Trampoline-Related Injuries to Children

FOLLOWING THEIR introduction in 1936, trampolines were used during the 1940s by US and British fighter pilots as training devices. During the next 2 decades, trampolines gained popularity among the general population in the United States and Europe.^{1 - 5} The first reports of trampoline-related injuries were by Zimmerman⁶ in 1956 and Ellis et al⁷ in 1960. More recent studies have continued to report trampoline-related injuries, including quadriplegia and death^{1 - 5 ,8 - 20} ; however, relatively few studies have focused on these injuries in the pediatric population.^{1 ,8 - 12} This investigation describes the epidemiological features of trampoline-related injuries among children treated in the emergency department (ED) of a large, urban children's hospital. To our knowledge, this represents the largest published series of pediatric trampoline-related injuries.

Medical records were reviewed for all children treated for a trampoline-related injury in the ED of Children's Hospital in Columbus, Ohio, from May 1, 1995, through April 30, 1997, giving a total of 214 children. Children's Hospital is a 313-bed tertiary care facility affiliated with The Ohio State University College of Medicine. Approximately 70 000 children are treated in the hospital ED annually. This study was approved by the Human Subjects Research Committee.

Data were obtained from the ED medical record for all patients. Inpatient medical records were reviewed for the patients admitted to the hospital. Attempts were made to contact parents of all study patients by mail or telephone to obtain information not contained in patient medical records. Follow-up contact occurred approximately 2 months after the ED visit and was completed for 120 patients (56.1%). The type of injury was grouped into categories during study analyses. Soft tissue injuries included contusions, abrasions, hematomas, strains, and sprains.

The data obtained from patient medical records included patient demographics, date and time of injury and ED treatment, type and anatomical location of injury, procedures performed, and patient disposition. Information obtained directly from parents included the trampoline characteristics and surrounding environment, the circumstances of injury, the surface or object struck, the presence of supervision, the child's training and experience on trampolines, the child's past trampoline-related injuries, continued trampoline use after injury, and parental awareness of trampoline hazards before the injury event.

Data were analyzed using commercially available statistical software (Epi Info, Version 5, USD Inc, Stone Mountain, Ga, 1990). Statistical evaluation included χ² analysis with the Yates correction and the 2-tailed Fisher exact test. Comparisons were considered to be statistically significant at P<.05. Relative risk (RR) with a 95% confidence interval (CI) was also calculated.

There were 214 children treated for trampoline-related injuries from May 1, 1995, through April 30, 1997. This represents an average of 1 child treated every 3.4 days. Children ranged in age from 1 to 16 years (mean [SD], 9.4 [3.6] years). Forty-four children (20.6%) were younger than 6 years. Males (n = 113 [52.8%]) predominated slightly (Figure 1). Although 63% of all patients treated for injuries in this ED were white, 93% (n = 199) of patients with trampoline-related injuries were white. Three patients were treated twice during the 2-year period. The time between the 2 injuries for these 3 patients was 4, 5, and 13 months. The number of trampoline-associated injuries (n = 190) peaked from April through October, accounting for 89.2% of cases. All injuries occurred between 9 am and 1 am, with 86.2% (169/196) of injuries occurring between 2 and 10 pm.

Emergency department treatment was sought within 6 hours of the injury in 139 cases (65.0%) and between 24 and 48 hours following the injury in 55 cases (25.7%). Injuries associated with long delays in seeking care were not always minor; 2 of the 5 children treated between 5 and 14 days following injury had fractures. Thirty-three patients (15.4%) arrived by ambulance, and the remainder arrived by private vehicle. Twenty-three patients were seen by their primary care physician or in the ED of another hospital before being referred to our ED; 14 (60.9%) of these patients had fractures.

The area of the body most commonly injured was a lower extremity (77 [36.0%]), followed by an upper extremity (68 [31.8%]), the head (31 [14.5%]), the trunk (21 [9.8%]), and the neck (17 [7.9%]) (Figure 2). The most common type of injury was a soft tissue injury (111 [51.9%]), followed by fracture (74 [34.6%]), laceration (25 [11.7%]), and other (4 [1.8%]) (Figure 3). The 4 injuries in the "other" category included 2 closed-head injuries with concussion, 1 acromioclavicular separation, and 1 permanent tooth intrusion. The Pediatric Trauma Score ranged from 8 to 12 (median, 11). The Injury Severity Score ranged from 1 to 9 (median, 1). The Glasgow Coma Score was 15 for all patients, except for 1 child who had a Glasgow Coma Score of 14.

