1Department of Social Sciences and Policy Studies, The Education University of Hong Kong, Tai Po, Hong Kong; 2Department of Information Systems, Business Statistics and Operations Management, The Hong Kong University of Science and Technology, Clear Water Bay, Hong Kong

Correspondence: Amanda MY Chu, Department of Social Sciences and Policy Studies, The Education University of Hong Kong, Tai Po, Hong Kong, Email [email protected]

Background: The COVID-19 pandemic presents the possibility of future large-scale infectious disease outbreaks. In response, we conducted a systematic review of COVID-19 pandemic risk assessment to provide insights into countries’ pandemic surveillance and preparedness for potential pandemic events in the post-COVID-19 era.
Objective: We aim to systematically identify relevant articles and synthesize pandemic risk assessment findings to facilitate government officials and public health experts in crisis planning.
Methods: This study followed the Preferred Reporting Items for Systematic Review and Meta-Analysis (PRISMA) guidelines and included over 620,000 records from the World Health Organization COVID-19 Research Database. Articles related to pandemic risk assessment were identified based on a set of inclusion and exclusion criteria. Relevant articles were characterized based on study location, variable types, data-visualization techniques, research objectives, and methodologies. Findings were presented using tables and charts.
Results: Sixty-two articles satisfying both the inclusion and exclusion criteria were identified. Among the articles, 32.3% focused on local areas, while another 32.3% had a global coverage. Epidemic data were the most commonly used variables (74.2% of articles), with over half of them (51.6%) employing two or more variable types. The research objectives covered various aspects of the COVID-19 pandemic, with risk exposure assessment and identification of risk factors being the most common theme (35.5%). No dominant research methodology for risk assessment emerged from these articles.
Conclusion: Our synthesized findings support proactive planning and development of prevention and control measures in anticipation of future public health threats.

Keywords: meta-analysis, coronavirus, pandemic risk management, WHO COVID-19 research database, data visualization

Introduction

The outbreak in 2019 of the novel coronavirus disease (COVID-19), which the World Health Organization (WHO) officially declared a global pandemic on 11 March 2020,1 is currently the most detrimental worldwide public health event of the twenty-first century. The disease’s rapid transmission not only has imposed tremendous pressures on the public health systems, but it also has severely disrupted the financial markets,2,3 our society and the global economy,4 and our environment.5 Furthermore, this threatening pandemic caused drastic harms to people’s mental health. People affected by COVID-19 showed relatively higher rates of adverse psychiatric outcomes like anxiety, depression, stress, and psychological distress.6 Another gloomy impact of COVID-19 to the society was that misinformation and fake news about transmission, prevention, and medical treatment7,8 were spread within and across online communities broadly and swiftly, causing the prevalence of incorrect knowledge about COVID-19. These wide-scale deadly effects made research works relating to COVID-19 pandemic risk assessment so important that governments, public health professionals, and scientists could gain insights from research findings for disease prevention and control strategies.

Since its first appearance, COVID-19 has been a hot topic of research in many fields, and especially in health-related disciplines. Even now, the research enthusiasm for COVID-19 has not abated, because the virus has continued to transform itself into new variants9 and has caused successive waves of large-scale transmission with exponential increases in new infections globally for the past 3 years.

As we have stepped into the post-COVID-19 era, enormous volumes of extant research studies on the various aspects of the COVID-19 pandemic have been published. During COVID-19 pandemic, people often used online social media platform to search for information on recent development, communicate their views, and express their feelings. The analysis conducted by Chandrasekaran et al10 on COVID-19–related tweets from Twitter data indicated that the contents could be broadly classified into 10 different themes. Four commonly concerned themes were spread and growth (15.45%), treatment and recovery (13.14%), impact on the health-care sector (11.40%), and government response to the pandemic (11.19%). In light of the above observations, it is noteworthy to work out some statistics describing the coverage distribution of the current COVID-19 pandemic risk assessments such that their diversity and applicability could be demonstrated to the concerned parties for acquiring a more comprehensive and detailed understanding on how to prevent, manage, treat, and address the issues. Nevertheless, staying vigilant to the spread of infectious disease and getting more well prepared are essential. Thus, the objectives of our study were to provide a systematic review of the COVID-19 pandemic risk assessment.

The articles in this review were selected from the World Health Organization (WHO) COVID-19 Research Database,11 which is a centralized database that pools publications from different health-care research databases such as Ovid, PubMed, Scopus, Web of Science, and others. At the time that we searched the articles, there were already more than 620,000 records, and the size of the pool is continually growing because it is updated weekly and new publications are added regularly. The WHO was recognized for its outstanding work in building the COVID-19 Research Database and for the excellence of its content.12 The relevant articles were characterized using the following six research questions (RQs).

RQ1: Study Location

RQ2: Types of Variables Used

RQ3: Availability of the Materials Used to Generate Research Outcomes

RQ4: Use of Data-Visualization Techniques

RQ5: Research Objectives

RQ6: Research Methodologies

This paper is subsequently organized as follows: In the Methods section, we describe thoroughly how the final eligible list of articles was selected in order to provide the best possible answers to our research questions. In the Results section, we develop our classification framework for the research questions and present the summary statistics of the eligible articles, with the aid of tables and charts. In the Discussion section, we provide our key findings, along with some recommendations for policymakers and health-care experts and researchers to deal with the potential for any future outbreak of disease. Finally, the Conclusions section gives a brief recap of the key conclusions that can be drawn from our findings.

