Introduction

Contemporary circus embodies the tension between risk and death with its “death-defying” stunts and emotive narratives.¹ Audiences are drawn to watching acrobatic performers face and overcome real or staged risks. Show creators seek to ensure that this sensation remains an aesthetic element² since the audience does not expect to see death.³ When catastrophic incidents (i.e., resulting in death, permanent disability, temporary paralysis or activity-induced heat stroke or cardiac event)⁴ occur publicly, media coverage is often characterized by sensationalized, risk-averse discourse that calls for increased safety, further contributing to the perception of danger in circus.⁵ Although these incidents are acknowledged and discussed by the industry, the data have not been compiled in a way that can support rational approaches to safety.

Catastrophic injuries in circus have not been characterized in the literature. The majority of research does not specify whether any have occurred and been classified with severe injuries,^6,7 although a recent consensus specified that this should occur.⁸ Selected catastrophic incidents have been included in the narrative of published articles^9,10 or as case studies,¹¹ but the objective characteristics of the incidents and consequent injuries remain unknown. Furthermore, circus performers are classified with other entertainers in occupational databases, which makes it infeasible to use these data to support circus-specific safety. Thus, there is no data to support a greater understanding of these injuries in the unique circus context, which is needed to investigate underlying mechanisms, establish enhanced safety protocols and legislation, foster the creation of injury surveillance mechanisms (e.g. protocols and registry) and provide public-facing information related to the actual risks involved in circus. We attempted to address this issue by conducting a systematic review of the scientific literature, but since our search strategy returned zero relevant papers, we turned to the gray literature as a first step¹² to characterize incidents resulting in catastrophic injury in circus.

Methods

A systematic search strategy was developed¹³ to identify incidents resulting in catastrophic injury in circus from the gray literature. The Google search engine was used to identify incidents that occurred between January 2000 and March 2020 in which a catastrophic injury was incurred by a circus artist (professional or student) or technician. The keywords circus and cirque were searched in combination with death, fatal, mortality, accident and spinal cord injury. Incidents involving suicide, natural disasters or circus animals were excluded. Incidents resulting in temporary paralysis were excluded due to vague details in reporting. Only incidents prior to March 2020 were included since sudden changes in training and performing practices due to COVID-19 pandemic-related closures may have resulted in incidents occurring due to atypical conditions. Context (performance, training, rehearsal, technician), age, sex, discipline (as previously described⁸) and mechanism of injury (as previously described,⁸ modified to add electrocution and classify falling to the mat, net or floor as “direct contact with ground” since it was not always clear where the artist landed) were extracted from articles. Attribution was inferred based on the incident description. External attribution was indicated if there was information about environmental contributions, including equipment failure. Internal attribution was indicated otherwise. A table of incidents was shared with two experts in the field who contributed additional events from their experience.

Descriptive analyses were completed to characterize the incidents. Mortality rates were calculated for artists and technicians in a professional circus company (/100,000 FTE) based on publicly available data, as well as in a circus college (/100,000 students) based on data from our college over a ten-year period (2009-2018), selected according to the period with the maximum number of known incidents.

Results

A total of 1,460 articles were retrieved, of which 208 were duplicates. The headlines of the remaining 1,252 articles were screened, and 774 were considered irrelevant and excluded. The full contents of the remaining 479 articles—which described 25 incidents—were reviewed, and the experts contributed an additional thirteen incidents, six of which came from a book published in Portuguese.¹⁴ Overall, 38 incidents that resulted in 39 catastrophic injuries (n = 20 fatal; n = 19 non-fatal, including n = 14 spinal cord injury, n = 3 serious head injury, n = 2 unknown; see Table 1) were described. Females accounted for 33% and 37% of fatal and non-fatal injuries, similar to the sex distribution of artists in professional companies. For artists (n = 32), incidents occurred during performance (n = 22) and training or rehearsals (n = 10). The majority were in the aerial discipline (n = 21), followed by ground acrobatics (apparatus propulsion) (n = 6) (see Table 1).