Fractures accounted for most injuries to upper extremities (61.8% [42/68]) compared with those to lower extremities (37.7% [29/77]) (P = .006; RR = 1.64; 95% CI, 1.16-2.31). Soft tissue injuries represented most (47 [61.0%]) lower extremity injuries compared with (25 [36.8%]) those to upper extremity injuries (P = .006; RR = 1.66; 95% CI, 1.16-2.38). Soft tissue injuries were also the most common type of injury (56.5% [96/170]) among children 6 years and older compared with children younger than 6 (34.1% [15/44]) (P = .01; RR = 1.66; 95% CI, 1.08-2.55). Nineteen injuries (43.2%) among children younger than 6 years were fractures compared with 55 (32.4%) for children 6 years and older; however, the association of age and fractures was not statistically significant (P = .24; RR = 1.33; 95% CI, 0.89-2.00). There also were no statistically significant associations between a child's age (< 6 years) and injury to the upper extremities (P = .86; RR = 0.91; 95% CI, 0.55-1.51) or lower extremities (P = .81; RR = 1.09; 95% CI, 0.72-1.67).

There was a strong association between lacerations and injury to the head, including the face and scalp (P<.001; RR = 67.9; 95% CI, 16.8-273.6). Most (92% [23/25]) lacerations were on the head, and 74.2% (23/31) of injuries to the head were lacerations. Lacerations accounted for 22.7% (10/44) of injuries among children younger than 6 years compared with 8.8% (15/170) of injuries to children 6 years and older (P = .02; RR = 2.58; 95% CI, 1.24-5.34). Among children younger than 6 years, 20.4% (9/44) of injuries were to the head, including face and scalp, compared with 12.9% (22/170) of injuries among children aged 6 years and older; however, the association of age younger than 6 years with injury to the head was not statistically significant (P = .31; RR = 1.58; 95% CI, 0.78-3.19).

Four patients (1.9%) were admitted to the hospital. All 4 patients had fractures, including 2 patients with a femur fracture, 1 patient with open fractures of the radius and ulna, and 1 patient with a displaced supracondylar fracture. The length of hospital stay ranged from 1 to 28 days (mean, 15 days). Three of these patients required orthopedic procedures in the operating room, including fracture reduction and pinning. Two of these patients went directly from the ED to the operating room. See Table 1 for a summary of procedures performed for all patients.

The trampoline was full-sized in 90.1% (128/142) of cases and was a minitrampoline in 9.9% (14/142); the type of trampoline was unknown in 72 cases. The surface underneath the trampoline was grass in 86% (104/121) of cases, dirt in 8.3% (10/121), concrete in 3.3% (4/121), or gym mats in 2.4% (3/121); the type of surface was unknown in 93 cases. Twenty-seven (22.5%) of 120 parents reported that the trampoline was set up near other objects that could cause injury if a person fell from the trampoline onto or against them. These objects included trees, poles, fences, playground equipment, and a pile of bricks. Table 2 shows the mechanism of injury for 188 patients. Fifteen percent (23/153) of children were injured when they fell on the padded or unpadded trampoline frame, and 5.1% (8/153)) of children were injured from falls onto the springs. Twenty-five (13.4%) of 187 children were injured when they attempted a flip.

There was more than 1 person on the trampoline at the time of injury in 76.2% (106/139) of cases. Among the 85 cases with the number of trampolinists known, there were 2 people on the trampoline in 49 cases (57.6%), 3 people in 23 cases (27%), 4 people in 10 cases (11.8%), and 5 people in 3 cases (3.5%). The patient weighed less than the others on the trampoline in 48 (56.5%) of these cases and was the heaviest person in 9 cases (10.6%). The trampoline was located in the backyard in 96% (119/124) of cases, including the patient's own backyard in 39.5% (49/124) of cases. Three injuries occurred in private gyms while a professional instructor was supervising. Adult supervision was present at the time of injury for 73.3% (22/30) of children younger than 6 years. Overall, 55.6% (65/117) of injuries occurred despite direct supervision by an adult at the time of injury. At least 1 spotter was present in 30.8% (33/107) of cases, including 1 spotter (10.3%), 2 (8.4%), 3 (5.6%), 4 (5.6%), or 6 spotters (0.9%). Twenty-two (18.6%) of 118 patients had received formal instruction on trampoline use before the injury, most of them in a gymnastics class. Sixty-six (56.9%) of 116 parents reported that their child had previously attempted a flip on a trampoline.

