2.3.1. Swiss cheese model.

The SCM is a complex accident causation model proposed by British psychologist James Reason in 1990, also known as the “REASON model” or “Aviation accident theory model”. The model was first applied to high-risk industries such as medical, transportation and mining [23–25]. Now it is extended to risk management and accident analysis in industries such as sociology and food safety, the mechanism of the accident is revealed and the complexity of the accident is explained [26,27]. Most scholars use the content of the model to systematically analyze all aspects of the accident, identify the potential risk factors that may lead to the accident, summarize the risk, and evaluate the overall risk level. In addition, according to the multi-layer defense mechanism proposed by the model, some scholars study the prevention mechanism and prevention strategy of the accident.

In recent years, many scholars have tried to improve and expand the SCM, and improve the applicability of the model. When Chinese scholars introduced the SCM, they used different analysis methods and made localized improvements based on China’s actual situation, the SCM is applied to Air Force flight training and safety accidents in specific industries in China [28,29]. Foreign scholars have extended the SCM in combination with other theories. Seshia et al. proposed a method of “Closing the Swiss cheese loophole” to improve patient safety [30]. Akuh and Atombo applied the SCM to the analysis of road transportation accidents and proposed an accident analysis method based on system theory [31]. Torres et al. combined the ACCIMAP and STAMP methods with the SCM, and proposed a new method for analyzing road transportation accidents based on the system theory, for the analysis of human errors in complex manual assemblies, the importance of dynamical system analysis is emphasized [32]. In summary, the SCM has been widely used in the study of safety accidents, but there are relatively few literatures on the analysis of tourism safety accidents.

In this study, the SCM serves as the theoretical basis for analyzing the configurational factors affecting sports tourism safety accidents in Tibet. This is reflected in three key aspects: first, the SCM emphasizes that accidents are the result of “systemic holes” created by multiple layers of defense, which is highly relevant to the multidimensional risk sources of high altitudes, extreme climates, and complex terrains in Tibet. Secondly, sports tourism activities usually require professional knowledge and skills, participants face high risks, and these activities often have strong time sensitivity and social influence, necessitates an analysis of the non-linear interactions between defense layers [33]. The “hole alignment” mechanism of the SCM reveals the nonlinear accumulation effect of defense layer failure, thus analyzing the interaction between different defense layers, rather than the linear effect of a single condition. Thirdly, sports tourism involves a wide range of factors, including natural, human, and technical risks, making the SCM highly applicable to the analysis of the influencing factors of sports tourism safety accidents in Tibet [34].

2.3.2 fsQCA method.

The Qualitative Comparative Analysis (QCA) method was proposed by Charles C. Ragin in 1987 [13]. Fuzzy-set QCA (fsQCA) is a case-oriented research method within QCA, based on Boolean algebra and set theory to examine the relationship between antecedent conditions and outcomes [35].

Compared with the traditional Quantitative analysis method, QCA method can capture the nonlinear relationship between variables and multiple concurrent causal effects, which is suitable for dealing with complex social phenomena [36]. It’s a method that facilitates the understanding of complex situations involving multiple factors potentially contributing to an outcome [37]. Since the 21 st century, fsQCA has been widely used in safety accident analysis, especially in the fields of medicine, transportation, commerce and engineering [38–41].

In terms of content, it focuses on the analysis of the complex causal relationship between the accident and various factors, reveals the path and mechanism of the accident, and points out the key causes of the accident. The outbreak mechanism and governance strategy of coal mine safety accidents, the occurrence mechanism of construction safety accidents, the risk of traffic operation, public opinion and public health events are the focus of scholars’ research [16,42–45]. In contrast, the application of fsQCA in the field of analyzing tourism accidents started late, and the number of literatures is not large, which indicates that fsQCA can be used to analyze tourism accidents. The team from the School of Tourism, China Huaqiao University, used fsQCA to analyze the safety accidents of various types of tourism, such as ice and snow tourism, outdoor sports, and night tourism [46–48]. It is an important institution for the study of tourism safety accident configuration causes, and has certain influence. Methodologically, scholars have combined fsQCA with methods such as PLS-SEM, Bayesian Network, and Meta-analysis, in conjunction with relevant theoretical frameworks from other disciplines [49–51]. However, the number of literatures combined with other methods is still small and still in the exploratory stage.

