

Data source

The National Electronic Injury Surveillance System , which is operated by the US Consumer Product Safety Commission (CPSC), mon- itors product-related and sports- and recreation-related injuries pre- senting to US EDs [31]. The NEISS represents a stratified probability sample of the N5300 hospitals with a 24-hour ED with at least 6 beds in the US and its territories [32,33]. Approximately 100 hospitals, in- cluding 8 children’s hospitals, contribute to the NEISS. Professional coders abstract data from ED medical records, including demographic characteristics, the product or activity involved, diagnosis, affected body region, disposition from the ED, and a brief narrative of the injury incident [31].

Case selection

Horse-related injuries occurring between January 1, 1990 and De- cember 31, 2017 were identified using the NEISS code 1239 (Horseback Riding: Activity, Apparel, or Equipment). Each case narrative was exam- ined, and cases were excluded if: a) a fatality was documented (because the NEISS does not adequately capture fatalities), b) the injured individ- ual was an occupant of an automobile, or c) the injury involved horse- related equipment or environments (barn, stable, or pasture) without the presence of a horse. Cases involving donkeys and mules were also excluded. A case was defined as an individual injured while engaged in horse (or pony)-related activities, such as riding, mounting or dismounting, caring for a horse while not mounted, use of equestrian equipment with a horse present, or riding in a horse drawn vehicle. There were 41,523 actual cases identified for analysis.

Study variables

The NEISS variable for age was divided into four age groups:

b5 years, b) 5-18 years, c) 19-60 years, and d) N60 years. The location of the injury incident was categorized as a) home (including NEISS cat- egories of home, apartment/condo, and manufactured/mobile home),

farm, c) public place of recreation (including street or highway, place of recreation or sport, and other public place), and d) other

(including school and industrial site). Injured body region was grouped as: a) head and neck (including NEISS categories of head, neck, face, eye-ball, mouth, and ear), b) upper extremity (including upper and lower arm, shoulder, elbow, wrist, hand, and finger), c) lower extremity (including upper and lower leg, knee, ankle, foot, and toe), d) trunk (in- cluding upper and lower trunk, and pubic region), and e) other (includ- ing injury to N25% of the body). injury diagnosis was categorized as:

laceration (including NEISS categories of laceration, amputation, and puncture), b) strain/sprain, c) fracture (including fracture and dis- location), d) concussion/closed head injury (CHI) (including internal Organ injuries of head, and hemorrhage of the head), e) soft tissue injury (including contusion/abrasion, and hematoma), and e) other (including nerve damage, hemorrhage, Electric shock, poisoning, internal organ in- jury to non-head body regions, and other). Disposition from the ED was grouped into: a) treated and released, b) admitted (including NEISS cat- egories of treated and transferred, treated and admitted, and held for b24 h for observation), and c) left against medical advice.

NEISS case narratives were individually reviewed to identify factors related to the mechanism of injury. Injuries were categorized as occur- ring while: a) mounted, b) not mounted, c) mounting/dismounting, or

d) other/unspecified. The primary mechanism of injury was categorized as: a) fell or was thrown, b) bucked/reared/spooked, c) while jumping,

d) horse tripped/fell, e) struck by or against (excluding kicked by a horse), f) kicked (including kicked by another horse while mounted),

g) stepped on/trampled/crushed, h) body part caught, i) pushed into/ pulled/jerked/knocked down, and j) other/unknown. In addition to the primary mechanism of injury, secondary circumstances related to mounted injuries were categorized as: a) no other circumstances re- ported, b) struck by or against, c) stepped on, trampled, or kicked by horse, d) horse fell/rolled, e) person fell or was thrown, f) was dragged, and g) other circumstances.

Statistical analysis and ethical considerations

NEISS data were analyzed using SPSS 25.0 for Windows (IBM Corpo- ration, Chicago, IL) and SAS Enterprise Guide 7.15 (SAS Institutes Inc., Cary, NC). National injury estimates were calculated by utilizing Sample weights provided by the CPSC [32,33]. An estimate is considered unsta- ble if the estimate is b1200 cases, the sample size is b20 cases, or the co- efficient of variation is N33%. All data reported are stable national

Table 1

Characteristics of nonfatal horse-related injuries treated in United States emergency departments by mount status, NEISS 1990-2017.