When the injury occurred, 46.4% (39/84) of children had been on the trampoline less than 15 minutes, and 33.3% (28/84) had been trampolining at least 30 minutes. Ten (8.6%) of 116 injured children had had no previous experience on a trampoline, but 75 (64.6%) of 116 children had used a trampoline on more than 10 occasions before the injury. Following recovery from injury, 54.8% (63/115) of children resumed use of a trampoline. For these children, new precautions reported by parents included only 1 person on the trampoline at a time, trampoline use only with adult supervision, and no flips allowed. Several parents indicated, however, that the new precautions were allowing only 2 people at a time on a trampoline, and children using the trampoline simultaneously should be of similar weight. Ten percent (12/120) of parents reported that this was not their child's first injury on a trampoline. Seventy-three percent (81/111) of parents indicated that they had been aware of the potential dangers of trampolines before the injury event.

The use of trampolines increased rapidly in the United States and Europe beginning in the 1950s,^{1 - 5} and along with this came more reports of trampoline-related injuries.^{1 - 20} These injury reports resulted in the removal of trampolining from national competition by the National Collegiate Athletic Association^{3 ,18} in 1971, the development of a voluntary trampoline safety standard²¹ by the American Society for Testing and Materials (ASTM) in 1974, and the publication of a fact sheet on trampoline safety²² by the Consumer Product Safety Commission (CPSC) in 1976.

In 1977, the American Academy of Pediatrics (AAP) issued its first statement²³ on trampoline safety, recommending that "trampolines be banned from use as part of the physical education programs in grammar schools, high schools, and colleges and also be abolished as a competitive sport." It indicated that from 1973 through 1975, trampolines were associated with permanent paralysis from an injury to the cervical spine more frequently than any other gymnastic sport and that trampolines were second only to football as a cause of permanent paralysis among all sports. These trampoline-related injuries had often occurred during supervised physical education activities.²³ Most schools in the United States removed trampolines from physical education programs following publication of this statement. By the late 1970s, the largest trampoline manufacturer in the United States, the Nissen Corporation, ceased production due to decreased sales and increased product liability litigation.³

The AAP's initial statement on trampolines was not without controversy. In 1978, the American Alliance for Health, Physical Education, and Recreation issued its own statement,²⁴ which said that "the use of the trampoline in physical education classes does not apparently constitute an unreasonable risk of serious injury providing that . . . controls are ensured." Eight controls and conditions for safe use of trampolines in physical education classes were outlined in this statement.

In 1981, the AAP revised its recommendations in a second statement,²⁵ indicating that it "does not endorse trampoline use, but a revision of the Academy's position to allow for a trial period of limited and controlled use by schools seems appropriate." Among 7 precautions outlined in this statement was the recommendation that "the trampoline should never be used in home or recreational settings."²⁵ The response to the revised statement was mixed, and some authors concluded that the AAP was ill-advised in altering its initial stricter safety recommendations.^{4 ,17 - 18} The AAP is planning to release a third statement on trampolines in 1998.

This study identified several statistically significant patterns of trampoline-related injury. Fractures were associated with injury to the upper extremities, soft tissue injuries were associated with injury to the lower extremities, and lacerations were associated with injury to the head region. Soft tissue injuries predominated among older children, and fractures were somewhat more common among younger children; however, this latter association did not achieve statistical significance. Lacerations were more common among younger children. Given the association of lacerations and head trauma, injuries to the head region were also somewhat more common among younger children; however, this relationship was not statistically significant.