This study chose fsQCA to analyze the configurational characteristics of safety accidents in Tibet’s sports tourism for three reasons: first, the factors influencing sports tourism safety accidents are complex, and fsQCA excels in analyzing multiple, concurrent causal relationships; second, fsQCA is suitable for analyzing smaller sample sizes, which aligns with the limited number of cases available for Tibet’s sports tourism safety accidents; third, the use of SCM to categorize safety factors makes fsQCA an appropriate tool for analyzing the complex interactions of multiple conditions contributing to accidents.

3.1.1. Type distribution characteristics.

From 2010 to 2025, the types of sports tourism safety accidents in Tibet mainly include seven categories: camping, cycling, self-driving, hiking, climbing, cruise ships and adventure (Table 1). Among them, hiking and cycling accidents are significantly more common than other types. This might be due to the fact that routes such as the Sichuan-Tibet Line, the Yunnan-Tibet Line, the Qinghai-Tibet Line, and the Xinjiang-Tibet Line are increasingly attracting the attention of sports tourism tourists. In addition, compared with others, these two types of sports tourism activities have higher requirements for tourists’ physical fitness and a greater risk of safety accidents in high-altitude areas.

Download:

• PNG
  larger image
• TIFF
  original image

Table 1. Distribution of types of sports tourism safety accidents in Tibet (2010-2025).

https://doi.org/10.1371/journal.pone.0334226.t001

3.1.2. Spatio-temporal distribution characteristics.

The temporal distribution of sports tourism safety accidents in Tibet is shown in the Fig 2. Overall, the occurrence time of sports tourism safety accidents in Tibet varies significantly by season. The number of accidents is the highest in July, with only three months having no accidents. This indicates that sports tourism has gradually become a part of People’s Daily lives, not just limited to the peak tourist season. Sports tourism safety accidents in Tibet mainly occur from May to July. This is mainly because during this period, the number of tourists is relatively large and the base of tourists is large, making safety incidents prone to occur [52].

Download:

• PNG
  larger image
• TIFF
  original image

Fig 2. Monthly distribution map of sports tourism safety incidents in the Tibet Autonomous Region (2010-2025).

https://doi.org/10.1371/journal.pone.0334226.g002

The spatial distribution of sports tourism safety accidents in Tibet from 2010 to 2025 is shown in Fig 3. Overall, accidents have occurred in all prefecture-level cities and regions, indicating that sports tourism safety accidents in Tibet have shown a trend of frequent occurrence in many places. Among them, the accident distribution in Nyingchi City and Ngari Prefecture is relatively concentrated. This is mainly because these two places are the main destinations for hiking and cycling tourism in Tibet, such as “Tongmai Natural Danger”, Mount Kailash, and Pangong Lake, where the number of tourists is larger and the accident rate is higher. This conclusion corroborates the one drawn in the type distribution characteristics mentioned earlier.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 3. Distribution map of areas where sports tourism safety incidents occurred in the Tibet Autonomous Region (2010-2025).

https://doi.org/10.1371/journal.pone.0334226.g003

3.2.1. Identification of influencing factors.

In order to comprehensively extract the causative variables of sports tourism safety accidents in Tibet, this study used the fishbone diagram method to conduct open-ended factor identification of sports tourism accident reports in Tibet. Starting from the particularity of tourist destinations and tourism activities, this article focuses the core of identifying influencing factors on high-risk tourism activities in extreme environments. Sort out the problems according to “people, events, time, environment and objects”, and determine the four core factors of personal factors, environmental factors, equipment factors and management factors as the main framework for analysis. Then, read the accident report sentence by sentence, extract all the phrases directly related to the accident. In the process of identifying influencing factors through accident reports, inter-coder checks were conducted. Finally we obtain 17 high-reliability factors to form the original entry pool (Table 2) for subsequent classification.