By synthesizing insightful findings with the help of tables and charts, this study aids policymakers, health-care experts, and researchers in creating preparedness and surveillance efforts for possible new waves of COVID-19 and/or the emergence of new infectious diseases in the future.

Methods Overview and Selection Process

We conducted a systematic review of literature reviews on COVID-19 pandemic risk assessment sourced from the World Health Organization (WHO) databases, according to the Preferred Reporting Items for Systematic review and Meta-Analysis (PRISMA) guidelines.13 Two researchers worked independently to select the final eligible articles in this review. First, one of them used the electronic search engine available in the WHO database to generate a list of potentially eligible articles. Next, each article in the potentially eligible list was retrieved either directly from the WHO database or from the journal website on which that article was published. Finally, the other researcher manually screened each article to ensure that those in the final eligible list satisfied our inclusion criteria and did not meet our exclusion criteria. All disagreements between the two researchers over the eligibility of particular articles were resolved through discussion with a third researcher.

Information Source and Search Strategy

We identified the relevant articles for this review by searching the World Health Organization (WHO) COVID-19 Research Database from its inception to 12 July 2022. This electronic database is freely and publicly accessible online. It searches, on a frequent basis, a vast number of popular databases to obtain current articles reporting global research on the coronavirus disease (COVID-19). During the time that we were searching the articles, the three largest sources of articles, in terms of the quantity in the WHO COVID-19 Research Database, were MEDLINE, Scopus, and Web of Science.

The search strategy was straightforward, because nearly all articles in the WHO COVID-19 Research Database are within the domain of the COVID-19 pandemic. No filters or limits were placed in the first screening process – we just screened out articles that lacked a title, name, or abstract, and then we removed duplicate records. Approximately 56% of the records remained and moved forward to the next screening process.

Eligibility Criteria

In our criteria, we included articles that related to one of three main scopes of study: (1) COVID-19, using the keywords “COVID” or “Coronavirus disease 2019”, (2) pandemic risk, using the keywords “pandemic risk”, and (3) risk assessment, using the keywords “risk assessment.”

The language of each article, the nature of the article, the research field of study, and the accessibility of each article was the four filters in our exclusion criteria. We excluded (1) non-English-language articles, (2) articles with a nature equivalent to letters/comments/abstracts, and (3) fields of study belonging to “clinical”, “medical”, “virology”, “finance”, “business”, “logistics”, “supply chain”, and “pharmacy”. For criterion (4), the accessibility of the article, inaccessible or nonidentifiable articles, including non-open-access papers, full-text pdfs, unavailable papers, and papers without a valid DOI (Digital Object Identifier) were excluded because we could not examine the entire papers to determine whether they were within our research focus.

After the screening, we conducted an additional manual scan of the eligible articles. Meta-analyses and articles not relating to our research questions were then excluded, leaving a final eligible list of 62 articles. A description of the inclusion and exclusion criteria of the articles for this study is presented in Table 1. The quality of the included articles was assessed using the Effective Public Health Practice Project (EPHPP) Quality Assessment Tool.14 Each included article was assessed by two reviewers, who conducted the assessment independently.

Table 1 Summary of the Inclusion and Exclusion Criteria

Results

The article search and selection process is shown in Figure 1. Initially, there were 626,900 records in the WHO COVID-19 Research Database. Only 410 records were entered into our “Eligibility” phase. In the last phase, “Included”, the number of records was further reduced to 62. These 62 articles were identified as the final candidates for analysis in this review.

Figure 1 PRISMA flow diagram of the article selection process.

We developed our classification framework based on six research questions. After the classification framework was developed, one of the reviewers performed a preliminary allocation by assigning each of the 62 included articles according to the respective classification types from each research question. To mitigate the risk of bias, another reviewer validated the preliminary allocation by assessing the individual classification of each article in each research question. Finally, a third reviewer examined all of the classification discrepancies between the first two reviewers and arrived at the final summary statistics for the six research questions, which we then described and visualized in the tables and figures shown below.

RQ1: Study Location

We divided the articles into four different sizes of the geographical areas in which the COVID-19 pandemic risk was assessed: regional areas, specific areas, local areas, and global coverage. The study location of one article was not classifiable because that article was a pandemic risk modelling evaluation and no location was indicated. Descriptions of the four sizes of geographical areas, with some examples, are given in Table 2.

Table 2 Description of the Four Sizes of Geographical Areas

Approximately one-third (32.3%) of the articles had conducted COVID-19 pandemic risk assessment in a local area, and another one-third had assessed the risk globally. Regional areas had attracted the least attention of researchers, constituting just 6.5% of the 62 included articles (Figure 2).

Figure 2 Distribution of the four sizes of geographical area studied (The numbers shown inside/outside the pie chart are the frequency count and the percentage of the 62 articles, respectively.).

Among the 17 articles concerned with risk assessment in specific areas, five articles focused on a single province in China, such as Qingdao17 and Hubei.18 India followed China as the second most popular specific area to have been studied, but the frequency was just two. The remaining 10 specific areas were each at a different location (Figure 2).

