Introduction
The trampoline was invented by American circus performer George Nissen in 1936. He originally designed the trampoline, patented as “tumbling device” in 1945, for acrobats. (Briskin and LaBotz 2012) During World War II, the trampoline was used to train fighter pilots. (Eberl et al. 2009; Smith 1998) After 1940, it became a piece of sports equipment, and from the late 1970s, trampolines were increasingly found in amusement parks, schools, and daycare centers. (Thi Huynh et al. 2018) Because the trampoline promotes coordination, fitness, muscle strength, and balance, it be-came a popular sport and recreational equipment.
Over the last twenty years, private trampolines have seen a rapid growth in popular-ity. According to the Swiss Advisory Center for Accident Prevention, 17% of all Swiss households with children under the age of 15 had a trampoline in 2020. (BFU 2020) However, the increasing popularity of private garden trampolines led to a concomitant rise in trampoline related accidents.
Trampoline accidents were first documented by Zimmerman in 1956. (Zimmerman 1956) According to the US Consumer Product Safety Commission, the most com-mon causes of trampoline accidents nowadays are collisions with other trampoline users, rough landings while jumping, falls onto the trampoline frame or falling off the trampoline altogether. (Klimek et al. 2013)To minimize the risk of accidents while trampolining, the Swiss Advisory Center for Accident Prevention has issued safety recommendations for it. Its recommendations are as follows:
(BFU 2017)
• Supervise children
• Regular trampoline maintenance
• Jump alone
• Set up trampoline without obstacles
• Use safety net
• No flips
A Canadian study, based on cross-sectional survey data, reported that most parents whose children regularly use recreational trampolines lack basic trampoline safety knowledge. (Beno et al. 2018) Such gaps in parental trampoline safety knowledge are likely an important factor in the increase of injuries sustained during children’s recreational trampoline use. (Meyerber et al. 2019) Considering the serious risks for injury, some healthcare professionals deem trampolines that are used for leisure activities of children as too dangerous, with some even calling for the ban of such trampolines. (Eberl et al. 2009; Furnival, Street, and Schunk 1999; Brown and Lee 2000; Hammer, Schwartzbach, and Paulev 1982). Between 2003 and 2009, the university hospital in Bern, Switzerland, recorded 286 trampoline accidents involv-ing children. The number increased from 13 patients in 2003 to 86 in 2009. (Klimek et al. 2013) Today, about 95% of all trampoline accidents happen at home, with fractures and soft tissue wounds of the extremities among the most common injuries. (Korhonen et al. 2018) Head injuries, on the other hand, occur less frequently, and currently no published data are available on dental injuries owing to trampoline use. (Cho et al. 2019)
The aim of this study was therefore to investigate dental accidents on garden tram-polines in Switzerland and to analyze possible correlative factors. The null hypoth-esis was that neither multi-person trampolining, nor any design or installation fea-tures of trampolines would have an association with the frequency of trampoline related dental injuries.
Material and method
Selection of aerial photographs
In a preliminary investigation, aerial photographs of various providers were as-sessed regarding their suitability to visually detect garden trampolines. Images pro-vided by map.search.ch, which are based on data from Swisstopo, the Swiss Fed-eral Office of Topography, were deemed the most suitable for the present study. Its aerial images, acquired through airplane-based vertical aerial photography, have a ground resolution of 0.1 m and undergo regular updating (swisstopo.admin.ch)
Search for eligible households
From May to June 2020, aerial photographs of each German speaking municipality of Switzerland, were visually screened by an investigator (AWH) for private resi-dences with a round garden trampoline. All 1438 German speaking political munic-ipalities of Switzerland, listed in the atlas "The 4 language areas of Switzerland by municipality 2016" published by the Swiss Federal Statistical Office, were included in the screening.
The aerial photographic map of each municipality was viewed on map.search.ch, employing the maximum zoom factor. (Fig. 1) Only round trampolines were consid-ered. Based on color and shadow features, trampolines could be distinguished from inflatable pools. In cases of uncertainty, another trampoline in the same municipality was sought out. Trampolines were ineligible for inclusion if they belonged to an apartment building or if it was unclear whose property they were located on. If the aerial photographs of a municipality featured more than one eligible private resi-dence, one of these residences was selected at random. In some municipalities, no trampoline could be found. The addresses of the selected private residences were collected using an in-build tool of map.search.ch and stored in a purpose-built, se-cured database. For the avoidance of doubt only one trampoline per municipality was selected for inclusion in our work.