There were no severe neurologic injuries or deaths among children in this study, but these outcomes do occur. There were 6 trampoline-related deaths reported to the CPSC from 1991 through mid 1995, and all of these injuries occurred to children younger than 16 years. Causes of death included "falling from," "jumping on," "doing flips," and "hanging from a spring by a necklace."²⁶ The 1.9% hospital admission rate in this study is slightly less than rates for other product-related pediatric injuries^{27 - 28} and also less than the 3.3% hospital admission rate reported for trampoline-related pediatric injuries nationally.²⁹

Twenty-one percent (n = 44) of trampoline-related injuries in this study occurred among children younger than 6 years. The CPSC recommends that children younger than 6 years not use trampolines³⁰ because of immature motor skills. The CPSC also recommends³⁰ that a ladder not be used with trampolines because it allows access to them by young children. This recommendation notwithstanding, a ladder is an available accessory for some trampolines. In 1991, a 3-year-old girl gained unsupervised access to a residential trampoline using a ladder provided for the trampoline and was strangled by her necklace, which caught in a spring as she dismounted.²⁶

Although trampoline springs are required to be covered by padding in other countries, such as England, New Zealand, and China (Jean Kennedy, Division of Corrective Actions, CPSC, oral communication, October 1997), "the suspension system need not be totally covered"³¹ in the United States. Trampolines sold in this country must have pads that cover the top surface of the trampoline frame, the spring-anchor attachments on the frame, and the outer hooks of the springs, but not necessarily the bodies of the springs. Fifteen percent (23/153) of children in this study were injured when they fell onto the trampoline frame, and 5.2% (8/153) of children received injuries from a fall onto the springs. Six percent of children in the study by Woodward et al¹ were injured when they fell through or on the springs. Some of these injuries may have been prevented if frame pads were required to cover the entire trampoline suspension system.

IN DECEMBER 1997, the F8 Subcommittee for Trampolines and Related Equipment of the ASTM met to discuss proposed changes to the ASTM voluntary standard on trampolines.³¹ There has been no change in this standard since 1995, but there is now discussion on several issues, including recommendations that trampolines not be used by children younger than 6 years, that ladders not be sold for trampolines, and that padding cover all of the springs (Jean Kennedy, Division of Corrective Actions, CPSC, oral communication, December 1997).

Approximately 13% (25/187) of children in this study were injured when they attempted a flip on the trampoline, and 56.9% (66/116) of children had previously attempted a flip, according to their parents. This is alarming because neurologic injuries related to trampolines are primarily due to trauma to the cervical spinal cord sustained during somersault attempts in the center of the trampoline bed.⁷ Many of these severe injuries occur with spotters present, because spotters at the edge of the trampoline bed cannot prevent an injury occurring in the center.¹⁵

Nine percent (17/188) of children in this study were injured when they fell from the trampoline or while getting off the trampoline. This is less frequent than reported in previous studies. Two US studies^{1 ,3} reported that 29% of injuries resulted from a fall off a trampoline, and 1 study from New Zealand² reported that 80% of injuries were due to this mechanism. Strategies to prevent injuries from falls off a trampoline include the use of spotters, a trampoline bed of adequate size, an energy-absorbing surface without obstructions around the trampoline, and lowering the height of the trampoline bed to ground level by locating the trampoline in a pit. Based on US data, however, these prevention strategies would address only a few of the injuries related to trampolines.

The use of a trampoline simultaneously by several persons has been identified as a risk factor for injury. There was more than 1 person on the trampoline at the time of injury in greater than three fourths (106 of 139) of cases in this study. This finding is similar to the percentages reported by Woodward et al¹ and Larson and Davis,³ who found that more than 1 person was on the trampoline at the time of injury in 77% and 64% of cases, respectively. In this study, 30.3% (57/188) of injuries were caused by a collision with another trampolinist or another trampolinist falling on the patient. It could not be determined in this study whether the simultaneous use of a trampoline by several persons was associated with a higher rate of injury, because the amount of time children spent trampolining with and without others simultaneously on the trampoline is unknown.

Children may be especially at increased risk for injury when trampolining with a heavier person because of the amount of force with which a trampoline bed recoils following a bounce by the heavier person.¹⁰ The patient weighed less than the other persons on the trampoline in 56.5% (48/85) of cases involving multiple simultaneous users with known weights in this study and was the heaviest person in 10.6% (9/85) of these cases. The amount of energy transferred to a child's body on impact with the surface is the major determinant of injury. Compared with adults, children are at an increased risk for injury when exposed to situations with potential for the transfer of large amounts of energy because of immature judgment, coordination, strength, and anatomical characteristics, such as open bone physes.