Regarding the 20 local-area articles, the countries of interest were quite diverse, with 12 different countries, only two of which had been studied in more than two articles. China (seven articles) was again the top country to arouse interest and to have been studied in a countrywide risk assessment, followed by the USA (three articles).

RQ2: Types of Variables Used

We classified the nature of the variables used in the included articles into six different types of data that the variables represented: (1) epidemic data, (2) population/demographic data, (3) mobility/transportation data, (4) socioeconomic data, (5) survey data, and (6) environmental data. Table 3 gives examples of each data type.

Table 3 Examples of the Six Types of Data Represented by the Variables

As shown in Figure 3, epidemic data was the most common type of variable used, appearing in nearly three-quarters (74.2%) of the 62 included articles, whereas the percentage was less than 50% for each of the other five types of variables. Environmental data were a relatively unpopular variable type and were used in only 9.7% of the 62 included articles.

Figure 3 Penetration rate by the different types of variables (= number of articles that used this type of variable/62).

Figure 4 measures how broadly the different types of variables were used in the 62 included articles (ie, how many types of variables in Table 2 were used). The minimum breadth (the use of only one type of variable) and maximum breadth (the use of five types of variables) were 1 and 5, respectively. Four articles were grouped into “Others” mainly because they lacked enough relevant information to precisely identify the variable types used in their studies.

Figure 4 Numbers of the types of variables used by the 62 included articles (Numbers shown inside the pie chart are the frequency count and the percentage based on 62 articles, respectively.).

Approximately 41.9% of the 62 included articles used only one of the six types of variables, as is shown in the upper half of Figure 4. The types of variables used by these 26 articles were epidemic data, mobility/transportation data, and survey data. Epidemic data variables (19 articles) were the distinctly most popular type of variable among the articles using a single type of variable, compared with mobility/transportation data (five articles) and survey data (two articles). More than half (51.6%) of the 62 included articles used two or more types of variables. Use of two types of variables (29%) followed use of a single type of variable as the second most common number of types of variables used in the risk assessment.

Among the 18 articles using two types of variables, the lower half of Figure 4 shows that epidemic data and population/demographic data (eight articles) were the most popular pair, followed by population/demographic data and survey data (four articles).

RQ3: Availability of the Materials Used to Generate Research Outcomes

Every study’s data collected and computing codes used to realize research outcomes are essential materials during the development of an article reporting on that research. The left side of Figure 5 summarizes the availability of the data and codes for the 62 articles we reviewed. More than half (54.8%) of the articles we analyzed did not mention whether their data and/or codes had open access. Approximately 11% of them quoted in their data availability statement that interested scholars could request data and codes from the authors. The remaining one-third of the 62 articles provided specific hyperlinks for downloading their materials.

Figure 5 Accessibility of data and codes (Numbers shown inside the pie chart are the frequency counts and the percentages based on 62 articles, respectively.).

The right side of Figure 5 shows the breakdown of the 21 articles that made their data and/or codes available: 20 articles made their data available; nine articles made their codes available; and eight articles made both their data and their codes available.

RQ4: Use of Data-Visualization Techniques

All 62 of the articles we reviewed used tables or charts or both to present their research findings and outcomes. Approximately 80% of them (50 articles) included tables. The percentage for those using charts was even higher, at 90.3% (56 articles). Forty-four of the articles (71%) displayed both tables and charts in their presentation. There are many different data-visualization techniques, and we found that seven types of graphic representation of data were used in the 62 articles: (1) time-series plots, (2) bar charts, (3) scatter plots, (4) box plots, (5) 3D plots, (6) network graphs, and (7) heat maps. These seven data-visualization techniques are detailed in Figure 6.

Figure 6 Description of seven common data-visualization techniques.

Of the seven data-visualization techniques mentioned above, three were used in at least half of the 62 included articles: time-series plots (38 articles or 61.3%), scatter plots (33 articles or 53.2%), and bar charts (31 articles or 50%). Box plots and network graphs were less popular data-visualization techniques in our reviewed articles, having been used in only nine articles (14.5%) and eight articles (12.9%), respectively. The relatively low number of articles that presented a network graph was expected because not many of the articles had conducted a network analysis as their research methodology. Figure 7 shows the relative popularities of the different data-visualization techniques.

Figure 7 Penetration rate by different data-visualization techniques (= number of articles that used the specific data visualization technique/62).

Figure 8 measures the breadth of the data-visualization techniques usage by the 56 included articles that used charts (ie, it shows how many types of data-visualization each article used). Of those 56 articles, the minimum (using one type) and maximum (using six types of visualization) breadth were 1 and 6, respectively. Minority groups made up two extremes: those using just one type of data visualization technique (7 articles or 12.5%) and those using more than four types (5 articles or 8.9%).

Figure 8 The number of types of data-visualization techniques used by each of the 56 articles that used charts. (Numbers shown inside the pie chart are, respectively, the frequency count and the percentage of the 56 articles that used charts.).

The most common breadth was 2, occupying approximately one-third (19 articles) of the 56 included articles. As shown in the upper half of Figure 8, out of those 19 pairs, only four different pairs of visualization type were used by more than one article. The most common pairs were “Time series plot & Bar chart” and “Time series plot & Scatter plot”, with six articles using each of those pairs. The less common pairs, with two articles using them, were “Time series plot & Heat map” and “Scatter plot & Heat map”.