Download:

• PNG
  larger image
• TIFF
  original image

Table 2. Pool of influencing factors.

https://doi.org/10.1371/journal.pone.0334226.t002

3.2.2. Classification of influencing factors.

The SCM posits that accidents within an organization are caused by factors at four levels (four slices of cheese), including organisational infuences, supervision, precondition, and specifie acts [18] (Fig 4). Based on the factors extracted from the fishbone diagram, this paper hierarchically classifies 17 factors with the aid of SCM. In the “Organisational Infuences” layer, “organizational professionalism” is summarized. In the “Supervision” layer, “rescue level” and “management level” are summarized. The “Precondition” layer summarizes “natural environment” and “tourist vulnerability”. The “Specifie acts” layer summarizes the “visitor behavior” factors, which are specifically divided into “action class behaviors” and “decision-making behaviors”.

Download:

• PNG
  larger image
• TIFF
  original image

Fig 4. The SCM of sports tourism accidents in the Tibet Autonomous Region of China.

https://doi.org/10.1371/journal.pone.0334226.g004

1. (1) Organisational infuences

A travel team constitutes a temporary organization comprising all participants and a tour leader, where the tour leader holds decision-making authority over the team [53]. At the organizational level, the secondary indicator “organizational professionalism” is established based on the nature of sports tourism groups, indicating the lack of organizational support during activities, informal organizational structures or affiliations, and insufficient professional staffing within the organization.

1. (2) Supervision

At the management level, two secondary indicators are defined: “rescue capabilities” and “management standards”. Rescue capabilities reflect organizers’ inadequate risk response capacities, including deficient emergency response systems from scenic areas’ authorities or local governments and insufficient post-accident handling due to negligence [54,55]. Management standards highlight insufficient macro-level regulation and oversight of the sports tourism industry by governments. Many sports tourism enterprises—such as travel agencies and scenic areas—exhibit poor operational practices, outdated operational philosophies, and inadequate risk forewarning systems or infrastructure maintenance [34].

1. (3) Precondition

This tier accounts for Tibet’s unique environmental context through two indicators: “natural environment” and “tourist vulnerability”. Natural environment refers to specific environmental conditions that may harm participants or cause losses [56]. Tourist vulnerability encompasses participants’ physical fitness, sports tourism experience and skills, and susceptibility to altitude sickness [57,58].

1. (4) Specifie acts

This tier focuses on tourist actions during sports tourism activities, including: inadequate equipment preparation prior to travel; poor decision-making due to herd mentality or weak safety awareness; risk-taking behaviors during activities, such as disregarding the tour leader’s instructions, refusing to cooperate with staff, or improper operational practices [59,60].

3.3.1. Conditional variables.

According to the analysis results of the Fishbone diagram of the accident case, the SCM constructed in the previous paper is extended, quantified and operationalized, and the specific condition variables are obtained (Table 3). There are generally three calibration methods for QCA: direct assignment, direct calibration method and indirect calibration method. The assignment of variables in fsQCA is a crucial step, determined by researchers based on their experience and relevant theoretical judgments. This process relies on external knowledge rather than the inherent attributes of the data itself [61]. Therefore, this study makes use of the existing theoretical and empirical evidence, combines the assignment suggestions of relevant experts, adopts the direct assignment method, and strictly follows the principle of “the guidance of theory and/or the relevance of practice” proposed by Ragin [62].

Download:

• PNG
  larger image
• TIFF
  original image

Table 3. Condition variables.

https://doi.org/10.1371/journal.pone.0334226.t003

In fsQCA, membership scores range from 0, indicating complete non-membership, to values between 0 and 0.5, signifying fuzzy non-membership. The midpoint, 0.5, represents the crossover point of the result set. Values between 0.5 and 1 indicate fuzzy membership, while a membership score of 1 denotes complete membership [22]. The variables RL, GL, NE, TV, and AB draw on the four-scale Likert scale and are assigned values using the four-valued fuzzy set (0, 0.33, 0.67, 1) [71]. The variable OP is assigned a value using a three-valued fuzzy set, with three score levels of 0, 0.75, and 1 selected to avoid intermediate value situations [72]. The variable DB indicates whether tourists make dangerous decisions and is a clear condition that is essentially dichotomous [73]. Therefore, when the condition exists, it is assigned a value of 1; when it does not exist, it is assigned a value of 0, which avoids information distortion and meaning distortion. To reduce the subjectivity of assignment, this study adopts double-blind and independent assignment. Two researchers respectively assigned values to each accident case based on the accident text. The two coders invited a third researcher to arbitrate the items with inconsistent assignments against the original text. Eventually, a 100% consensus was reached, forming a conditional variable dataset for fsQCA. The specific assignment rules are outlined in Table 4.

