In this study, we aim to evaluate the current trends in pediatric fractures related to trampolines.
The National Electronic Injury Surveillance System was queried for fractures occurring between 2008 and 2017 in individuals aged 0 to 17 years. Sex, anatomic region, locale of injury, admission status, and year of injury were recorded. Incidence rates were calculated by using national census data. Poisson regression analysis was used to test for changes in fracture incidence across the time period. Logistic regression analyses were used to test temporal trends in the odds of a fracture occurring at a place of recreation or sport and a patient with a fracture being admitted.
Between 2008 and 2017, there was a 3.85% (95% confidence interval [CI]: 0.51–7.30) increase in the incidence of trampoline-related pediatric fractures per person-year. The incidence of pediatric trampoline-related fractures increased from 35.3 per 100 000 person-years in 2008 to 53.0 per 100 000 person-years in 2017. There was no change in the odds of a trampoline fracture requiring hospitalization (odds ratio per 1 year: 1.02; 95% CI: 0 6–1.07; P = .5431). There was a significant increase in the odds of a fracture occurring at a place of recreation or sport (odds ratio per year: 1.32; 95% CI: 1.21–1.43; P < .0001).
Between 2008 and 2017, there was a significant increase in the national incidence of trampoline-related fractures. We identified a significant increase in the proportion of trampoline fractures that occurred at a place of recreation or sport. Advocacy campaigns should consider these sites in their prevention efforts.
Trampolines are a common source of pediatric injury and frequently result in a fracture. Additionally, trampoline parks are becoming an increasingly popular avenue for participation.
In conjunction with an overall increasing number of trampoline-related pediatric fractures presenting to US emergency departments between 2008 and 2017, there is also a greater number of these fractures occurring at a place of recreation or sport.
Surveillance is a key component of injury prevention. Injury monitoring is necessary to understand injury burden and, more importantly, to identify changes in injury patterns that can be used to improve the effectiveness of prevention efforts. Numerous studies have used data from large national databases to document temporal changes in trampoline injury patterns.1–5 Between 2002 and 2011, trampoline-related injuries accounted for an estimated 1 million emergency department (ED) visits with estimated health care costs in excess of $1 billion.3 Among trampoline-related injuries in the United States, fractures are frequently reported, accounting for 29% of ED-related visits.3
Trampoline injury prevention has traditionally focused on trampoline injuries in the home. The American Academy of Orthopaedic Surgeons (AAOS) and the American Academy of Pediatrics (AAP) have advocated against childhood trampoline use with targeted campaigns, strongly recommending against trampolines in the home. Karkenny et al5 evaluated the effectiveness of these injury prevention campaigns by studying the relationship between implementation of these campaigns and temporal changes in the incidence of trampoline fractures. A 1999 AAP policy statement was associated with a plateau in trampoline fractures, and more importantly, policy statements from both the AAOS and AAP in 2005 and 2006, respectively, were associated with a 4.3% drop in trampoline fracture incidence in subsequent years, suggesting a public response to these statements.
Within the last decade there has been a shift in culture surrounding trampolines. In 2000, the Sydney Olympic games included trampolining as a medal event, illustrating an evolving organized athletic interest in trampolines. Recreationally, trampoline parks have soared in popularity from just 3 parks in existence in 2009 to >1000 worldwide in 2018.6 These avenues for participation and renewed interest in trampolining as a recreational and competitive sport may be expected to result in changes in reported injuries and/or shifts in the locale of injuries.1,7
Our purpose in this study was to estimate the incidence of pediatric fractures related to trampoline use and, more importantly, to identify changing patterns related to trampoline fractures in our culture today. Our ultimate goal in this work is to educate parents, guide policy, and inform existing prevention efforts.
Data were retrospectively collected from the National Electronic Injury Surveillance System (NEISS), which is maintained by the US Consumer Product Safety Commission (CPSC). NEISS is based on a probability sample of 100 US hospitals and is used by the CPSC to provide nationwide estimates of ED visits for product-related injuries (for example, trampolines, skateboards, scooters, etc). Information from each of the 100 participating EDs is reviewed by a NEISS coordinator at the participating hospital and further checked by CPSC-designed software to ensure acceptable entries. The enrolled EDs are representative of trauma centers, rural and urban community hospitals, and adult and children’s hospitals that are open 24 hours per day with a minimum of 6 beds. Study weights allow for nationally representative estimates. Weights rely on the geographic location and patient volume of the participating EDs.