Characteristics	Study totala			Mounted		Not mounted		Mounting/dismounting
	n (%b)	95% CI		n (%b)		n (%b)		n (%b)
Study total (row %c)	1,836,536 (100.0)	1,494,788-2,178,284		1,632,288 (88.9)		109,274 (6.0)		40,625 (2.2)
Location of incident
Public place of recreation	503,341 (45.4)	360,907-645,775		460,363 (47.0)		19,804 (27.9)		10,829 (44.2)
Home	469,171 (42.3)	354,302-584,041		396,165 (40.5)		43,721 (61.7)		10,168 (41.5)
Farm	133,554 (12.0)	78,308-188,800		119,647 (12.2)		7179 (10.1)		3503 (14.3)
Other	2795 (0.3)	1030-4560		2459 (0.3)		179 (0.3)d		15 (0.1)d
Subtotal	1,108,861 (100.0)	880,288-1,337,435		978,635 (100.0)		70,882 (100.0)		24,515 (100.0)
Body region injured
Upper extremity	542,657 (29.6)	456,718-628,596		481,105 (29.6)		30,121 (27.6)		7196 (17.7)
Trunk	532,522 (29.1)	424,267-640,777		502,376 (30.9)		15,645 (14.3)		7411 (18.2)
Head and neck	414,546 (22.6)	308,040-521,052		381,869 (23.5)		19,776 (18.1)		4156 (10.2)
Lower extremity	323,176 (17.7)	271,811-374,542		244,419 (15.0)		43,352 (39.7)		21,710 (53.5)
Other	17,357 (0.9)	12,171-22,544		16,345 (1.0)		309 (0.3)d		136 (0.3)d
Subtotal	1,830,259 (100.0)	1,489,268-2,171,249		1,626,114 (100.0)		109,202 (100.0)		40,609 (100.0)
Diagnosis
Fracture	569,571 (31.1)	459,321-679,821		520,561 (31.9)		22,734 (20.8)		11,294 (27.8)
Soft tissue	520,569 (28.4)	438,910-602,228		446,075 (27.4)		48,354 (44.3)		9592 (23.6)
Sprain or strain	294,818 (16.1)	244,490-345,145		262,325 (16.1)		10,345 (9.5)		12,712 (31.3)
Concussion/closed head injury	198,390 (10.8)	127,552-269,228		190,576 (11.7)		4248 (3.9)		1938 (4.8)
Laceration	116,694 (6.4)	96,356-137,031		89,824 (5.5)		16,936 (15.5)		1872 (4.6)
Other	133,423 (7.3)	99,959-166,886		120,103 (7.4)		6477 (5.9)		3217 (7.9)
Subtotal	1,833,464 (100.0)	1,491,981-2,174,947		1,629,464 (100.0)		109,094 (100.0)		40,625 (100.0)
Disposition
Treated and released	1,602,277 (87.3)	1,329,239-1,875,316		1,412,971 (86.6)		102,194 (93.7)		37,121 (91.4)
Admitted	226,359 (12.3)	146,373-306,345		211,958 (13.0)		6758 (6.2)		3445 (8.5)
Left against medical advice	6215 (0.3)	3376-9055		5906 (0.4)		91 (0.1)d		60 (0.1)d
Subtotal	1,834,852 (100.0)	1,493,841-2,175,863		1,630,835 (100.0)		109,043 (100.0)		40,625 (100.0)
Primary mechanism of injury
Fell or was thrown	1,222,750 (66.6)	995,757-1,449,743		1,206,827 (73.9)		1640 (1.5)		12,807 (31.5)
Bucked/reared/spooked	215,157 (11.7)	140,331-289,984		205,737 (12.6)		3434 (3.1)		2826 (7.0)
Struck by or against	74,454 (4.1)	61,717-87,192		55,131 (3.4)		7219 (6.6)		1852 (4.6)
Kicked	65,543 (3.6)	51,697-79,388		11,186 (0.7)		46,045 (42.1)		3911 (9.6)
Horse tripped/fell	63,660 (3.5)	50,358-76,963		51,263 (3.1)		821 (0.8)d		485 (1.2)d
Stepped on/trampled/crushed	36,181 (2.0)	27,825-44,536				29,613 (27.1)		3316 (8.2)
Body part caught	35,888 (2.0)	29,340-42,435		15,640 (1.0)		4684 (4.3)		2984 (7.3)
Pushed into/pulled/jerked/knocked down	23,828 (1.3)	18,812-28,845		7714 (0.5)		9931 (9.1)		1123 (2.8)d
While jumping	14,880 (0.8)	9450-20,309		14,724 (0.9)				156 (0.4)d
Other/unknown	84,195 (4.6)	63,067-105,323		64,066 (3.9)		5888 (5.4)		11,164 (27.5)
Subtotal	1,836,536 (100.0)	1,494,788-2,178,284		1,632,288 (100.0)		109,274 (100.0)		40,625 (100.0)
a Included “other/unknown” mount status.