Ninety-six percent (119/124) of injuries in this study were associated with the use of a backyard trampoline. Ninety-three percent of trampoline-related pediatric injuries in the United States²⁹ and 59% to 71% of trampoline-related injuries in New Zealand² occur in the residential setting. The AAP²⁵ and many injury researchers^{1 - 2 ,9 ,15 - 16 ,29} have recommended against the recreational use of backyard trampolines.

The annual number of children treated in EDs for trampoline-related injuries doubled from 1990 to 1995 in the United States,²⁹ most likely because of an increasing use of backyard trampolines. Although imperfect, an indirect measure of trampoline usage is trampoline sales. An estimated 5-fold increase in trampoline sales during recent years was reported by a leading trampoline manufacturer (CNBC "Steals and Deals" television show, July 29, 1997). As many as 500 000 trampolines intended for backyard use are sold each year nationally, representing $125 million in retail sales.³² These trampolines can be purchased for less than $250.

Trampolines are sold with warning labels and safety instructions, which generally follow the voluntary standards established by the ASTM.³¹ Product warning labels and public education, however, have not adequately prevented injuries to children associated with other consumer products,^{28 ,33 - 34} and the findings of this study demonstrate that these prevention strategies are also insufficient to prevent trampoline-related injuries. Despite manufacturer warnings against multiple simultaneous users on the trampoline and warnings of the risks of severe neurologic injury when doing flips, 76.2% (106/139) of injuries in this study occurred when more than 1 person was on the trampoline, and 56.9% (66/116) of parents reported that their child had attempted a flip. A knowledge of injury risks associated with trampolines was not an effective deterrent to trampoline use. Seventy-three percent (81/111) of parents indicated that they had been aware of the potential dangers of trampolines before the injury event, and 10% (12/120) reported that their child had been injured previously on a trampoline. Moreover, 54.8% (63/115) of children continued to use a trampoline after the current injury event. Adult supervision cannot be relied on to decrease trampoline-related injuries. Approximately 56% (65/117) of study children , including 73.3% (22/30) of children younger than 6 years, were being directly supervised by an adult at the time of injury. Even supervision by trained spotters has been ineffective in preventing serious trampoline-related injury.¹⁵ The inadequacy of warning labels, public education, and adult supervision is consistent with injury prevention theory, which states that injury prevention strategies that require increased human effort or action are least likely to be effective.³⁵

Injuries associated with trampolines and infant walkers predominantly occur among the pediatric population and are almost entirely due to falls.^{28 ,33 ,36} There were 11 infant walker–related deaths reported from 1989 through 1993,³⁷ compared with 6 trampoline-related deaths from 1991 through mid 1995.²⁶ The number of trampoline-related injuries treated in US EDs is more than twice that for infant walker–related injuries³³ and, in contrast with infant walker–associated injuries, is increasing at a rapid rate.²⁹ The AAP and the American Medical Association are among the groups calling for a ban on infant walkers, and the latter has described infant walkers as a "lethal form of transportation."³⁸ Health and child advocacy organizations should take similarly strong positions regarding backyard trampolines, especially given the rapidly increasing number of pediatric injuries associated with their use nationally.²⁹ These organizations should call for a ban on the sale of trampolines intended for private recreational use. The trampoline should be regarded as a training device for selected athletes and not as a toy. Other safer recreational activities and sports, including other forms of gymnastics, can provide children with the same skills and enjoyment obtained from trampolining.

The potential for selection bias regarding who seeks care in the ED of an academic children's hospital is a limitation of this study. The findings of this study may not be representative of trampoline-related injuries to children treated in other medical settings. Data obtained from parents following the ED visit could have been affected by recall bias, a respondent's unwillingness to admit to the truth, or a respondent's wish to please the investigators.

Trampoline-related injuries to children treated in the ED are almost exclusively associated with the use of backyard trampolines. The prevention strategies of warning labels, public education, and adult supervision are inadequate to prevent these injuries. Children should not use backyard trampolines, and the sale of trampolines for private recreational use should be halted.

Editor's Note: Perhaps a better name for these contraptions would be "tramp-o-children." —Catherine D. DeAngelis, MD

Accepted for publication February 23, 1998.

Corresponding author: Gary A. Smith, MD, DrPH, Division of Emergency Medicine, Children's Hospital, 700 Children's Dr, Columbus, OH 43205 (e-mail: gsmith@CHI.OSU.EDU).