Variables available in the database include age, sex, race, ethnicity, diagnosis, body region, locale of injury, product, and a brief narrative of the incident. The NEISS is a reliable and reproducible source of data for epidemiological studies in the United States.8 More information regarding the use of data for the NEISS can be found on the CPSC Web site.9 This study was considered exempt from institutional review because the database uses deidentified, publicly available data.
NEISS was queried for all fractures (NEISS code 57) among children and adolescents 0 to 17 years old who were treated at a participating hospital between 2008 and 2017 (n = 257 145 cases; nationally representative n = 7 109 079). Age (0–4, 5–9, 10–14, and 15–17 years), year of injury, sex, locale of injury (recreation and/or sport related, home, or other), admission status (admitted versus not admitted), and anatomic region were collected for each case. Trampoline-related injuries (NEISS code 1233) were isolated. The anatomic region of fracture was stratified into 3 major groups: upper-extremity fractures, axial skeletal fractures, and lower-extremity fractures. For ease of group comparisons, the pelvis was grouped with the axial skeleton. It should be noted that NEISS can be queried for multiple injury types (for example, concussion, abrasion, dislocation, sprain, etc). Fracture was selected as the primary search term because it is a specific diagnosis, allowing for a homogenous data set for study.
A Poisson regression analysis was used to test for changes in trampoline fracture incidence, per person-year, across the time period. Weighted estimates provided by NEISS were used for all calculations. National census data representing all children and adolescents (<18 years of age) in the United States for the years 2008 to 2017 were used as the denominator in all incidence calculations. Variance estimates were obtained by using the jackknife method of variance estimation. Significance was assessed on the basis of the 95% confidence intervals (CIs). The parameter estimates, rate ratios, representing the linear increase in trampoline fracture incidence per year were considered statistically significant if the lower or upper CI excluded 1 (null or no difference equals a rate ratio of 1). Among trampoline fractures, multivariable logistic regression analyses were used to test for temporal trends in odds of a patient with a fracture being admitted and a fracture occurring at a place of recreation or sport versus at home. The NEISS coding manual defines a place of recreation or sport as a bowling alley, an amusement park, sports fields or stadiums, lakes, mountains, beach resorts, parks, or other recreational areas. Home is defined as the patient’s own home, someone else’s home, a foster home, rooms inside a home, porch or patio, yard or garden, garage or driveway, or sidewalk of a home.10 We also evaluated the incidence of trampoline fractures as a percentage of all pediatric fractures recorded in NEISS between 2008 and 2017. A logistic regression analysis was used to test for temporal changes in the odds of a trampoline fracture across the decade evaluated. Among all fractures, we also tested whether temporal changes in the odds of a fracture occurring at a place of recreation or sport differed between trampoline- and nontrampoline-related fractures (year × trampoline interaction). Sex and age were included as covariates in the logistic regression models. The complex survey design was accounted for by using the Proc SurveyLogistic procedure in SAS 9.4 (SAS Institute, Inc, Cary, NC). Jackknife 95% CIs were calculated for all parameters included in the logistic models.
Between 2008 and 2017, there was an estimated total of 989 338 pediatric trampoline-related ED visits. Fractures represented 26.9% (n = 266 373) of these visits. Of all pediatric trampoline fracture patients, 47.71% were girls, 60.3% were white, and 11.7% required hospital admission (Table 1). There was no change in the odds of a pediatric trampoline fracture requiring hospital admission during the decade evaluated (odds ratio [OR] per 1 year: 1.02; 95% CI: 0.96–1.07; P = .5431). Pediatric trampoline fractures most frequently affected the upper extremity (56.3%; 95% CI: 54.5%–58.0%), followed by the lower extremity (37.0%; 95% CI: 35.2%–38.8%) and the axial skeleton (6.8%; 95% CI: 6.2%–7.3%). Forearm (radius and/or ulna, 21.5%; 95% CI: 19.8%–23.2%) and lower leg (tibia and/or fibula, 18.3%; 95% CI: 16.7%–19.9%) were the most common body regions afflicted by trampoline-related fracture (Table 2). It should be noted that according to the NEISS coding manual, if an individual presented to the ED with injuries to multiple body parts, the “most seriously hurt” body part was the 1 that was recorded and coded for.10
In the decade evaluated, the incidence of pediatric trampoline fractures per person-year increased by an annual average of 3.85% (95% CI: 0.51%–7.30%). The highest incidence of fractures by age occurred in the 5- to 9-year-old age group (Fig 1, Table 1). In relation to locale of injury, pediatric trampoline fractures were most likely to occur at home; however, this proportion decreased throughout the decade evaluated (Supplemental Table 5). In contrast, pediatric trampoline factures were increasingly noted to occur at a place of organized recreation or sport (OR per year: 1.32; 95% CI: 1.21–1.43; P < .0001) throughout the decade. In other words, the odds of a pediatric trampoline fracture occurring at a place of recreation or sport increased by an average of 32% per year during the decade evaluated (Table 3).