The breadth value 3 followed the breadth value 2 and was the second most common size of data visualization techniques used by the 56 articles that used charts. As is shown in the lower half of Figure 8, the most popular triples were “Bar chart & Heat map & Time series plot” and “Bar chart & Heat map & Scatter plot”, with three articles using each of those triplets. Another three articles with the breadth of 3 used two of same techniques, “Time series plot & Scatter plot”, but their third techniques were different.

RQ5: Research Objectives

We found that the primary research objective of each of the articles could be classified into five major themes (see also Figure 9): (1) COVID-19 risk exposure assessments using risk indicators/indexes or identifying risk factors (22 articles or 35.5%); (2) reviews on the effectiveness of policy measures for COVID-19 control and prevention (11 articles or 17.7%); (3) predictions/estimations of COVID-19-related parameters (10 articles or 16.1%); (4) investigations on the patterns of COVID-19 transmission/geographical spread of COVID-19 (eight articles or 12.9%), and (5) specific-focus articles (11 articles or 17.7%).

Figure 9 Breakdown of the 62 articles by research objective. (Numbers shown inside the pie chart are respectively the frequency count and the percentage of the 62 articles.).

Table 4 gives further descriptions of the (1) COVID-19 risk exposure assessments that used risk indicators/indexes or identifying risk factors; (2) reviews on the effectiveness of policy measures for control and prevention; (3) predictions/estimations of COVID-19 related parameters, (4) patterns of transmission/spread of COVID-19 and (5) specific focuses.

Table 4 Further Descriptions of the Research Objectives

RQ6: Research Methodologies

Generally, the articles we reviewed used more than one research method to produce their research outcomes. In each article, we focused on the core aspects of the various methods used, and we identified six core statistical research methodologies from 54 of the included articles: (1) exploratory data analysis (eight articles or 12.9%); (2) network analysis (five articles or 8.1%); (3) time-series analysis (four articles or 6.5%); (4) Susceptible, Infected, and Recovered (SIR)/Susceptible-Exposed-Infectious-Removed (SEIR) Models (seven articles or 11.3%); (5) proposed frameworks/systems (nine articles or 14.5%); and (6) special models/techniques (21 articles or 33.9%). The remaining eight articles (or 12.9%) either provided insufficient information to determine which core methods were used or they used a narrative description/qualitative analysis/context analysis as their core method (Figure 10).

Figure 10 Types of research methodologies used by the 62 included articles. (Numbers shown inside the pie chart are the frequency count and the percentage of the 62 articles, respectively.).

Methodology outlines for the research methodologies for exploratory data analysis, network analysis, time-series analysis, and SIR/SEIR models are summarized as group levels in Table 5. The proposed framework/system and special model/technique approaches are described individually in Tables 6 and 7, respectively, because they are quite unique in nature.

Table 5 Methodology Outlines of the Exploratory Data Analysis, Network Analysis, Time-Series Analysis, and SIR/SEIR Models

Table 6 Methodology Outline of the Proposed Frameworks/Systems Approach

Table 7 Methodology Outline of the Special Models/Techniques Approach

Discussion Principle Findings

From our classification, which we derived solely from observing the nature of the various research questions, we found that the 62 included articles reflected a wide variety of research focuses on different aspects of the COVID-19 pandemic risk assessment. With the exception of the research question “Availability of Materials to Generate Research Outcomes”, each research question classification contained at least four class types. Except epidemic data (74.2% or 46/62) in “Types of Variables Used” and time-series plot (61.3% or 38/62) in “Use of Data-Visualization Techniques”, the distribution of class types was quite diversified, with no distinct class type that was prominent.

The study locations examined by these 62 included articles comprised worldwide coverage. Four class types, based on the size of the geographic areas studied, were identified: Local areas (32.3% of articles), Global coverage (32.3% of articles), Specific areas (27.4% of articles), and Regional areas (6.5% of articles). Seven out of the 20 local-area articles and five out of the 17 specific-areas articles focuses in China. No other single country has such a high frequency of appearance. The study locations of the remaining, much larger proportion of articles were scattered across the globe, either in a single country other than China or in a mix of different countries.

In the era of big data, we are not surprised that as many as six different types of data were used in the 62 articles. As was suggested by the titles of the articles, epidemic data were the most widely used type of data (in 74.2% of articles), while environmental data were the least frequently retrieved type of data (9.7% of articles). Articles using a mixture of types of data (32 articles) did not substantially outnumber those using just one single data type (26 articles), thus suggesting that fully utilizing the diversity of available types of data might not be a prevalent phenomenon in COVID-19 pandemic research work.

It goes without saying that thorough documentation, such as making one’s research data and codes available to readers and other researchers, is essential in published papers in order to facilitate the understanding and the interpretation of one’s research outcomes. Approximately one-third of the 62 articles we reviewed (21 articles) gave clear instructions for open access to the data and the codes for their studies. Even if counting as open access the seven articles that offered possible accessibility to their data and codes upon request from the corresponding authors, more than half of the 62 included articles (34 articles) still did not provide this option. We understand that full transparency of data and codes may not always be possible, due to competing interests or other sensitivity issues, but it is worthwhile for authors to consider at least a limited disclosure of their research materials in order to improve the reliability and the appropriate use of research outcomes by policymakers and healthcare-related professionals.