For artists (n = 32), the mechanism of injury—with one exception—was direct contact with the ground (n = 31; one incident also involved direct contact with another artist) with greater attribution to intrinsic (n = 20) compared to extrinsic (n = 12) factors for ground contact (n = 1, not reported; n = 1, classified as both intrinsic and extrinsic). For technicians (n = 6), the mechanisms of injury were direct contact with an object (i.e. hit by props/equipment: n = 4) and electrocution (n = 2).

Over ten years, there were no deaths at a circus college and five deaths at a professional circus company (three artists and two technicians), resulting in estimated ten-year mortality rates of 0/100,000 for students, 37.5/100,000 FTE for artists and 25.0/100,000 FTE for technicians.

Discussion

This study provides the first approximation of the characteristics of catastrophic injuries in circus, which most commonly involve a fall (direct contact with the ground) during professional performance. Although the details available through gray literature were insufficient to support concrete recommendations for risk assessment and safety protocols, they provide basic information and support the need for more systematic data collection and reporting of catastrophic incidents in circus. Contrary to previous work estimating that 75% of severe accidents experienced by professional aerialists were caused by rigging failure,⁹ this review found that from 2000 until the onset of the COVID-19 pandemic, the attribution of intrinsic vs. extrinsic factors was approximately equal (seventeen vs. nineteen incidents, one both). This may be partly due to recent advancements in engineering and biomechanical knowledge.^15,16 Extrinsic risk factors that have been associated with general injury in circus include scheduling,¹⁷ discipline,¹⁸ warm-up,¹⁹ clothing, temperature and mats/safety equipment,²⁰ and numerous others have been postulated.^21,22 Many of these factors have the potential to interact with or be falsely attributed to intrinsic factors, as seen with one incident with sufficient details to attribute both internal and external factors. Thorough details beyond those reported in media stories are needed to better understand the circumstances contributing to catastrophic injury in circus.

Intrinsic attribution may be faultily assumed to be due to reckless, sensation-seeking behaviour. However, artists tend to be more motivated by mastery and control than sensation-seeking.⁹ It may also be assumed that circus tricks are constantly becoming more difficult or “riskier” when, in fact, the risk is offset by increasing mastery and knowledge through movement research and pedagogical advancement.¹⁰ Artists have suggested that they are more likely to experience a mishap that may lead to injury when performing “simpler” moves or familiar sequences, when they fall into “autopilot” and are not 100% focused, than when they are performing more difficult stunts.²³ Other intrinsic risk factors such as age,^18,24 disordered eating behaviours,^18,25 self-efficacy, fatigue, emotional exhaustion, injury,²⁶ perceived risk, personality and emotion regulation²⁷ are associated with general injury in circus. These studies collectively suggest that holistic interventions²⁸ addressing psychosocial and physical factors should be prioritized for future study of performer safety, keeping in mind that several contextual factors can affect artists’ mindsets and performance.^21-23

The majority of fatal catastrophic injuries occurred during performance. While exact rates could not be calculated, mortality rate estimates in circus performers (37.5/100,000 FTE) and technicians (25.0/100,000 FTE) are similar to those of occupations with a risk of fall from height (roofers, 54.0/100,000 FTE; helpers in the construction industry, 40.0/100,000 FTE; grounds maintenance workers, 19.8/100,000 FTE).²⁹ From an occupational health and safety perspective, an international registry of circus-related injuries akin to general trauma or workforce registries would benefit the industry by enabling objective investigation of injury rates and characteristics. This, coupled with an environmental scan to estimate the workforce more closely, would allow for a proper evaluation of performance, injury and safety in circus. These data would provide sound information to estimate incident rates for comparison with other occupational mortality and long-term disability data, and they are essential for thorough risk assessment and subsequent rational decision-making to provide adequate safety.

Conclusion

While the data extracted from the gray literature were insufficient to support concrete recommendations to inform safety protocols, we demonstrated that the majority of catastrophic incidents occurring in circus contexts were characterized as falls (direct contact with the ground) during professional performance. Mortality rates were within the range of occupations with a risk of falling from height. The analysis of non-systematic data, as shown in this article, appears limited and insufficient to establish recommendations or consensus, although it is an important step in delimiting the problem.¹² Improving protocols and tools for systematically recording accidents is an important next step to support evidence-informed decision-making in circus safety.