A secondary analysis was used to estimate temporal trends in the incidence of trampoline fractures as a proportion of all pediatric fractures. In contrast to the incidence rate (fractures per person-year) presented above, this method allowed us to determine if changes in trampoline fracture incidence are distinct from changes in overall fracture patterns and, more importantly, whether injury patterns observed among trampoline fractures are different from injury patterns among all fractures. When modeled as a proportion of all pediatric fractures, changes in the proportion of fractures over the decade (Fig 2) closely mirrored changes in the incidence of trampoline fractures per person-year described above in the primary analysis (Fig 3). In 2008, trampoline-related fractures accounted for 3.59% (95% CI: 3.04%–4.14%) of all pediatric fractures. By 2017, 6.16% (95% CI: 5.31%–7.01%) of all fractures were related to trampolines (Fig 2). After adjusting for sex and age, there was a significant increase in the proportion of fractures that were related to trampoline injuries between 2008 and 2017 (OR per 1-year increase in time: 1.06; 95% CI: 1.05–1.08; P < .0001). Additionally, we tested whether temporal changes in the odds of a fracture occurring at an organized sport location were different for trampoline- versus nontrampoline-related fractures (year-by-trampoline interaction). There was a significant (P < .0001) difference in the slope, representing a linear association between time and the odds of a pediatric fracture occurring at an organized sport-related location among trampoline-related fractures (OR per year: 1.32; 95% CI: 1.22–1.43; P < .0001) versus nontrampoline-related fractures (OR per year: 0.99; 95% CI: 0.96–1.01; P = .3399).
Trampoline injuries remain a common source of orthopedic injuries and a significant public health issue. In the decade of 2008 to 2017, the incidence of fractures related to trampolines per person-year increased by an average of 3.85% per year. In 2017, trampolines were the cause of nearly 1 in 16 pediatric fractures presenting to US EDs. Despite this increase in the number of fractures, the hospitalization admission rate for these injuries remained relatively constant at ∼12%. Our study identified an important shift in the locale of trampoline injuries. Pediatric trampoline-related fractures were increasingly more likely to occur at a recreational or sports facility over the course of the study (Fig 4). It is expected that in any longitudinal data set fluctuations will occur from year to year. However, reporting overarching trends gives a more general picture of the impact trampolines may have on society, and for this reason, trends are the focus of the current study.
The modern trampoline design was developed in 1936 by George Nissen. The recreational and sporting use of trampolines can be traced to the 1950s with the first report of injury in 1956.11 As reports of both major and minor injuries increased over the next decade,12,13 particularly in relation to gymnastic sports, the National Collegiate Athletic Association eliminated trampolines from national competition in 1971. The AAP followed in 1977 with a recommendation to ban trampolines from schools and competitive sport in an effort to reduce the number of injuries in the pediatric population.14 Serious injuries remained a significant issue in the 1990s with a reported 14% yearly rise in trampoline-related fractures between 1991 and 1999,5 leading the AAOS in 1996 and the AAP in 1999 to issue policy statements recommending the abolishment of trampolines from homes, playgrounds, and education programs. Fracture incidence plateaued until 2003, when another increase was noted from 2004 to 2005.5 Again, policy statements from both the AAOS and AAP were published in 2005 and 2006, respectively, reaffirming each society’s previous statements. Loder et al3 and others report a subsequent overall decrease in the amount of trampoline injuries between 2006 and 2010, followed by a relative plateau in nationwide numbers from 2010 to 2013, which is potentially reflective of a public responsiveness to these policy statements. Our current findings indicate that this plateau has not continued. Pediatric trampoline fracture incidence has increased remarkably from 38.7 per 100 000 person-years in 2014 to 53.0 per 100 000 person-years in 2017.
Historically, trampoline fractures were predominantly related to injuries that occurred in the home environment. Our findings suggest that the recent increase in trampoline fractures may be partly due to the increased popularity of recreational and sport trampoline parks, which have grown from 35 in existence in North America in 2011 to nearly 400 by March 2018.15 Kasmire et al1 reported in 2016 that trampoline park injuries increased by nearly 12-fold between 2010 and 2014. Regarding our study, it should be considered that, as previously stated, the NEISS definition of a place of recreation or sport includes locales other than trampoline parks, so it is possible that our findings reflect this limitation and that not all visits in this category are from trampoline parks. Additionally, NEISS provides granular data only, and thus, it is possible that the increase in designation of trampoline fractures as having occurred outside of the home may be due to shifts in coding practices, a factor that would impact all fractures. To test this hypothesis, we tested whether the incidence of fractures outside of the home also increased in fractures unrelated to trampoline injuries. We found that the odds of a fracture occurring in a place of recreation or sport was constant among nontrampoline fractures, <1% change in incidence per year, providing additional evidence that the shift in the locale of injuries was unique to trampoline injuries.