Download:

• PNG
  larger image
• TIFF
  original image

Table 4. Assignment rules.

https://doi.org/10.1371/journal.pone.0334226.t004

3.3.2. Outcome variables.

This paper uses the “Interim Provisions on the Implementation of Tourism Safety Management” (1994) and the National Tourism Administration Order No.41, “Tourism Safety Management Measures” (2016), combined with specific contents from case reports, to assign variables for casualties and missing persons in accidents as the result variables for analyzing accident influencing factors. The severity of the casualties is processed using fuzzy values, with the assignment method detailed in Table 5.

Download:

• PNG
  larger image
• TIFF
  original image

Table 5. Outcome variables and assignment rules.

https://doi.org/10.1371/journal.pone.0334226.t005

3.4.1. Analysis of the necessity of individual conditions.

Before conducting the fuzzy-set configurational analysis, it is essential to determine whether a single condition variable serves as a necessary condition for the occurrence of sports tourism safety accidents in Tibet. fsQCA 3.0 was employed to quantitatively analyze the necessity of individual conditions, and the results are presented in Table 6. In severe accidents, the consistency levels of condition variables do not exceed 0.9, indicating that they are not necessary conditions for severe accident occurrence. The analysis shows that no single condition variable is sufficient to cause severe accidents or general accidents, thus necessitating a multivariate combinatory analysis.

Download:

• PNG
  larger image
• TIFF
  original image

Table 6. Analysis of the necessity of a single condition.

https://doi.org/10.1371/journal.pone.0334226.t006

3.4.2. Analysis of combinations of condition variables.

After completing the necessity analysis, the truth tables corresponding to severe accidents and general accidents are constructed. For the analysis, fsQCA 3.0 was employed, with the consistency threshold set at 0.8, the Proportional Reduction in Inconsistency (PRI) threshold at 0.7, and the case frequency threshold at 1, considering the small sample size (15–40). Cases that do not meet the consistency standard are coded as 0 and excluded [60].

Based on the Boolean minimization rules for configurational paths, the truth tables of the two types of accidents were analyzed to generate three types of solutions: parsimonious, complex, and intermediate solutions. In this study, the intermediate solution is used as the standard configuration result, supplemented by the parsimonious solution. Conditions that appear in both the intermediate and parsimonious solutions are regarded as core conditions, while those appearing only in the intermediate solution are considered peripheral conditions. Core conditions exert a greater influence on the outcome variables than peripheral ones [74].

Using fsQCA 3.0 software, a total of nine configurational paths were identified for the influencing factors of both severe and general accidents, as shown in Table 7.

Download:

• PNG
  larger image
• TIFF
  original image

Table 7. Configuration path of influencing factors of Tibet’s sports tourism safety accidents.

https://doi.org/10.1371/journal.pone.0334226.t007

Overall, the consistency levels of both individual and overall solutions for severe and general accidents exceed the minimum threshold of 0.7. Specifically, there are three configurational paths for severe accidents with an overall solution consistency of 0.888 and a coverage of 0.425; and six paths for general accidents with an overall solution consistency of 0.876 and a coverage of 0.634. These results indicate that the nine configurations of antecedent conditions exhibit a high degree of consistency in explaining the occurrence of sports tourism safety accidents in Tibet, and can thus be considered sufficient conditions for such accidents.

3.4.3. Robustness check.

To verify the robustness of the configuration conclusion, the sufficiency threshold of the truth table was raised from 0.80 to 0.85. After the threshold was raised, the case frequency threshold remained unchanged at 1, and the overall causal path framework also remained constant. Only minor adjustments were made to the core conditions of the severe accident configurations H1a and H1b: the missing core changed from “low tourist quality” to “low management level”. However, univariate consistency shows that “low tourist quality” (0.390) is still higher than “low management level” (0.371), suggesting that this substitution does not substantially weaken the explanatory power.