^b Column percentages were calculated using each category’s total and may not sum to 100.0% due to rounding error.

^c Row percentages may not sum to 100.0% due to rounding error.

^d Estimate is potentially unstable due to sample size b 20 cases, estimate b 1200, or coefficient of variation N 33%.

estimates unless otherwise noted. US Census Bureau July 1 intercensal and postcensal estimates from 1990 to 2017 were used to calculate in- jury rates [34]. Statistical analyses included linear regression, Rao’s- Scott ?² tests, and relative risks (RRs) with 95% confidence intervals (CIs). Linear regressions were performed piecewise to analyze trends using breakpoints determined from Figs. 1-3. The entire study period was used in linear regressions of age. The estimated annual rate of change from the regression model, denoted by “m,” was reported along with the p value of the associated t-test used to assess statistical significance. The level of significance for all analyses was ? = 0.05. This study was judged as exempt by the institutional review board of the authors’ institution.

Results

General characteristics

From 1990 through 2017, an estimated 1,836,536 (95% CI: 1,494,788-2,178,284) individuals presented to US EDs due to horse- related injuries, averaging 65,591 injuries annually or 22.6 injuries per 100,000 US residents. Although most injuries occurred while mounted (88.9%), 6.0% of injuries occurred while not mounted, and 2.2% occurred

during mounting or dismounting (Table 1). Females accounted for 63.7% of horse-related injuries and had a higher injury rate than males (28.2 vs 16.7 injuries per 100,000 population). The median age of those injured was 29.8 years (Inter-quartile range: 15.5-44.1 years) with most of the injuries occurring among patients ages 19-60 (63.2%) and 5-18 (29.6%). The injury rate per 100,000 population was

highest among 5-18 year-olds (34.5), followed by 19-60 year-olds (25.0), N60 year-olds (7.6), and b5 year-olds (5.1).

Body region injured and diagnosis

The upper extremity (29.6%) and trunk (29.1%) were the most com- mon body regions injured, followed by the head and neck (22.6%) (Table 1). The majority of horse-related injuries were diagnosed as a fracture (31.1%) or soft tissue injury (28.4%), and 10.8% were diagnosed as a concussion/CHI. Females patients were more likely to sustain a head and neck injury (RR: 1.27; 95% CI: 1.20-1.35) or be diagnosed with a concussion/CHI (RR: 1.54; 95% CI: 1.33-1.79) than male patients. Simi- larly, patients b 5 years of age were more likely to have a head and neck injury (RR: 2.17; 95% CI: 1.96-2.41) or be diagnosed with a concussion/ CHI (RR: 1.44; 95% CI: 1.16-1.77) than older patients. Patients N 60 years of age were more likely to sustain a trunk injury (RR: 1.43; 95% CI:

Table 2

Selected primary mechanisms and circumstances associated with nonfatal horse-related injuries that occurred while mounted, NEISS 1990-2017.

Selected mechanisms/circumstances n (%^a) 95% CI Fell or was thrown

injured from being kicked (42.1%) or stepped on/trampled/crushed (27.1%). Female patients (RR: 2.90; 95% CI: 2.10-4.00) were more likely to be injured while jumping than males, and male patients (RR: 1.69; 95% CI: 1.48-1.93) were more likely to be injured from a horse tripping and falling than females. Patients b 5 years of age (RR: 2.15; 95% CI:

And no other circumstances reported 1,079,662

(89.5)

886,682-1,272,641

1.46-3.14) were more likely to be injured from being kicked than

older patients, while patients N60 years of age (RR: 1.50; 95% CI:

And struck by or against 51,346 (4.3) 38,137-64,555 And was stepped on, trampled, or kicked by 39,856 (3.3) 29,852-49,860

horse

And horse fell/rolled	19,357 (1.6)	10,272-28,442	Among individuals injured while mounted, secondary circum-
And was dragged	13,853 (1.1)	9596-18,111	stances associated with the primary injury mechanisms are provided

1.29-1.76) were more likely to be injured when the horse bucked/ reared/spooked than younger patients.

And person fell or was thrown 1623 (0.1) 848-2398

And other circumstances 1130 (0.1)^b 257-2002

in Table 2. A person falling or being thrown was the most common sec- ondary circumstance that occurred after a horse bucked, reared, or

Subtotal 1,206,827

(100.0)

Bucked/reared/spooked

983,652-1,430,001

spooked (80.3%) or while jumping (67.8%).