Although the data indicate a greater fracture burden due to injuries that occurred at a place of recreation or sport, the current data do not allow us to draw any inferences regarding changes in the risk of a trampoline fracture at these locations. The increase in trampoline fractures at a place of sport or recreation may reflect a constant level of risk, with the increase in the number of cases being attributable to an increase in exposure rather than an increase in actual risk. Additional research is needed to define the risk level at trampoline parks and sport-related trampoline events relative to other, well-studied recreational and sport-related activities that are also a common source of pediatric fractures.
Childhood injury prevention is an important aspect of our health care system because of the potential for both intermittent and lifelong implications of injury. The orthopedic surgeon, pediatrician, and emergency medical personnel have key roles and responsibilities in their ability to advocate for safety in this arena. Karkenny et al5 highlights both medical personnel and their respective societies as effective advocates in a review of advocacy campaigns in relation to pediatric injury rates. Although the correlation between an injury prevention campaign and injury rates can only result in the “perceived effectiveness” of a campaign because of the vast number of possible confounding factors, it is our best mechanism to proactively discern potential changes in public behavior or awareness as a consequence of the campaign. To date, AAOS and AAP position statements have traditionally focused on dangers in trampoline injuries in the home environment. However, the current study demonstrates that trampoline-related fractures are increasingly occurring outside of the home (Table 3). In turn, advocacy and injury prevention campaigns need to adapt their message to accommodate shifts in the location of trampoline-related fractures to address the risk of injury outside of the home.
The strength of the current study is the use of a nationwide database to provide nationally representative estimates of trampoline fracture injury rates. National estimates are only possible if there are at least 20 reported cases, the weighted estimate is at least 1200, and the coefficient of variation is <33%. Therefore, some of the national estimates were not calculated (Tables 1 and 2). Another limitation of our study is that we do not have a denominator for the number of youth who are using trampolines to determine the true risk of trampolines. Furthermore, the mechanism of injury and circumstances surrounding the injury, such as the number of children on the trampoline, severity of injury, treatment, and final outcome, were not available. The NEISS database also relies on providers thoroughly noting the details of each case and the staff importing and coding the information from the visit reports accurately. For example, our data are limited by a large number of trampoline fractures occurring in “unknown” locales, with 40% of trampoline-related fractures in our study period being coded as “not recorded” regarding the locale of injury. In an effort to describe potential implications of these unknown locales, we have included a table (Table 4) that summarizes the distribution of injuries by body region among trampoline fractures coded as having occurred in the home, at a sports or recreational facility, and at an unknown location. For further description of the numbers of known and unknown locales of trampoline and all other fractures, please refer to Supplemental Table 5. Finally, estimates were only obtained from emergency visits and may underrepresent the true burden of pediatric trampoline injuries that present to other facilities, such as urgent care centers.
The United States has seen a significant rise in trampoline-related fractures between 2008 and 2017. Historically, injury prevention campaigns have been directed toward home trampoline use and have purportedly proved effective in reducing pediatric trampoline injury rates.5 The recent increase in fracture incidence coincides with an increase in the popularity of recreational and sporting trampoline parks. As these new avenues of participation become increasingly popular, future advocacy and injury prevention campaigns should potentially broaden their focus to address the changing locale of pediatric trampoline fractures.
Dr Georgopoulos conceptualized the study and reviewed and revised the manuscript; Dr Miller critically reviewed the manuscript for important intellectual content and revised it accordingly; Mr Brazell collected data, conducted initial analyses, drafted the initial manuscript, created figures, and reviewed and revised the manuscript; Ms Carry performed statistical analyses, created tables, and reviewed and revised the manuscript; Ms Holmes created figures, reviewed data, and reviewed and revised the manuscript; and all authors approved the final manuscript as submitted and agree to be accountable for all aspects of the work.
FUNDING: No external funding.
American Academy of Orthopaedic Surgeons
American Academy of Pediatrics
Consumer Product Safety Commission
National Electronic Injury Surveillance System
POTENTIAL CONFLICT OF INTEREST: The authors have indicated they have no potential conflicts of interest to disclose.
FINANCIAL DISCLOSURE: The authors have indicated they have no financial relationships relevant to this article to disclose.