What is particularly noteworthy is the change in the overall coverage rate: at the 0.80 threshold, the overall coverage rate is 0.425, while at the 0.85 threshold, it drops to 0.405. This reduction indicates that although raising the consistency threshold enhances the homogeneity within each configuration, it simultaneously excludes some “marginal cases” within the 0.80–0.85 range, thereby narrowing the accident extension that the identified configuration can explain. This indicates that although the threshold of 0.85 is highly applicable to extreme consistency cases, its explanatory power is limited for heterogeneous situations in the consistency transition zone. Therefore, this paper analyzes the configuration path with a consistency threshold of 0.8 and a frequency threshold of 1, and the conclusion obtained has high robustness.

3.5.1. Mechanism analysis of serious accidents.

The sufficiency analysis of severe accidents generates three causal configuration paths, which are classified into two major categories: natural factor dominated type and human factor dominated type.

1. (1) Natural Factor Dominated Type

Configuration H1a: In this path, the core factors are the presence of natural environment, the lack of organizational professionalism, and the absence of tourist vulnerability. The auxiliary factors are the lack of rescue level and tourist behavior. This path shows that the special natural environment and sudden natural disasters in Tibet constitute the primary driving factors of the accident. When such force majeure risk sources occur, even if the tourism organization has a high professional level and the tourists’ behavior complies with the norms, the rescue system is difficult to respond quickly due to poor spatial accessibility and limited technical equipment. The disaster impact will still break through the safety threshold with an overwhelming advantage, resulting in serious casualties. The typical case corresponding to this configuration is the avalanche hiking accident in Nyingchi City. This case was a hiking project with a large number of participants and high professionalism. Most of the tourists involved were inexperienced outdoor sports enthusiasts. The route was not fixed and they encountered natural disasters that could not be predicted.

Configuration H1b: In this path, the core factors are the presence of natural environment, the lack of organizational professionalism, and the absence of tourist vulnerability. The auxiliary factors are the presence of tourist behavior and the lack of management level. This path indicates that outdoor sports tourism activities have high environmental requirements. When the static risk at the destination remains persistently high, sports tourism tourists usually have a mentality of taking chances and taking risks. Especially the “expert confidence” of experienced tourists will induce risky and unsafe behaviors, thereby activating potential risks into actual accidents. The typical case corresponding to this configuration is the “Xizang Adventure King” accident of falling into the water. The party involved had many years of experience in glacier exploration, but due to overconfidence in the stability of the ice tongue structure, they chose to take pictures at the edge of the waterfall that had not been surveyed. Eventually, they fell into the glacier due to the instantaneous disintegration of the ice surface.

Configuration H1a and H1b show a high similarity, and both of them jointly verify that the poor natural environment is an important safety hazard of sports tourism in Tibet. However, the difference lies in that both present two typical causative patterns in the tourist behavior subsystem: “fault-free type” and “risky type”. H1a emphasizes the suddenness of disasters, and H1b highlights the “experience paradox” of professional tourists in high-risk scenarios.

1. (2) Human Factor Dominated Type

Configuration H2: In this path, the core factors are the presence of rescue level and tourist behavior, with the presence of natural environment, the lack of management level, and the absence of tourist vulnerability acting as auxiliary factors. This path indicates that in the context of high-altitude sports tourism, the natural environment itself constitutes a high baseline risk, and the unsafe behaviors of tourists, as a triggering mechanism, transform potential risks into immediate dangers. Therefore, a crucial interaction effect is formed between tourists’ unsafe behaviors and the low level of rescue. The essence of this path is a systematic failure with a triple coupling of “behavior triggering - response failure - environmental amplification”. This configuration corresponds to typical cases of rockfall accidents on the Sichuan-Tibet Highway and a landslide accident during a hiking trip in Nyingchi City.