3.4. Disposition

And person fell or was thrown 165,293 (80.3) 103,494-227,091

And struck by or against 14,728 (7.2) 10,006-19,450

And horse fell/rolled 9923 (4.8) 6262-13,584 And was stepped on, trampled, or kicked by 9291 (4.5) 5582-13,000

horse

And was dragged 3843 (1.9) 2553-5133

And no other circumstances reported 2309 (1.1) 1315-3302

And other circumstances 352 (0.2)^b 44-659

Most (87.3%) individuals with horse-related injuries were treated and released from the ED and 12.3% were admitted (Table 1). Patients injured while mounted were 2.10 (95% CI: 1.59-2.77) times more likely to be admitted than patients injured while not mounted. Patients N 60 years of age (RR: 2.41; 95% CI: 2.01-2.90) were more likely to be

Subtotal 205,737

(100.0)

Horse tripped/fell

132,207-279,268

hospitalized than younger patients. Patients diagnosed with a fracture (RR: 3.22; 95% CI: 2.89-3.60) or a concussion/CHI (RR: 2.00; 95% CI:

1.79-2.23) were more likely to be hospitalized than patients diagnosed

And horse fell/rolled 32,690 (63.8) 25,078-40,302

And person fell or was thrown 13,783 (26.9) 10,315-17,252

And other circumstances 3721 (7.3) 2676-4766 And no other circumstances reported 1069 (2.1)^b 401-1738

Subtotal 51,263 (100.0) 39,979-62,548

Struck by or against

And no other circumstances reported 44,368 (80.5) 36,907-51,828

And person fell or was thrown 7844 (14.2) 4954-10,734

And other circumstances 2920 (5.3) 1944-3896

Subtotal 55,131 (100.0) 45,215-65,047

While jumping

And no other circumstances reported 659 (4.5)^b 282-1037

And person fell or was thrown 9978 (67.8) 6089-13,866

And other circumstances 4086 (27.8) 2336-5837

Subtotal 14,724 (100.0) 9298-20,149

^a Column percentages were calculated using each category’s total and may not sum to 100.0% due to rounding error.

^b Estimate is potentially unstable due to sample size b 20 cases, estimate b 1200, or coefficient of variation N 33%.

1.35-1.51) or a fracture (RR: 1.43, 95% CI: 1.33-1.55) than younger patients.

The relationship between body region injured and diagnosis with mounting status is provided in Table 1. Although the lower extremity (39.7%) was the most common body region injured while not mounted, trunk (30.9%) and upper extremity (29.6%) were the most common body regions injured while mounted. When comparing injuries occur- ring while mounted and not mounted, injuries occurring while mounted were 1.53 (95% CI: 1.40-1.68) times more likely to be a frac- ture than other diagnoses, and injuries occurring while not mounted were 1.62 (95% CI: 1.48-1.78) times more likely to be a soft tissue injury than other diagnoses.

Mechanism and circumstances of injury

Among all horse-related injuries, 66.6% resulted from falling or being thrown, and 11.7% resulted from the horse bucking/rearing/spooking (Table 1). The primary mechanisms of injury differed between patients mounted and not mounted. While falling or being thrown was the most common mechanism among those mounted (73.9%) or those mount- ing/dismounting (31.5%), the majority of those not mounted were

with other injuries. Patients injured from a horse bucking, rearing, or spooking were 1.55 (95% CI: 1.36-1.78) times more likely to be admit- ted than patients injured from other injury mechanisms.

3.5. Secular trends

The annual number of horse-related injuries decreased significantly by 25.3% (m = -2919, p = 0.002) from 69,311 injuries in 1990 to 51,755 injuries in 1996, increased significantly by 51.2% (m = 7034, p b 0.001) from 1996 to 78,230 injuries in 2000, and then decreased sig- nificantly by 37.6% (m = -1590, p b 0.001) from 2000 to 48,796 inju- ries in 2017 (Fig. 1). Similarly, the annual injury rate per 100,000 population decreased by 30.8% (m = -1.4, p b 0.001) from 27.8 in

1990 to 19.2 in 1996, increased by 33.0% (m = 2.2, p b 0.001) from

1996 to 27.7 in 2000, and then decreased by 46.0% (m = -0.7, p b

0.001) from 2000 to 15.0 in 2017 (Fig. 1). The annual injury rates among Males and females (Fig. 1) and among 5-18 year-olds and 19-60 year-olds (Fig. 2) followed similar trends to the overall annual in- jury rate trend (statistics not given). Among individuals N60 years old, the annual injury rate increased significantly by 139.6% (m = 0.3, p b 0.001) from 4.2 in 1990 to 10.1 in 2017 (Fig. 2).

