COVID-19 infection risk assessment and management at the Tokyo 2020 Olympic and Paralympic Games: A scoping review

The Tokyo 2020 Olympic and Paralympic Games was one of the largest international mass-gathering events held after the beginning of coronavirus disease 2019 (COVID-19) pandemic. In this scoping review, we extracted papers discussing COVID-19 risk assessment or management at the Tokyo 2020 Games to determine the nature of studies that were conducted. Among the 75 papers obtained from two search engines (PubMed and ScienceDirect) and four papers collected from hand-searches, 30 papers were extracted. Only eight papers performed both COVID-19 prior risk assessment and quantitative evaluation of effectiveness measures, highlighting the importance of rapid, solution-focused risk assessment. Furthermore, this review revealed that the findings regarding the spread of COVID-19 infection to citizens in the host country were inconsistent depending on the assessment methods and that assessments of the spread of infection outside the host country were lacking.


Introduction
Infection risk assessment and management are important issues for mass-gathering events such as sports, music concerts, festivals, and religious events because these events inherently involve gatherings of an unspecified large number of people [1]. As the largest international mass-gathering events, the Olympic and Paralympic Games have previously been threatened by infectious disease outbreaks, such as the Zika virus at the Rio de Janeiro 2016 Games [2]. After the beginning of the coronavirus disease 2019 (COVID- 19) pandemic, the importance of infection risk assessment and management was recognized [3], and risk assessment and management efforts for mass-gathering events have changed drastically. The Tokyo 2020 Olympic and Paralympic Games (hereafter referred to as Tokyo 2020 Games), which were scheduled to start on July 24, 2020, were postponed until July 23, 2021. In principle, the Tokyo 2020 Games were held without spectators. In addition, a set of infection control efforts together known as the "bubble strategy" was thoroughly implemented among athletes, staff members, and other game officials; these measures included isolation from the citizens in Japan and testing at travel and during the Tokyo 2020 Games [4]. An analysis of the infection risk-assessment and management efforts at the Tokyo 2020 Games can yield perspectives for holding various international mass-gathering events, including the Beijing 2022 Games. However, no review has comprehensively outlined the studies on COVID-19 infection risk assessment and management that were conducted prior to or after the Tokyo 2020 Games. Such a review can be expected to provide scientific knowledge on the preparation for and validation of infection risk during mass-gathering events.
Therefore, we conducted a scoping review on the risk assessment and management of COVID-19 at the Tokyo 2020 Games. Among the papers identified in a systematic literature search, we selected papers that aimed to discuss COVID-19 risk assessment or management at the Tokyo 2020 Games and clarified the types of studies that were conducted on the basis of the contents of these papers. In particular, we focused on how prior risk assessments and quantitative evaluations of measure effectiveness were conducted and summarized the lessons and perspectives for studies on infection risk assessment and management of mass-gathering events.
The flow of the screening process used to identify the papers covered in this scoping review is shown in Fig. 1.
Papers were extracted using the PubMed and ScienceDirect search engines on September 16, 2022. No restrictions were placed on the type of paper (e.g., original article, review, or letter), publication date, and language. In PubMed, 68 papers were extracted by using the keywords "(tokyo AND (olympic* OR paralympic*) AND (covid* OR sars* OR coronavirus))" for the title/abstract. In ScienceDirect, 19 papers were extracted using the keywords "(tokyo AND (olympic OR paralympic) AND (covid OR sars OR coronavirus))" for Title, abstract or author-specified keywords (wildcards were not necessary in the ScienceDirect search). In total, 75 papers were identified, after excluding 12 duplicate papers. All the papers were written in English.
The first screening was conducted on 63 papers, after excluding 12 papers without abstracts. The first screening was based on the criterion of identifying papers that aimed to discuss COVID-19 risk assessment or management at the Tokyo 2020 Games. Both quantitative and qualitative studies were included in this assessment. For each paper, two experts (M.M. and K.F.) independently judged whether they met the criteria by reading the titles and abstracts. The kappa coefficient of agreement was 0.72. If the judgment did not agree between the two experts, a decision was made after consultation with a third expert (W.N.). In total, 34 papers were excluded because they did not meet the inclusion criteria.
Before the second screening, four papers [5][6][7][8] that were not included in the above search but were of high importance for discussion in this review were added by hand-searches. One of these papers was an official review of the Tokyo 2020 Games [8]. This report was written in Japanese but was included because it will be published in English in the future (https://library.olympics.com/), and it is important as an official report. The full texts of the 33 papers were obtained and their full texts were read and evaluated on the basis of the above criteria to determine their suitability for inclusion in the scoping review. The following items were extracted from the papers that met the criteria.
• Targets for the risk assessment or management of COVID-19 (multiple choices allowed): Not available, athletes and staff members, citizens in Japan, citizens in other countries, others (the details were specified for others, as for the following items).
• The COVID-19 risk-assessment approach (the risk assessment here refers to the assessment of the infection status such as the number of infected individuals, and does not include the evaluation of measure effectiveness) (multiple choices allowed): Not available, epidemiology (including studies involving human participants and virus genomes extracted from human participants as well as reports of simple statistical data on the infection status, such as the number of infected individuals), model simulation (including quantitative microbial risk assessment using data such as virus concentration), others.
• Whether the risk assessment was prior or ex-post (here, prior risk assessments refer to risk assessments based on information before the Tokyo 2020 Games rather than risk assessments made prior to the Games, and were often performed to provide knowledge that would contribute to the preparedness of the Games. On the other hand, ex-post risk assessment refers to assessments conducted using information during or after the Tokyo 2020 Games.) (Multiple choices are allowed.): Not available, prior assessment, ex-post assessment • Contents related to risk management (multiple choices allowed): Not available, testing (including testing and quarantine in travel and in the Tokyo 2020 Olympic and Paralympic Village [hereinafter, "Village"]), vaccination, basic measures (including ventilation, masks, physical distance, hand-washing, disinfection, and regulation of human flow), comprehensive measures (representing a set of various measures such as overall measures used in the Tokyo 2020 Games), others.
• Evaluation of the effectiveness of measures related to COVID-19 (single choice allowed): Not available, quantitative evaluation, qualitative evaluation.
Criteria-based judgments and the extraction of the above items were made independently by two experts for each paper: M.M. was in charge of all papers and K.F. and W.N. were in charge of half each. When the judgments or extraction made by the two experts did not agree, a third expert was consulted. Ultimately, 30 papers remained .
Furthermore, M.M. summarized the submission date, publication date, country of the corresponding author, and summary of the paper, which were confirmed by another expert (K.F. or W.N.). M.M. then categorized the pre-and post-risk assessments based on the content of the paper, which was further confirmed by two experts (K.F. and W.N.).
For papers that performed prior risk assessment, M.M. also investigated whether the paper was published as a preprint. In addition to searches based on the text information in the paper, M.M. searched for the first author's name in arXiv, bioRxiv, medRxiv, SSRN, and PeerJ PrePrints on October 14, 2022. Table 1 summarizes the studies identified in this review and their characteristics. Fig. 2 shows the distribution of the papers in relation to submission or publication before the Tokyo 2020 Games, the corresponding author's country, the targets of risk assessment or management, the risk-assessment approach, prior or ex-post risk assessment, the content of risk management, and the evaluation of measure effectiveness.  Of the 30 papers, eight were submitted before the Tokyo 2020 Games and six were published. In assessments of the corresponding author's country, Japan was the most frequent (19), followed by China (4) and the United Kingdom (3). The most common target groups for risk assessment or management were athletes and staff members (18) and citizens in Japan (15), whereas citizens in other countries (5) and spectators in stadiums (2) were less frequently evaluated. Regarding the risk-assessment approach, 15 papers used epidemiological assessments, 12 performed model simulations, and two conducted wastewater-based epidemiology analyses in the Village. The number of papers that performed prior and ex-post risk assessments were 10 and 17, respectively. Regarding the contents of risk management, 10 papers focused on tests, seven on comprehensive measures, six each on vaccines and basic measures, and two on the holding of the Tokyo 2020 Games without spectators. The effectiveness of the measures was quantitatively evaluated in 11 papers and qualitatively evaluated in 13 papers.