3.5.2. Mechanism analysis of general accidents.

The sufficiency analysis of general accidents generates six causal configuration paths, which are divided into four main categories: tourist-altitude sickness type, tourist-experience type, tourist-behavior type, and institutional factors dominated type.

1. (1) Tourist-Altitude Sickness Type

Configuration L1a: In this path, the core factors are the presence of tourist vulnerability and operational behavior, with the presence of organizational professionalism, natural environment, and the absence of rescue level and management level as auxiliary factors. This path indicates that even if the external supervision and rescue system is relatively complete, individual physical fragility and operational errors can still directly trigger altitude sickness related accidents. At this point, the negative effects of the organization and the environment were compressed to a “potential but unintervened” state, and the accident was mainly attributed to the immediate coupling of the tourists’ own physiological threshold breakdowns and behavioral deviations. This configuration corresponds to the typical case of a child suffering from altitude sickness on Mount Sapu.

Configuration L1b: In this path, the core factors are the presence of tourist vulnerability and operational behavior, with the presence of organizational professionalism, natural environment, and decision-making behavior, as well as the absence of rescue level, serving as auxiliary factors. Compared with L1a, this path extends the causal chain from a single operational error to a dual absence of “cognition and operation”. Insufficient information search before travel and distorted risk assessment during the trip constitute the defect of pre-decision-making, while real-time operational errors act as triggers, turning potential high reversal into explicit accidents. This configuration corresponds to the typical cases of heatstroke on the Dongda Mountain cycling trip and a fall during a cycling trip in the Qiangtang Wilderness.

Configuration L1a and L1b exhibit a high degree of similarity, and the two paths jointly reveal the vulnerability of the “individual physiological - behavioral - cognitive” continuum in plateau sports tourism. When tourists with a relatively low threshold of their own physical fitness are simultaneously exposed to the low-oxygen environment at high altitudes and high-intensity exercise loads, operational errors and decision-making flaws will superimpose and amplify the physiological stress response, significantly increasing the probability of general accidents. The lack of professionalism of the organization increases the possibility of unsafe behavior among tourists, and the poor environmental conditions increase the possibility of accidents such as altitude sickness among tourists.

1. (2) Tourist-Experience Type

In this pathway, the presence of tourist vulnerability and the absence of organizational professionalism play core roles, while the existence of tourist decision-making behaviors and the absence of rescue capabilities, management standards, natural environmental factors, and tourist operational behaviors serve as supplementary factors. This path indicates that although the existence of professional organizations can provide standardized processes and immediate error correction mechanisms at the operational end, it cannot offset the experience gap and cognitive arrogance at the individual level of tourists. Tourists’ lack of prior knowledge about the environmental risks of the plateau and the expansion of their self-efficacy will jointly give rise to “blind confidence” type wrong decisions. Even if the organization has professional qualifications, tourists’ decision-making mistakes can still directly penetrate the organizational barrier and trigger general accidents. Therefore, the essence of this path is that “experiential vulnerability” and “cognitive distortion” are exposed in a localized and case-by-case manner through the protective net of professional organizations. This proves that in high-altitude sports tourism, the quality of tourists’ cognitive decision-making is a key influencing factor for safety outcomes, while the functional boundary of whether an organization is professional only lies at the behavioral execution level. The corrective effect on cognitive dissonance is limited. A representative case is the Mount Wula altitude sickness incident.

1. (3) Tourist-Behavior Type

Configuration L3a: In this pathway, the presence of tourist operational behaviors and the absence of natural environmental factors and tourist decision-making behaviors serve as core conditions, while the presence of organizational professionalism and the absence of rescue capabilities, management standards, and tourist vulnerability act as supplementary factors. This path indicates that in Tibet, where self-driving and cycling Tours are prevalent, tourists lacking professional and sports tourism experience are prone to accidents due to improper operation by themselves, and non-professional organizations are also unable to provide immediate corrections at the operation end. Therefore, the operational mistakes of tourists may cross the environmental threshold. Even when the natural environment of the sports tourism destination is good, general accidents still occur. A corresponding case is the Ngari No-Man’s Land Off-Road vehicle crash.