The annual numbers of horse-related injury by diagnoses are given in Fig. 3. While the annual number of fractures, sprains or strains, and soft tissue injuries (Fig. 3) generally followed the overall number of in- juries trend (statistics not given), the annual number of lacerations de- creased significantly by 58.0% (m = -123, p b 0.001) during the 1990-2017 study period. The annual number of concussion/CHI diagno- ses increased significantly by 337.2% (m = 368, p b 0.001) from 1990 to 2009, before decreasing significantly by 32.7% (m = -217, p = 0.035)

from 2009 to 2017.

Discussion

The annual number of horse-related injuries fluctuated throughout the study period, decreasing from 1990 to 1996, then increasing from 1996 to 2000 before decreasing again from 2000 to 2017. The annual number of injuries roughly correlate with the annual number of horses registered in the United States [35], suggesting a possible relationship between injuries and the popularity of horse-related activity. Both horse registrations and estimates of the number of individuals partici- pating in horse-related activities have declined over the past decade

[35,36], which may help explain the decreasing numbers of injuries ob- served during the latter part of the study period. The period of increase in the number of horse-related injuries was driven, in part, by an in- crease in the diagnosis of concussions/CHIs.

Prevention of concussion/CHI remains a critical focus for equestrian activities. Severe horse-related injuries and fatalities are often attribut- able to head injuries [8,18,37-40]. The use of a helmet in horse-related activities is advocated by medical and equestrian organizations [41,42], based on evidence that helmets reduce both the number and se- verity of injuries [38,40,43,44]. However, prior studies have suggested that helmet use among riders is inconsistent [45]. Currently, the use of a helmet that is compliant with ASTM standard F1163 is recommended or required while riding [41,42,46], with evidence suggesting protective benefits of helmet use even while unmounted [40,47]. Emergency med- icine practitioners should be aware of the possibility of concussion or closed head injuries for patients presenting with a horse-related injury, because these injuries could require urgent or emergent attention.

From 2009 through 2017, a 33% decrease was observed in concus- sion/CHI diagnoses. Around 2009, several high-profile events occurred that may have contributed to this decline, including publicized TBI- related fatalities [14,48], signing of legislation and regulations requiring equestrian helmet use [42,48], and increased efforts to promote eques- trian TBI awareness [49]. Despite these public health gains, the number of horse-related TBIs remains high, underscoring the need to increase promotion of helmet use and other prevention efforts.

During the study period, the annual injury rate among adults N60 years of age increased by approximately 140%. One possible expla- nation for this trend is increased interest among older adults in horse- related activities. The increased injury rate also may reflect an aging of those participating in horse-related activities, with the cohort of expo- sure entering the next study age group over the period of the study. Be- cause riding and horse-related recreation statistics in the US are incomplete, the explanation for this trend is uncertain.

Most horse-related injury prevention efforts are focused on females, young adults, and equestrian competitors [12,24,44]. These injury pre- vention efforts are in line with evidence from this and prior studies that demonstrate that younger and female riders represent a large pro- portion of those injured during horse-related activities [3,27,29]. The large number of injured females likely reflects differences in participa- tion in equestrian activities between females and males. Given the in- creased injury rate among older adults, prevention efforts focusing on this age group merit attention [24]. Older adults were noted to be at in- creased risk of fractures, injuries to the trunk, and hospitalizations from horse-related injuries compared with younger patients. Older adults may approach riding and horse-related activities differently than youn- ger individuals, which warrants additional investigation. Furthermore, this age group may require special attention from emergency medicine providers when presenting with horse-related injuries due to the in- creased likelihood of co-morbid conditions and medication use, such as anticoagulants, which could complicate or increase the severity of the initial ED presentation.

The most commonly injured body region while mounted was the trunk, which is consistent with prior studies utilizing the NEISS after ac- counting for coding differences [3]. Trunk injuries are diverse, including fractures, pneumothoraxes, internal organ contusions, and crush inju- ries, and can be severe [7,20,22]. Trunk protection during horse- related activities is often minimal to non-existent, given the relatively large region to protect and the lack of evidence-based protective equip- ment. Recent injury prevention technologies, such as inflatable vests or air jackets have become increasingly popular among riders [16], but continue to require further study to confirm their effectiveness. Emer- gency medicine providers should maintain a high index of suspicion for internal organ injuries for patients presenting with horse-related in- juries. Appropriate diagnostic evaluation and consultation with surgical or trauma services may be warranted, depending on injury type and severity.