Use of tables and charts certainly helps explain the process of the research work clearly and effectively to the readers. From the initial stage of data exploration to the later stage of presenting the research findings, we saw many tables and charts in different forms and types. Approximately 80% of the articles (50 articles) used tables and even more articles (56 articles or 90.3%) used charts, whereas 44 articles (71%) used both. We observed seven different types of charts, and as expected, time-series plots were the most common type (used by 38 articles or 61.3%) because the data under study were the time patterns of several waves of COVID-19 transmission. Two special chart types that may not be found commonly in most other studies are particularly useful for visualizing the dissemination of the COVID-19 pandemic: heat maps (used by 28 articles or 45.2%), which displayed the severity of the infection by areas, and network graphs (used by eight articles or 12.9%), which showed the COVID-19 connectedness using straight lines between different places. For the 56 articles using charts, a vast majority (49 articles) used more than one type of chart in order to broaden their visualization effects.

The presumably hot objective of “risk exposure assessment by using risk indicators” did not draw overwhelming interest in the 62 included articles. Although it had the largest proportion (22 articles or 35.5%) of articles, the proportion was smaller than 50%. In addition to risk exposure assessment by using risk indicators, four other research objectives (each with a greater than 10% proportion of the articles) were as follows: effectiveness of policy measures (17.7%), prediction/estimation of COVID-19-related parameters (16.1%), patterns of COVID-19 transmission (12.9%), and specific focuses (17.7%). Given the variety of research objectives in the 62 included articles, policymakers, public health officials, and health-care professionals are urged to rely on the synthesized findings of this systematic review to meet various purposes, such as evaluating the effectiveness of current public health measures, making informed decisions on policies for prevention and control, clinical practices and further research77 for early detection of an outbreak, better preparation, and burden reduction on public health systems in the event of new waves of infectious disease transmission.

No research methodology was dominant in the 62 articles – in contrast, many different methods were employed, as shown in Figure 10. One common observation was that, no matter which methodology was employed (except for the six articles using either a narrative description or context analysis), most articles applied inferential statistics analyses such as factor analyses, time-series analyses, regressions, Bayesian inferences, and the like, to generate their research outcomes. Their process flows were clearly outlined, and their research methodologies were well documented. Such thorough documentation definitely increases the credibility of articles,78 giving full knowledge of what has already been done,79 and facilitating others’ ability to replicate research outcomes, with high confidence for the appropriate use by interested parties.

Systematic Review

For policymakers having an interest in topics, which requires reviewing lots of primary papers and articles in a standardized manner, we suggest the following five key stages. First of all, setting up the objectives by clearly pre-defining specific research questions in the context of what are already known. Second, identifying an explicit and reproducible methodology describing eligibility criteria and search strategy for finding relevant research and collecting data. Third, specifying the methods used to assess the validity of the selected information such that they meet the eligibility criteria and how to identify potential risk of bias such as selection bias on target population, performance bias on treatments and reporting bias on result findings. Fourth, providing pre-planned methodological and analytical approach on how to analyze quantitative data and synthesize qualitative evidence. Lastly, describing how to interpret the results, summarize the findings and recommend actionable plans.

Limitations and Future Research

Some limitations apply to this systematic review. First, it is possible that some relevant studies could not be included. Even though we used a comprehensive and highly relevant source, WHO COVID-19 research database, there may still be chances that some relevant articles were not captured. Another possibility for missing some relevant articles is that certain articles were screened out by the exclusion criteria such as non-English language articles and articles with no open/free access. In addition, the number of included articles in this systematic review might be considered to be not sufficient because of the limited number of articles that met the eligibility criteria. Unlike survey sampling, there is no universal measurement to determine the appropriate size for systematic review. When retrieving published studies of systematic review, it is common to find that the size of the final list is usually less than 100, some may even be less than 30. So, we believe that this limitation does not affect the validity of our findings.

Future research should continue to track the latest development of COVID-19 as it progresses. Two new variants (Omicron and Arcturus) have emerged after the date of searching relevant studies for this systematic review. In addition to capturing more recent relevant articles that studied the new waves of transmission, a critical appraisal tool should be developed in order to assess the quality of the included articles from different assessment criteria such as study design, statistical analysis, and outcomes. This helps to quantify the strengths and weaknesses of the included articles and hence facilitate more in-depth discussion and better interpretation of the findings.

Conclusions

The impact of the COVID-19 pandemic has been enormous, presenting unprecedented challenges to public health. Therefore, researchers conduct risk assessment based on available data and methods to identify risk factors and/or study their effects and consequences. Although we have entered the post-pandemic period for COVID-19, history tells us that we should continue to stay vigilant against both the emergence of a new variant of COVID-19 and also of a new infectious disease.

This systematic review gathered relevant research works about the global COVID-19 pandemic risk assessment by conducting an extensive systematic search in the WHO COVID-19 Research Database, and we here provide useful synthesized findings of what has been done to evaluate the COVID-19 pandemic risks. Policymakers and those who are responsible for public health can refer to our detailed summary of the various research objectives, which we have classified in this systematic review, and can learn from one or more of them depending on the priorities of their country. This information can support informed decisions and plans for informed actions to analyze and monitor the spread of new infectious diseases that are likely to arise in the future.