Questionnaire survey
In mid-June 2020, a questionnaire was mailed to each address collected in the da-tabase. It included a hand-signed cover letter, outlining the aim of the study and the irreversibly anonymized nature of the survey, and a self-addressed stamped enve-lope to reply. Owing to the voluntary nature of the survey and the irreversibly anon-ymized data collection, the local ethics committee waived the requirement for ethi-cal approval (EKNZ Req-2020-00605). The questionnaire consisted of one open-ended and ten close-ended questions. The questionnaire comprised questions on design and installation features of the garden trampoline, details on persons who regularly use the trampoline and on multi-person trampolining, any trampoline re-lated accidents including details on the type of accident and sustained injuries. The questions included in the questionnaire are reported in Table 1 in detail. Question-naires returned by the end of August 2020 were considered in the analysis.
Statistical analysis
Data from the questionnaires were recorded in an Excel spreadsheet. Descriptive analysis included number of cases and percentages for categorical parameters (e.g., safety net). For continuous parameters (e.g., diameter of the trampoline), the mean with standard deviation was calculated. Associated p-values were calculated using appropriate significance tests (Chi2 test, t-test, and Wilcoxon rank sum test). To estimate which parameters had an influence on the occurrence of accidents, lo-gistic regressions were calculated (accident yes versus no). The resulting estima-tors were odds ratios (OR) with the corresponding 95% confidence intervals and p-values. For all test procedures, a 5% (two-sided) probability of error was set as the significance level. Owing to the purely descriptive nature of the study, no adjustment was made for the probability of error for multiple comparisons. All analyses were carried out with the statistical program R version 3.5.1 (R Foundation for Statistical Computing, Vienna, Austria).
Results
Of the 1212 letters sent out, the post office returned six undelivered letters to send-er. Within ten weeks, 730 questionnaires were returned. This corresponds to a re-sponse rate of 60.2%. 93 households replied that they either no longer owned a trampoline or had never owned one. A total of 637 questionnaires of private house-holds with a garden trampoline were available for analysis.
Most households owned an onground trampoline (95.5%, n=577) with a safety net (72.1%, n=413). (Fig. 2) The diameter of the trampolines averaged 373.5 cm (80-600, SD 81.4 cm). The average age of a trampoline user was 14.3 years (1-67, SD 9.5 years). 52.4% (n=319) of trampoline users in this study were female, 47.6% (n=290) were male. In 11.1% (n=71) of the households, the trampoline was used by only one person, in 47.4% of the households by two people, in 25.7% (n=164) by three people, and in 15.7% (n=100) by four or more people. 86.7% (n=548) of the respondents reported that several persons use the trampoline simultaneously. Trampolines used for multi-person trampolining featured a safety net more frequent-ly (89.7%) compared with trampolines that were reported to never see multi-person trampolining (79.9%) (p=0.002). Large trampolines (mean 396.8 cm, SD 87.6cm) and trampolines used by older children (mean 15.1 years, SD 7 years) were more likely to have no safety net (p<0.001). 83.5% (n=532) of respondents stated that an accident had never occurred on their trampoline, 16.5% (n=105) reported that there had been a trampoline related accident. The average age of a person involved in an accident was 13.5 years (2-53, SD 4.7 years). In 33% of accidents, only one per-son was on the trampoline, and in 67% two or more people were. More accidents involved males (52.9%) than females, (47.1%), with no statistically significant differ-ence between genders (p=0.283).
A total of 99 injuries were reported. The injuries were divided into five groups for evaluation: teeth, head, arms, legs, and trunk. Dental injuries occurred in 23.2% (n=23) of the accidents. Head injuries, such as concussions, lacerations, or injuries to the eyes and lips, occurred in 21.2% (n=21) of cases. Arms and legs were frac-tured, sprained, or otherwise injured in 25.3% (n=25) of the accidents. The trunk was involved in 5.1% (n=5). A physician or dentist was consulted in two-thirds of the accidents.
Of the 23 dental accidents, 9 were dental fractures, 7 were concussions, 4 were avulsions, and 3 were dislocation injuries. In these cases, the teeth were struck on the person’s own knee (n=6), on the trampoline itself (n=6), on the head of another trampoline user (n=5), or there was a fall from the trampoline (n=1). In five cases, no details on the circumstances of the accident were provided.
Dental accidents occurred 3.6 times more frequently (95% Cl: 0.5.27) when the trampoline was used by several persons at the same time. Dental accidents were 2.4 times more frequent on inground trampolines (95% Cl: 0.5.11) than on onground trampolines.
For the occurrence of accidents in general, the diameter of the trampoline had the greatest influence (p=0.044). Trampolines on which an accident occurred were, on average, 20 cm larger compared with trampolines on which no accident took place. (Fig. 2) Neither the presence of a safety net, nor the type of trampoline, nor the number of people on the trampoline, nor the age and gender of the trampoline us-ers had a significant influence on the frequency of accidents (p?0.084).