Results
Prior risk assessments were classified into four groups ( Table 2): 1) risk assessment for athletes and staff members [6,7,16], 2) risk assessment for spectators in stadiums [8,10], 3) evaluation of the number of infection cases before the Tokyo 2020 Games [9], and 4) evaluation of the potential of the Tokyo 2020 Games to cause spread of the infection among citizens in Japan [12,13,18,29]. Of these, eight papers also performed quantitative evaluations of measure effectiveness, including the effectiveness of tests and vaccines in reducing the infection risk mainly among athletes and staff members [6,7,16], the effectiveness of basic measures in the stadium in reducing the infection risk among spectators [8,10], and the effectiveness of testing, vaccines, basic measures, and holding non-spectator events in controlling the spread of infection among citizens in Japan [12,18,29].
The ex-post risk assessments were classified into three groups: 1) evaluation of the number of infection cases or tests among athletes and staff members [8,15,[19][20][21]24,25,28,32], 2) evaluation of whether the Tokyo 2020 Games caused the spread of infection among citizens in Japan [5,8,17,22,26,31,33], and 3) evaluation of the external transmission of substrains originating in Japan by using genome sequence analysis of severe acute respiratory syndrome coronavirus 2 (SARS-CoV-2) [23,30]. Of these, four studies also performed quantitative evaluations of measure effectiveness [8,17,25,26]. Three papers evaluated the effectiveness of vaccines, basic measures, comprehensive measures, and holding non-spectator events in reducing the spread of infection among citizens in Japan [8,17,26], and one study evaluated the effectiveness of tests in reducing infection risk among athletes and staff members [25].