Configuration L3b: In this pathway, the presence of tourist operational behaviors and the absence of natural environmental factors and tourist decision-making behaviors remain core conditions, while the presence of organizational professionalism, management standards, tourist vulnerability, and the absence of rescue capabilities serve as supplementary factors. This path indicates that non-professional organizations and low management levels jointly weaken risk identification, hardware preparation and process control. The vulnerability of tourists further reduces the team’s sensitivity to danger signals and self-restraint ability. In this context, operational errors are no longer occasional instantaneous fluctuations but are systematically created under the dual effects of organizational failure and individual insufficiency. A representative case is the Nyingchi Road collapse incident.

Configurations L3a and L3b share high similarity: accidents occur in relatively safe natural environments due to tourist operational errors. However, unlike L1a, L3a involves tourists with higher competence and adequate management standards, emphasizing the accident’s unpredictability and contingency.

1. (4) Institutional Factors Dominated Type

Configuration L4: In this pathway, the presence of tourist decision-making behaviors and management standards, coupled with the absence of rescue capabilities and tourist operational behaviors, act as core conditions, while the presence of organizational professionalism, natural environmental factors, and the absence of tourist vulnerability serve as supplementary factors. The logical core of this path lies in the triple coupling of “system - cognition - consequence”. When the official governance system experiences a failure of early warning and a gap in emergency response, even if tourists have basic operational skills, they are very likely to make high-risk decisions due to information vacuum, thereby transforming institutional deficiencies into observable general accidents. A corresponding case is the Namtso Lake disorientation incident, where the managers failed to issue early warnings or close the area despite unseasonable snowfall rendering hiking unsafe.

4.1.1. Spatiotemporal characteristics.

Existing studies confirm significant seasonal and regional variations in sports tourism accidents, with accident rates peaking during high tourist seasons due to surges in visitor numbers and activity intensity [75]. The current study corroborates this by identifying May to July as the period with the highest accident occurrence in Tibet. Additionally, the spatial distribution analysis reveals a concentration of accidents in high-altitude regions like Nyingchi City. This aligns with the research conclusion that geographical features (such as steep terrain and hypoxic environment) attract tourists and increase risks [76]. However, the current study extends this by linking the spatiotemporal patterns to high-risk activities such as hiking and cycling, which are particularly prevalent in these regions. Unlike previous studies that primarily focused on seasonal variations, this research also highlights the concentration of accidents in specific high-altitude areas, providing a more nuanced understanding of the spatial distribution of risks [77].

4.1.2. Influencing factors.

The SCM framework used in this study categorizes influencing factors into active failures (direct triggers) and latent failures (systemic weaknesses). This classification builds upon prior research that has examined human, environmental, and material risks [78]. Identifying natural environmental hazards, such as extreme weather, and tourist behavior, as active failures, alongside organizational gaps and regulatory oversight as latent failures, provides a more nuanced understanding of accident causation. This study systematically categorizes these factors within the SCM framework, offering a clearer distinction between direct triggers and systemic weaknesses.

4.1.3. Risk coupling paths.

Unlike studies attributing accidents to single causes (e.g., “human error” [4] or “weather” [76]), this configurational analysis proves that no necessary condition alone suffices for accidents—consistent with Reason’s SCM but previously untested in plateau tourism [18]. For instance, while tourist negligence is prevalent in Path 7, its effects are neutralized by concurrent regulatory gaps (latent failure). This forces a paradigm shift from factor-centric to pathway-centric risk governance.

A key contribution of this study is the identification of nine configuration paths leading to severe and secondary accidents through fsQCA. This approach highlights the nonlinear interplay between systemic deficiencies (latent failures) and situational triggers (active failures). Previous research has often focused on individual factors in isolation, whereas the current study demonstrates that accidents typically result from the coupling of multiple factors. For instance, the configuration paths for severe accidents, such as “environmental hazard + equipment failure + poor rescue coordination,” and secondary accidents, such as “tourist negligence + weak supervision + communication breakdown,” underscore the complexity of accident causation [79]. This finding is consistent with the study which argued that a combination of factors, rather than a single cause, is typically responsible for accidents [80]. The current study’s emphasis on the interaction between organizational, environmental, and behavioral factors provides a more holistic view of risk management. Unlike previous studies that may have overlooked the interdependencies between different factors, this research highlights the importance of considering multiple pathways to accident occurrence.