Abbreviations

PRISMA, Preferred Reporting Items for Systematic Reviews and Meta-Analyses; WHO, World Health Organization.

Data Sharing Statement

The authors confirm that all articles identified through searching WHO database: Global research on coronavirus disease (https://search.bvsalud.org/global-literature-on-novel-coronavirus-2019-ncov/) from its inception to 12 July 2022 and all screened articles are freely available to public at the time of writing by accessing the website (assessed on 20 July 2022).

Acknowledgments

This work was supported by the Research Impact Case Grant from the Department of Social Sciences and Policy Studies, The Education University of Hong Kong, and The Hong Kong University of Science and Technology research grant “Risk Analytics and Applications” (grant number SBMDF21BM07). The funders had no role in study design, data collection and analysis, decision to publish, or preparation of the manuscript.

Disclosure

The authors report no conflicts of interest in this work.

References

1. Coronavirus disease 2019 (COVID-19) situation report—51; World Health Organization: Geneva, Switzerland; 2020. Available from: https://www.who.int/publications/m/item/situation-report---51. Accessed March3, 2023.

2. So MKP, Chu AMY, Chan TWC. Impacts of the COVID-19 pandemic on financial market connectedness. Finance Res Lett. 2021;38:e101864. doi:10.1016/j.frl.2020.101864

3. Goodell JW, Goutte S. Co-movement of COVID-19 and Bitcoin: evidence from wavelet coherence analysis. Finance Res Lett. 2021;38:e101625. doi:10.1016/j.frl.2020.101625

4. Nicola M, Alsafi Z, Sohrabi C, et al. The socio-economic implications of the coronavirus pandemic (COVID-19): a review. Int j Surg. 2020;78:185–193. doi:10.1016/j.ijsu.2020.04.018

5. Zambrano-Monserrate MA, Ruano M, Sanchez-Alcalde L. Indirect effects of COVID-19 on the environment. Science of the Total Environment. 2020;728:e138813. doi:10.1016/j.scitotenv.2020.138813

6. Xiong J, Lipsitz O, Nasri F, et al. Impact of COVID-19 pandemic on mental health in the general population: a systematic review. J Affective Disorders. 2020;277:55–64. doi:10.1016/j.jad.2020.08.001

7. Nsoesie EO, Cesare N, Müller M, Ozonoff A. COVID-19 misinformation spread in eight countries: exponential growth modeling study. J Med Internet Res. 2020;22(12):e24425. doi:10.2196/24425

8. Lee JJ, Kang K, Wang MP, et al. Associations between COVID-19 misinformation exposure and belief with COVID-19 knowledge and preventive behaviors: cross-sectional online study. J Med Internet Res. 2020;22(11):e22205. doi:10.2196/22205

9. Lundberg AL, Lorenzo-Redondo R, Hultquist JF, et al. Overlapping Delta and Omicron outbreaks during the COVID-19 pandemic: dynamic panel data estimates. JMIR Public Health Surveillance. 2022;8(6):e37377. doi:10.2196/37377

10. Chandrasekaran R, Mehta V, Valkunde T, Moustakas E. Topics, trends, and sentiments of Tweets about the COVID-19 pandemic: temporal infoveillance study. J Med Internet Res. 2020;22(10):e22624. doi:10.2196/22624

11. WHO COVID-19 research database; WHO. Available from: https://search.bvsalud.org/global-literature-on-novel-coronavirus-2019-ncov/. Accessed March3, 2023.

12. WHO COVID-19 research database: user guide and information; WHO. Available from: https://www.who.int/publications/m/item/quick-search-guide-who-covid-19-database. Accessed March3, 2023.

13. Moher D, Liberati A, Tetzlaff J, Altman DG. Preferred reporting items for systematic reviews and meta-analyses: the PRISMA statement. PLoS Med. 2009;6(7):e1000097. doi:10.1371/journal.pmed.1000097

14. Armijo-Olivo S, Stiles CR, Hagen NA, Biondo PD, Cummings GG. Assessment of study quality for systematic reviews: a comparison of the Cochrane Collaboration Risk of Bias Tool and the Effective Public Health Practice Project Quality Assessment Tool: methodological research. J Evaluation Clin Practice. 2012;18(1):12–18. doi:10.1111/j.1365-2753.2010.01516.x

15. Impouma B, Mboussou F, Wolfe CM, et al. COVID-19 in the WHO African region: using risk assessment to inform decisions on public health and social measures. Epidemiol Infect. 2021:149. doi:10.1017/s0950268821001126

16. Krueger T, Gogolewski K, Bodych M, et al. Risk assessment of COVID-19 epidemic resurgence in relation to SARS-CoV-2 variants and vaccination passes. Communicat Med. 2022;2(1):84. doi:10.1038/s43856-022-00084-w

17. Zhang S, Wang M, Yang Z, Zhang B. A novel predictor for micro-scale COVID-19 risk modeling: an empirical study from a spatiotemporal perspective. Int J Environ Res Public Health. 2021;18(24):e13294. doi:10.3390/ijerph182413294

18. Liu Y, Zheng F, Du Z, et al. Evaluation of China’s Hubei control strategy for COVID-19 epidemic: an observational study. BMC Infect Dis. 2021;21(1). doi:10.1186/s12879-021-06502-z