Discussion
In this study, we conducted a scoping review to analyze the types of studies conducted on COVID-19 risk assessment and management for the Tokyo 2020 Games, with a particular focus on whether a prior risk assessment was conducted and whether the measures' effectiveness was quantitatively evaluated.
As a result, 30 papers were identified. Ten studies performed prior risk assessments, of which only eight also quantitatively evaluated measure effectiveness. Prior risk assessments mainly focused on the infection risk among athletes and staff members, citizens in Japan, and spectators in stadiums during the Tokyo 2020 Games. Only four of the prior risk assessments were published in journals and only five, including the preprints, were made publicly available. Prior risk assessment is essential in public health crises, such as the COVID-19 pandemic, when rapid decision-making is required, as in the case of the Tokyo 2020 Games. Furthermore, wide discussion of these results is also important. Therefore, it is necessary to raise awareness of the importance of studies that enable rapid prior risk assessment and to provide a platform for open discussion of the findings. Assuming that sufficient attention is paid to the use of the findings obtained [34], preprint servers can be expected to be useful in accelerating discussions.
Of the 17 papers that included ex-post risk assessments, nine were about the number of infection cases or tests among athletes or staff members. Most of them were descriptive reports of the infection cases or the number of tests, but one study quantitatively evaluated the cost-effectiveness analysis on testing strategy by using a model simulation [25]. This study aimed to assess the risk among athletes and staff members at the Beijing 2022 Games based on findings from the Tokyo 2020 Games. In addition, two papers reported wastewater-based epidemiological surveillance in the Village [21,32]. The authors of these papers noted that SARS-CoV-2 was detected in wastewater samples discharged from the Village prior to the positivity of individual testing to humans, and that the combination of wastewater-based epidemiology data and other monitoring results, such as individual testing in humans, contributed to enhanced infection control measures in the Village. As a new approach to infection risk assessment and management, wastewaterbased epidemiological surveillance is expected to be used in future Olympic and Paralympic Games and other mass-gathering events.
Among the studies that conducted ex-post risk assessments, seven focused on discussing whether infection risk had spread to citizens in Japan. Among them, three papers suggested that the infection had spread [5,22,31], one mentioned that the spread was neither dominant nor negligible [26], two considered that the infection had not spread [8,33], and the remaining paper supported Fig. 2. Distribution of paper types. The number to the right of the bar represents the number of papers. RA: risk assessment; RM: risk management. a One paper had corresponding authors from two countries, namely, Qatar and the United Kingdom. b Epidemiology included studies involving human participants and genomes extracted from human participants. c Testing included quarantine; basic measures included ventilation, masks, physical distance, hand-washing, disinfection, and regulation of human flow; comprehensive measures represented a set of various measures, such as overall measures used in the Tokyo 2020 Games. d Papers that performed both prior risk assessment and quantitative evaluation of measure effectiveness. e Papers that performed both ex-post risk assessment and quantitative evaluation of measure effectiveness. f Papers that performed prior risk assessment and were submitted before the Tokyo 2020 Games. g Papers that performed prior risk assessment and were published before the Tokyo 2020 Games.

Table 2
Classification of contents in prior or ex-post risk assessments.