4.2.1. Theoretical implications.

First, Theoretical Extension of the SCM: This study innovatively applies the SCM to high-altitude sports tourism safety, expanding its theoretical scope beyond traditional industries (e.g., aviation, healthcare) to address complex environmental and human-factor interactions in extreme tourism contexts. The refined framework integrates multi-layered defense mechanisms (natural, behavioral, managerial, and rescue layers), offering a systematic lens for analyzing nonlinear risk couplings in fragile ecosystems.

Secondly, Methodological Contribution of fsQCA in Tourism Safety Research: By employing fuzzy-set Qualitative Comparative Analysis (fsQCA), this research advances methodological approaches in tourism safety studies. It demonstrates how fsQCA effectively uncovers multifactorial causal configurations driving accidents, addressing limitations of linear regression models in capturing complexity. This approach sets a precedent for small-sample, case-based analyses in understudied high-risk tourism regions.

Thirdly, High-Altitude-Specific Risk Coupling in Sports Tourism: This study pioneers the identification of plateau-environment-driven risk configurations unique to high-altitude sports tourism. By analyzing the interplay of hypoxia, extreme weather, and glacial terrain with tourist behavior, it reveals how altitude-exacerbated vulnerabilities amplify risks in adventure activities like trekking or cycling. These findings fill a critical gap in existing literature, which often overlooks the nonlinear synergies between high-altitude ecosystems and sports tourism dynamics, offering a theoretical foundation for future research on extreme-environment tourism safety.

4.2.2. Practical implications.

First, this study proposes a multi-layered risk prevention system for sports tourism in high-altitude regions. The system includes natural defense, behavioral defense, management defense, and rescue defense layers, each targeting specific risk factors identified in the study. This comprehensive approach provides practical guidance for policymakers, tourism managers, and safety professionals to develop targeted interventions that address the multifaceted nature of sports tourism accidents. Implementing such a system can significantly enhance safety and reduce the incidence of accidents in high-altitude sports tourism destinations.

Secondly, the study highlights the critical role of visitor behavior and decision-making in accident causation. Therefore, enhancing visitor education and awareness through targeted programs and campaigns is essential. These initiatives should focus on educating tourists about the risks associated with high-altitude sports tourism, providing them with necessary skills and knowledge to make informed decisions, and promoting responsible behavior. By improving visitor preparedness and awareness, the likelihood of accidents can be significantly reduced, thereby enhancing overall safety in sports tourism environments.

Lastly, the findings emphasize the importance of robust emergency response and rescue capabilities in mitigating the severity of accidents. This includes upgrading rescue equipment, deploying advanced technologies such as drones and helicopters for rapid response, and establishing comprehensive emergency communication systems. Additionally, forming collaborative rescue alliances involving government agencies, military, and local communities can enhance the efficiency and effectiveness of rescue operations. Strengthening these capabilities is crucial for minimizing the impact of accidents and ensuring the safety of tourists in high-altitude regions.

4.2.3. Implications for other high-altitude contexts.

Although this study focuses on the Tibet Autonomous Region, the integrated SCM-fsQCA framework and identified risk configurations offer transferable insights for similar high-altitude or extreme-environment destinations (e.g., the Andes, Alps, or Himalayas). The core risk mechanisms—such as hypoxia-driven vulnerabilities, nonlinear couplings between environmental hazards and human/organizational factors, and the criticality of rescue coordination—are universally relevant in high-altitude tourism settings. For instance, the “environmental hazard + equipment failure + poor rescue coordination” path (Path 3 in fsQCA results) mirrors accident patterns observed in Alpine mountaineering [2,34]. The proposed four-layer defense system (natural, behavioral, managerial, rescue) can be adapted to other regions by calibrating location-specific factors (e.g., glacial terrain in the Andes or avalanche risks in the Alps). Future studies should validate these findings in diverse geographic contexts to establish a generalized high-altitude safety management paradigm.