Discussion
The study showed, based on survey data of 637 Swiss households with a garden trampoline, that the diameter of a trampoline’s bounce mat had a significant impact on the frequency of trampoline related accidents. More accidents happened on trampolines with a larger bounce mat. Consequently, the null hypothesis had to be rejected. Dental injuries – tooth fractures, concussions and dislocation injuries showing comparable rates - were sustained in 23.2% of the accidents. The present study, therefore, suggests that concerted efforts are needed to reduce the frequency and severity of trampoline related accidents.
A retrospective study reported that in a British hospital 11.1% of all fractures in chil-dren who had to be treated under general anesthesia were due to trampoline acci-dents. This was about the same as for soccer accidents and higher than for bicy-cling, skating, horseback riding, or climbing. (Bhangal, Neen, and Dodds 2006) Ac-cording to the present study, a trampoline related accident occurred in 16.5% of the households with a garden trampoline. Compared with other high-risk leisure activi-ties and sports, such as riding a kick scooter (29.1%) or mountain biking (53.1%), trampoline use entailed fewer risks of sustaining injuries.
(Baumgartner et al. 2012; Müller et al. 2008) However, owing to the closed-ended questions used in present study, no minor accidents were reported by respondents. If minor accidents had been included, the accident frequency might have been higher.
A dental accident occurred on 3.6% of the total 637 trampolines assessed in the present study. This rate is lower compared with mountain biking (5.7%), but higher than for kick scooter riding (3.1%) or skiing (2.2%). (Müller et al. 2008; Baumgartner et al. 2012; Innerhofer et al. 2013) As in mountain biking and kickscooter riding, tooth fractures occurred most frequently in trampolining. An investigation into the causes of general injuries, such as bruises, sprains, and fractures sustained while using a trampoline revealed that most injuries occurred owing to incorrect landing on the bouncing mat (42%). Falling off the trampoline (27%) was ranked second among the causes of accidents, followed by injuries that occurred on the trampoline frame (19%) or resulted from colliding with another trampoline user (10%). Dental injuries were listed among accidents frequently caused by the trampoline frame. (Alexander et al. 2010) In accordance with previous studies, 86.7% of respondents in the present study reported that their trampoline was regularly used by more than one person at a time. (Shields, Fernandez, and Smith 2005) The results of the pre-sent survey indicate a 3.6-fold increase in the likelihood of trampoline related dental injuries when the trampoline is used by two or more persons simultaneously. More-over, the results of the present study suggest that the diameter of the bounce mat diameter has a significant impact on the frequency of trampoline related accidents. This may be due to the fact that bigger bounce mats facilitate higher jumps, which, in turn, involve a heightened risk for injury. In addition, multi-person trampolining is easier on bigger sized bounce mats and multi-person use is considered an im-portant risk factor for trampoline related accidents. It is therefore important to in-crease parental and public awareness regarding the potential dangers of multi-person trampolining regardless of trampoline size.
Compared with other studies, the victims of trampoline related accidents tended to be older according to the present study. (Thi Huynh et al. 2018; Woodward, Furnival, and Schunk 1992; Choi et al. 2018) However, most of the other studies' data were collected in hospitals, where bone fractures were particularly prominent. Yet, fractures are more common in children younger than 6 years. (Klimek et al. 2013; Choi et al. 2018) In particular, when the trampoline is used by more than one person at a time, larger and heavier children may generate more recoil from the jumping surface. This recoil can cause significant injury to a young child. (Woodward, Furnival, and Schunk 1992)
The present study revealed that the presence of a safety net had no statistical effect on the occurrence of trampoline related accidents. This finding is in line with data reported in a previous study, which showed that bone fractures occur more fre-quently on trampolines with a safety net.(Klimek et al. 2013) Presumably, safety nets may encourage trampoline user to make more daring and risky jumps.
The present study has some limitations that require careful consideration. The study was subject to the inherent methodological and reporting limitations of question-naire based surveys. Considering the voluntary nature of the survey, the response rate of 60.2% was high, which may be indicative of trampoline owners’ interest in trampoline related accidents. Nevertheless, the potential of response and participa-tion biases ought to be taken into account. A random selection of Swiss households with a garden trampoline was made to reduce the risk of selection bias. Though no data on the service time of a respondent’s trampoline were gathered, the random selection of households with garden trampolines ensured that the average service time of trampolines included in the study was close to the factual average across Switzerland. It is, however, crucial to take account of the fact that only trampolines located in gardens of private residences were eligible for inclusion. Trampolines belonging to apartment buildings were excluded in the present study because it was unfeasible to identify persons who regularly used such trampolines. Conse-quently, the sample of the study is biased towards persons and families living in detached houses and it is conceivable that, say, multi-person trampoline use is more frequent on trampolines belonging to apartment buildings.