Prior risk assessment
Ex-post risk assessment the holding of the Games without spectators [17]. Thus, the findings were inconsistent among studies. Of the two papers that mentioned no infection spread, one was an official report of the Tokyo 2020 Games, based on the fact that the effective reproduction number of COVID-19 in Tokyo had decreased since before the Tokyo 2020 Games. However, there was no evaluation of what would have happened if the Games had not taken place (i.e., a counterfactual setting) [8]. Another study was based on a comparative analysis of the number of infections among the prefectures [33]. However, all three papers that suggested the spread of infection used synthetic control methods as counterfactual settings. The difference in the presence of spread of infection depended on the analytical methods used in the counterfactual settings. Therefore, when discussing the presence or absence of infection spread among citizens in the host country after mass-gathering events, the validity and reliability of the counterfactual settings should be discussed in more depth. For example, it would be useful to validate the analysis by assessing the infection risk at a variety of mass-gathering events, or during periods when no mass-gathering events are being held. Two studies used SARS-CoV-2 genome sequence analysis, and both showed that substrains originating from Japan were transmitted overseas [23,30]. However, there was disagreement regarding these views. One report suggested that the number of infectious cases was limited [23] and another pointed out that the overall picture might not be fully understood [30]. No study evaluated the spread of infection outside Japan due to the Tokyo 2020 Games by using methods other than genome sequence analysis (e.g., epidemiology or model simulations based on the number of infected individuals).
This review provides lessons and perspectives on infection risk assessment and management in future mass-gathering events. First, very few studies conducted both prior risk assessments and quantitative evaluations of measure effectiveness. Acceleration of solution-focused risk assessments [35] and open discussions on the implementation of the measures are important in this regard. For the Tokyo 2020 Games, a research team, Mass gathering Risk COntrol and Communication (MARCO [36]), performed a prior risk assessment of COVID-19 among spectators at the opening ceremony at the Tokyo 2020 Games and quantitative evaluation of the effectiveness of the measures in risk reduction [10] They also evaluated the effectiveness of testing systems during travel and in the Village in terms of risk reduction among athletes and staff members [6,7]. These assessments were used to design a venue for public viewing for the Tokyo 2020 Games (although public viewing ended up being canceled due to concerns about the spread of infection regardless of the assessment results) and to implement wastewater-based epidemiological surveillance in the Village [21,32]. For future massgathering events, the findings obtained from prior infection risk assessment can be expected to be implemented into the actual event design. Second, the studies evaluating whether the Tokyo 2020 Games would cause the spread of infection among citizens in Japan did not yield consistent results. For model simulations, the validity of the models can be further evaluated using the observed values during the validation period. In epidemiological studies, the validity and reliability of counterfactual settings needs to be improved from multiple perspectives. Third, studies on the infection risk outside the host country are extremely limited, necessitating deeper evaluation of this point using methods other than viral genome sequence analyses.
This study has some limitations. First, only two search engines were used (PubMed and ScienceDirect). Second, this review was based on searches conducted on September 16, 2022, and papers published after that date were not included, except for a few papers found by manual searches. Third, the papers identified in this study were written only in English, with the exception of the official report on the Tokyo 2020 Games. This may have resulted in publication bias. Fourth, preprints were identified solely on the basis of relevant descriptions in the paper and searches in five preprint servers (arXiv, bioRxiv, medRxiv, SSRN, and PeerJ PrePrints). The findings for other servers (e.g., servers of the authors' institutions) could not be verified. Therefore, we cannot rule out the possibility that the number of preprints published before the Tokyo 2020 Games was underestimated.

Conclusion
This scoping review is the first to characterize the studies on COVID-19 risk assessment and management at the Tokyo 2020 Games, especially focusing on perspectives such as prior risk assessment and quantitative evaluation of measure effectiveness. This review yielded lessons and perspectives of studies on infection risk assessment and the management of mass-gathering events. Our findings are as follows.
• Of the 30 papers extracted, 10 and 17 papers performed COVID-19 prior and ex-post risk assessments, respectively.
• Prior risk assessments focused on the infection risk among athletes and staff members, citizens in Japan, and spectators in stadiums during the Tokyo 2020 Games.
• Ex-post risk assessments included evaluation of the number of infection cases or tests among athletes and staff members, discussion on the extent of infection risk to citizens in Japan, and examination of the transmission of Japan-originated substrains outside Japan using genome sequence analysis.
• Very few studies conducted both prior risk assessments and quantitative evaluations of measure effectiveness, highlighting the importance of rapid solution-focused risk assessment.
• Results regarding the spread of COVID-19 infection to citizens in the host country were inconsistent among the assessment methods. This warrants further evaluation of the validity of the model simulations as well as the validity and reliability of counterfactual setting for epidemiological studies.
• Studies on the infection risk outside the host country are limited, implying the needs for more in-depth evaluation of transmission abroad.

Ethical approval
Not required.

Declaration of Competing Interest
Countermeasures Liaison Council jointly established by the Nippon Professional Baseball Organization and the Japan Professional Football League as experts without any reward. F.K., W.N., and T.Y. were advisors to the Japan National Stadium. W.N. and T.Y. are advisors to Japan Professional Football League. Masaaki.K. received research funding from SHIONOGI & CO., LTD and AdvanSentinel, Inc. and patent royalties from SHIONOGI & CO., LTD. The findings and conclusions of this article are solely the responsibility of the authors and do not represent the official views of any institution.