Mechanisms of horse-related injury are diverse and differ between individuals injured while mounted and unmounted (Tables 1 and 2); these observed mechanisms are similar to those previously described [3,27]. Given the diversity of injury mechanisms and the complexity of interactions between humans and horses, a multi-faceted approach to horse-related injury prevention is required. Injury prevention recom- mendations have been described previously [3,43], and include proper training for riders and horse caregivers, appropriate fitting of equip- ment and rider attire, and research on the use of additional safety equip- ment, such as quick-release stirrups. Psychosocial factors affecting human and horse behaviors is also the subject of research [15,50], which could inform future injury prevention efforts.

Study limitations

This study has several limitations. Individuals treated in non-ED set- tings, such as urgent care centers and physicians’ offices, and those who were injured but did not seek treatment, are not captured by the NEISS. Therefore, our study findings underestimate the true number of horse- related injuries in the US. In addition, the NEISS does not capture fatal- ities well, and therefore, only non-fatal injuries are reported in this study. Population-based injury rates were calculated using national cen- sus data because data on the number of participants (including by age and sex) in horse-related activities were unavailable. Occupants of auto- mobiles involved in horse-related crashes were not included in this study. NEISS case narratives are limited by the amount of detail included in ED medical records. Narratives may not consistently provide informa- tion about the mechanisms or circumstances of injury. Despite these limitations, the strength of this study is that it is based on a nationally Representative sample over a 28-year study period.

Conclusions

Despite the decline in horse-related injuries, the number of injuries remains high, emphasizing the need for increased prevention efforts. Clinicians should be aware of the spectrum of ED presentations of horse-related injuries; cognizance of the mechanism of injury can help inform the index of suspicion for injury type and severity, which can guide initial stabilization and treatment of the patient. Mechanisms of injury associated with horses are diverse and require a multi-faceted in- jury prevention approach. Although females and younger individuals account for the majority of horse-related injuries, injuries among older adults and concussions/CHIs warrant special attention.

Funding

This research did not receive any specific grant from funding agen- cies in the public, commercial, or not-for-profit sectors.

Financial disclosure statement

The authors have nothing to disclose relevant to this study.

Declaration of Competing Interest

The authors have no conflicts of interest relevant to this study.

References

1. Salem DJ, Rowan AN. The state of the animals IV: 2007. Humane Society of the United States. Washington, D.C.: Humane Society Press; 2007
2. Bixby-Hammett DM. Accidents in equestrian sports. Am Fam Physician 1987;36: 209-14.
3. Thomas KE, Annest JL, Gilchrist J, Bixby-Hammett DM. Non-fatal horse related inju- ries treated in emergency departments in the United States, 2001-2003. Br J Sports Med 2006;40:619-26. https://doi.org/10.1136/bjsm.2006.025858.
4. Hawson LA, McLean AN, McGreevy PD. The roles of equine ethology and applied learning theory in horse-related human injuries. J Vet Behav 2010;5:324-38. https://doi.org/10.1016/j.jveb.2010.06.001.
5. DeAraugo J, McLaren S, McManus P, McGreevy PD. Improving the understanding of psychological factors contributing to horse-related accident and injury: context, loss of focus, cognitive errors and rigidity. Animals 2016;6:12. https://doi.org/10.3390/ ani6020012.
6. Carrillo EH, Varnagy D, Bragg SM, Levy J, Riordan K. Traumatic injuries associated with horseback riding. Scand J Surg 2007;96:79-82. https://doi.org/10.1177/ 145749690709600115.
7. Havlik HS. Equestrian sport-related injuries: a review of the current literature. Curr Sports Med Rep 2010;9:299-302. https://doi.org/10.1249/JSR.0b013e3181f32056.
8. Bixby-Hammett D, Brooks WH. Common injuries in horseback riding: a review. Sports Med 1990;9:36-47. https://doi.org/10.2165/00007256-199009010-00004.
9. McCrory P, Turner M. Equestrian injuries. Med Sport Sci 2005;48:8-17. https://doi. org/10.1159/000084280.
10. Kruger L, Hohberg M, Lehmann W, Dresing K. Assessing the risk for Major injuries in equestrian sports. BMJ Open Sport Exerc Med 2018;4:e000408. https://doi.org/10. 1136/bmjsem-2018-000408.
11. Ball CG, Ball JE, Kirkpatrick AW, Mulloy RH. Equestrian injuries: incidence, injury patterns, and risk factors for 10 years of major traumatic injuries. Am J Surg 2007; 193:636-40. https://doi.org/10.1016/j.amjsurg.2007.01.016.
12. Christey GL, Nelson DE, Rivara FP, Smith SM, Condie C. Horseback riding injuries among children and young adults. J Fam Pract 1994;39:148-52.
13. Forrester JA, Weiser TG, Forrester JD. An update on fatalities due to venomous and nonvenomous animals in the United States (2008-2015). Wilderness Environ Med 2018;29:36-44. https://doi.org/10.1016/j.wem.2017.10.004.
14. Thomas K. Equestrians’ deaths spread unease in sport. The New York Times. Avail- able from: http://www.nytimes.com/2008/04/09/sports/othersports/09equestrian. html; 2008, Apr 9, Accessed date: 26 January 2019.
15. Thompson K, McGreevy P, McManus P. A critical review of horse-related risk: a re- search agenda for safer mounts, riders and equestrian cultures. Animals 2015;5: 561-75. https://doi.org/10.3390/ani5030372.
16. Thomas K. Added safety in the saddle. The New York Times. Available from: http:// www.nytimes.com/2010/08/24/sports/24airbag.html; 2010, Aug 23, Accessed date:

    26 January 2019.

    Abdulkarim A, Juhdi A, Coffey P, Edelson L. Equestrian injury presentations to a re- gional trauma centre in Ireland. Emerg Med Int 2018;2018:7394390https://doi. org/10.1155/2018/7394390.

17. Bixby-Hammett DM. Horse-related injuries and deaths in North Carolina, 1995- 1999. N C Med J 2006;67:161-2.
18. Clarke CN, Tsuei BJ, Butler KL. Equine-related injury: a retrospective analysis of out- comes over a 10-year period. Am J Surg 2008;195:702-4. https://doi.org/10.1016/j. amjsurg.2007.11.007.
19. Davidson SB, Blostein PA, Schrotenboer A, Sloffer CA, VandenBerg SL. Ten years of equine-related injuries: severity and implications for emergency physicians. J Emerg Med 2015;49:605-12. https://doi.org/10.1016/j.jemermed.2015.03.025.
20. Newton AM, Nielsen AM. A review of horse-related injuries in a rural Colorado hos- pital: implications for outreach education. J Emerg Nurs 2005;31:442-6https://doi. org/10.1016/j.jen.2005.08.009.
21. Cripps RA. Horse-related injury in Australia. Australian injury prevention bulletin 24, AIHW Cat. No. INJ26, May 2000. Research Centre for Injury Studies, Flinders Univer- sity; 2000. [Accessed 26 Jan 2019]. Available from: http://www.aihw.gov.au/reports/ injury/horse-related-injury-australia/contents/table-of-contents.
22. Sorli JM. Equestrian injuries: a five year review of hospital admissions in British Co- lumbia, Canada. Inj Prev 2000;6:59-61.
23. Smartt P, Chalmers D. A new look at horse-related sport and recreational injury in New Zealand. J Sci Med Sport 2009;12:376-82. https://doi.org/10.1016/j.jsams. 2008.04.001.
24. Lang J, Sathivelu M, Tetsworth K, Pollard C, Harvey K, Bellamy N. The epidemiology of horse-related injuries for different horse exposures, activities, and age groups in Queensland, Australia. J Trauma Acute Care Surg 2014;76:205-12. https://doi.org/ 10.1097/TA.0b013e3182a9007e.
25. Bixby-Hammett DM. Pediatric equestrian injuries. Pediatrics 1992;86:1173-6.
26. Ghosh A, Di Scala C, Drew C, Lessin M, Feins N. Horse-related injuries in pediatric pa- tients. J Pediatr Surg 2000;35:1766-70. https://doi.org/10.1053/jpsu.2000.19247.
27. Sandiford N, Buckle C, Alao U, Davidson J, Ritchie J. Injuries associated with recrea- tional horse riding and changes over the last 20 years: a review. JRSM Short Rep 2013;4:2042533313476688. https://doi.org/10.1177/2042533313476688.
28. Moss PS, Wan A, Whitlock MR. A changing pattern of injuries to horse riders. Emerg Med J 2002;19:412-4.
29. Chitnavis JP, Gibbons CL, Hirigoyen M, Lloyd Parry J, Simpson AH. Accidents with horses: what has changed in 20 years? Injury 1996;27:103-5.
30. U.S. Consumer Product Safety Commission. NEISS: The National Electronic Injury Surveillance System, a tool for researchers. Washington, D.C.; 2000. [Accessed 26 Jan 2019]. Available from: http://www.cpsc.gov/s3fs-public/pdfs/blk_media_ 2000d015.pdf.
31. Schroeder T, Ault K. The NEISS sample (design and implementation) 1997 to pres- ent. U.S. Consumer Product Safety Commission: Washington, D.C; 2001 Accessed 26 Jan 2019]. Available from: http://www.cpsc.gov/PageFiles/106617/2001d011- 6b6.pdf.
32. Schroeder T, Ault K. The NEISS sample (design and implementation) from 1979 to 1996. Washington, D.C.: U.S. Consumer Product Safety Commission; 2001. [Accessed 26 Jan 2019]. Available from: http://www.cpsc.gov/s3fs-public/2001d010-6b6.pdf.
33. U.S. Census Bureau. Annual estimates of the resident population by single year of age