19. Nazia N, Law J, Butt ZA. Identifying spatiotemporal patterns of COVID-19 transmissions and the drivers of the patterns in Toronto: a Bayesian hierarchical spatiotemporal modelling. Sci Rep. 2022;12(1). doi:10.1038/s41598-022-13403-x

20. Wani MA, Farooq J, Wani DM. Risk assessment of COVID-19 pandemic using deep learning model for J&K in India: a district level analysis. Environ. Sci. Pollut. Res. 2021;29(12):18271–18281. doi:10.1007/s11356-021-17046-9

21. Donnat C, Bunbury F, Kreindler J, et al. Predicting COVID-19 transmission to inform the management of mass events: model-based approach. JMIR Public Health Surveillance. 2021;7(12):e30648. doi:10.2196/30648

22. Pluchino A, Biondo AE, Giuffrida N, et al. A novel methodology for epidemic risk assessment of COVID-19 outbreak. Sci Rep. 2021;11(1). doi:10.1038/s41598-021-82310-4

23. Li Q, Tang Z, Coleman N, Mostafavi A. Detecting early-warning signals in time series of visits to points of interest to examine population response to COVID-19 pandemic. IEEE Access. 2021;9:27189–27200. doi:10.1109/access.2021.3058568

24. Shimizu K, Negita M. Lessons learned from Japan’s response to the first wave of COVID-19: a content analysis. Healthcare. 2020;8(4):426. doi:10.3390/healthcare8040426

25. Parajuli RR, Mishra B, Banstola A, et al. Multidisciplinary approach to COVID-19 risk communication: a framework and tool for individual and regional risk assessment. Sci Rep. 2020;10(1). doi:10.1038/s41598-020-78779-0

26. Jia JS, Lu X, Yuan Y, Xu G, Jia J, Christakis NA. Population flow drives spatio-temporal distribution of COVID-19 in China. Nature. 2020;582(7812):389–394. doi:10.1038/s41586-020-2284-y

27. Li T, Luo J, Huang C. Understanding small Chinese cities as COVID-19 hotspots with an urban epidemic hazard index. Sci Rep. 2021;11(1). doi:10.1038/s41598-021-94144-1

28. Grima S, Rupeika-Apoga R, Kizilkaya M, Romānova I, Gonzi RD, Jakovljevic M. A proactive approach to identify the exposure risk to COVID-19: validation of the pandemic risk exposure measurement (PREM) model using real-world data. Risk Management Healthcare Policy. 2021;14:4775–4787. doi:10.2147/rmhp.s341500

29. Amer F, Hammoud S, Farran B, Boncz I, Endrei D. Assessment of countries’ preparedness and lockdown effectiveness in fighting COVID-19. Disaster Medicine and Public Health Preparedness. 2020;15(2):e15–e22. doi:10.1017/dmp.2020.217

30. Yao L, Dong W, Wan JY, Howard SC, Li M, Graff JC. Graphical trajectory comparison to identify errors in data of COVID-19: a cross-country analysis. J Personalized Med. 2021;11(10):955. doi:10.3390/jpm11100955

31. Chakraborty T, Ghosh I. Real-time forecasts and risk assessment of novel coronavirus (COVID-19) cases: a data-driven analysis. Chaos, Solitons Fractals. 2020;135:e109850. doi:10.1016/j.chaos.2020.109850

32. Maged A, Ahmed A, Haridy S, Baker AW, Xie M. SEIR model to address the impact of face masks amid COVID-19 pandemic. Risk Anal. 2022;43(1):129–143. doi:10.1111/risa.13958

33. Vieira KM, Potrich ACG, Bressan AA, Klein LL, Pereira BAD, Pinto NGM. A pandemic risk perception scale. Risk Anal. 2021;42(1):69–84. doi:10.1111/risa.13802

34. Guo Z, Liu X, Zhao P. A vector field approach to estimating environmental exposure using human activity data. ISPRS Int J Geo-Information. 2022;11(2):135. doi:10.3390/ijgi11020135

35. Kanga S, Meraj G, Sudhanshu M, Nathawat M. Analyzing the risk to COVID-19 infection using remote sensing and GIS. Risk Anal. 2021;41(5):801–813. doi:10.1111/risa.13724

36. Mahmood M, Mateu J, Hernández-Orallo E. Contextual contact tracing based on stochastic compartment modeling and spatial risk assessment. Stochastic Environ Res Risk Assessment. 2021;36(3):893–917. doi:10.1007/s00477-021-02065-2

37. Kheirallah KA, Al-Nusair M, Aljabeiti S, et al. Jordan’s pandemic influenza preparedness (PIP): a reflection on COVID-19 response. Int J Environ Res Public Health. 2022;19(12):7200. doi:10.3390/ijerph19127200

38. Barthe G, Viti RD, Druschel P, et al. Listening to Bluetooth beacons for epidemic risk mitigation. Sci Rep. 2022;12(1). doi:10.1038/s41598-022-09440-1

39. Baik I. Region-specific COVID-19 risk scores and nutritional status of a high-risk population based on individual vulnerability assessment in the national survey data. Clin Nutr. 2022;41(12):3100–3105. doi:10.1016/j.clnu.2021.02.019