    and sex for the United States: April 1, 2010 to July 1, 2017. Washington, D.C.; 2018. [Accessed 26 Jan 2019]. Available from: http://www.census.gov/data/tables/time- series/demo/popest/2010s-national-total.html.

    Kleine K. Current trends in the horse industry. American Association of Equine Prac- titioners. Available from: http://animalagriculture.org/Resources/Documents/Conf% 20-%20Symp/Conferences/2015%20Annual%20Conference/Speaker% 20Presentations/Equine/Equine-Kleine-stateoftheindustry.pdf; 2015, Accessed date:

    26 January 2019.

    Sport & Fitness Industry Association. Sports marketing surveys U.S.A. 2013 sports, fitness and leisure activities topline participation report. Available from http:// www.espn.com/pdf/2013/1113/espn_otl_sportsreport.pdf; 2013, Accessed date: 26

    January 2019.

    Nguyen HS, Lew S. Equestrian-related traumatic brain injury in the pediatric popu- lation. Pediatr Neurosurg 2016;51:279-83https://doi.org/10.1159/000446402.

34. Bier G, Bongers MN, Othman A, et al. Impact of helmet use in equestrian-related traumatic brain injury: a matched-pairs analysis. Br J Neurosurg 2018;32:37-43. https://doi.org/10.1080/02688697.2017.1409874.
35. Wolyncewicz GEL, Palmer CS, Jowett HE, Hutson JM, King SK, Teague WJ. Horse- related injuries in children – unmounted injuries are more severe: a retrospective review. Injury 2018;49:933-8. https://doi.org/10.1016/j.injury.2017.12.003.
36. Kriss TC, Kriss VM. Equine-related neurosurgical trauma: a prospective series of 30 patients. J Trauma 1997;43:97-9.
37. American Academy of Pediatrics Committee on Sports Medicine and Fitness. Horse- back riding and head injuries. Pediatrics 1992;89:512.
38. USEF Communications Department. USEF passes new safety helmet rules for eventing and dressage riders. Available from: http://www.usef.org/media/press- releases/6569_usef-passes-new-safety-helmet-rules-for-eventing-and-dressage- riders; 2011, Accessed date: 26 January 2019.
39. Watt GM, Finch CF. Preventing equestrian injuries. Locking the stable door. Sports Med 1996;22:187-97. https://doi.org/10.2165/00007256-199622030-00005.
40. Bond GR, Christoph RA, Rodgers BM. Pediatric equestrian injuries: assessing the im- pact of helmet use. Pediatrics 1995;95:487-9.
41. Nelson DE, Rivara FP, Condie C. Helmets and horseback riders. Am J Prev Med 1994; 10:15-9.
42. ASTM International. ASTM F1163-15: standard specification for protective headgear used in horse sports and horseback riding. West Conshohocken, PA; 2015. [Accessed 26 Jan 2019]. Available from: https://www.astm.org/Standards/F1163.htm.
43. Meredith L, Ekman R, Thomson R. Horse-related incidents and factors for predicting injuries to the head. BMJ Open Sport Exerc Med 2018;4:e000398. https://doi.org/10. 1136/bmjsem-2018-000398.
44. Allen CR. Gov. Crist signs horse rider helmet law. South Florida sun sentinel. Avail- able from: https://www.sun-sentinel.com/news/sfl-horse-helmet-law-p060809- story.html; 2009, Jun 8, Accessed date: 26 January 2019.
45. Riders4Helmets. Riders4Helmets international awareness day. Available from https://www.riders4helmets.com/about/; 2019, Accessed date: 26 January 2019.
46. Chapman M, Thompson K. Preventing and investigating horse-related human injury and fatality in work and non-work equestrian environments: a consideration of the workplace health and safety framework. Animals 2016;6:33. https://doi.org/10. 3390/ani6050033.