40. Zhang Y, Li Y, Yang B, Zheng X, Chen M. Risk assessment of COVID-19 based on multisource data from a geographical viewpoint. IEEE Access. 2020;8:125702–125713. doi:10.1109/access.2020.3004933

41. Grima S, Kizilkaya M, Rupeika-Apoga R, Romānova I, Gonzi RD, Jakovljevic M. A country pandemic risk exposure measurement model. Risk Management Healthcare Policy. 2020;13:2067–2077. doi:10.2147/rmhp.s270553

42. Bird JJ, Barnes CM, Premebida C, Ekárt A, Faria DR. Country-level pandemic risk and preparedness classification based on COVID-19 data: a machine learning approach. PLoS One. 2020;15(10):e0241332. doi:10.1371/journal.pone.0241332

43. Zhou L, Liu J-M, Dong X-P, McGoogan JM, Wu Z-Y. COVID-19 seeding time and doubling time model: an early epidemic risk assessment tool. Infect Diseases Poverty. 2020;9(1). doi:10.1186/s40249-020-00685-4

44. Gunthe SS, Patra SS. Impact of international travel dynamics on domestic spread of 2019-nCoV in India: origin-based risk assessment in importation of infected travelers. Globalization Health. 2020;16(1). doi:10.1186/s12992-020-00575-2

45. Wang Y, Wang L, Zhao X, et al. A semi-quantitative risk assessment and management strategies on COVID-19 infection to outpatient health care workers in the post-pandemic period. Risk Management Healthcare Policy. 2021;14:815–825. doi:10.2147/rmhp.s293198

46. Liu J, Liu L, Tu Y, Li S, Li Z. Multi-stage Internet public opinion risk grading analysis of public health emergencies: an empirical study on microblog in COVID-19. Information Processing and Management. 2022;59(1):e102796. doi:10.1016/j.ipm.2021.102796

47. Ronchi E, Lovreglio R. EXPOSED: an occupant exposure model for confined spaces to retrofit crowd models during a pandemic. Safety Science. 2020;130:e104834. doi:10.1016/j.ssci.2020.104834

48. Wang Z, Yao M, Meng C, Claramunt C. Risk assessment of the overseas imported COVID-19 of ocean-going ships based on AIS and infection data. ISPRS Int J Geo-Information. 2020;9(6):351. doi:10.3390/ijgi9060351

49. Tyagi N, Gurian PL, Kumar A. Using QMRA to understand possible exposure risks of SARS-CoV-2 from the water environment. Environ. Sci. Pollut. Res. 2021;29(5):7240–7253. doi:10.1007/s11356-021-16188-0

50. Gunaratne C, Reyes R, Hemberg E, O’Reilly U-M. Evaluating efficacy of indoor non-pharmaceutical interventions against COVID-19 outbreaks with a coupled spatial-SIR agent-based simulation framework. Sci Rep. 2022;12(1). doi:10.1038/s41598-022-09942-y

51. Trentini F, Manna A, Balbo N, et al. Investigating the relationship between interventions, contact patterns, and SARS-CoV-2 transmissibility. Epidemics. 2022;40:e100601. doi:10.1016/j.epidem.2022.100601

52. Cheng Q, Zheng S, Xiong Z, Lin M. Characterizing the dynamic evolution of interagency collaborative decision-making networks in response to COVID-19 in China: a policy document analysis. Healthcare. 2022;10(3):590. doi:10.3390/healthcare10030590

53. Chu AMY, Chan TWC, So MKP, Wong W-K. Dynamic network analysis of COVID-19 with a latent pandemic space model. Int J Environ Res Public Health. 2021;18(6):3195. doi:10.3390/ijerph18063195

54. Pan J, Tian J, Xiong H, et al. Risk assessment and evaluation of China’s policy to prevent COVID-19 cases imported by plane. PLoS Negl Trop Dis. 2020;14(12):e0008908. doi:10.1371/journal.pntd.0008908

55. Jian S-W, Kao C-T, Chang Y-C, Chen P-F, Liu D-P. Risk assessment for COVID-19 pandemic in Taiwan. Inter J Infect Dis. 2021;104:746–751. doi:10.1016/j.ijid.2021.01.042

56. Kandel N, Chungong S, Omaar A, Xing J. Health security capacities in the context of COVID-19 outbreak: an analysis of International Health Regulations annual report data from 182 countries. Lancet. 2020;395(10229):1047–1053. doi:10.1016/s0140-6736(20)30553-5

57. Duffey RB, Zio E. Prediction of COVID-19 infection, transmission and recovery rates: a new analysis and global societal comparisons. Safety Science. 2020;129:e104854. doi:10.1016/j.ssci.2020.104854

58. Nielsen J, Rod NH, Vestergaard LS, Lange T. Estimates of mortality attributable to COVID-19: a statistical model for monitoring COVID-19 and seasonal influenza, Denmark, spring 2020. Eurosurveillance. 2021;26(8). doi:10.2807/1560-7917.es.2021.26.8.2001646

59. Jung S-m, Akhmetzhanov AR, Hayashi K, et al. Real-time estimation of the risk of death from novel coronavirus (COVID-19) infection: inference using exported cases. J Clin Med. 2020;9(2):523. doi:10.3390/jcm9020523

60. Marani M, Katul GG, Pan WK, Parolari AJ. Intensity and frequency of