Abstract
This study explored a Bayesian belief networks (BBNs) approach, developing two distinct models for prioritizing the seven indicators related to the “rapid response to and mitigation of the spread of an epidemic” category within the context of both the specific category and the Global Health Security Index (GHS index). Utilizing data from the 2021 GHS index, the methodology involves rigorous preprocessing, the application of the augmented naive Bayes algorithm for structural learning, and k-fold cross-validation. Key findings show unique perspectives in both BBN models. In the mutual value of information analysis, “linking public health and security authorities” emerged as the key predictor for the “rapid response to and mitigation of the spread of an epidemic” category, while “emergency preparedness and response planning” assumed precedence for the GHS index. Sensitivity analysis highlighted the critical role of “emergency preparedness and response planning” and “linking public health and security authorities” in extreme performance states, with “access to communications infrastructure” and “trade and travel restrictions” exhibiting varied significance. The BBN models exhibit high predictive accuracy, achieving 83.3% and 82.3% accuracy for extreme states in “rapid response to and mitigation of the spread of an epidemic” and the GHS index, respectively. This study contributes to the literature on GHS by modeling the dependencies among various indicators of the rapid response dimension of the GHS index and highlighting their relative importance based on the mutual value of information and sensitivity analyses.
Similar content being viewed by others
Explore related subjects
Discover the latest articles, news and stories from top researchers in related subjects.Avoid common mistakes on your manuscript.
1 Introduction
Global health security (GHS) is a critical and interconnected concept that addresses the collective efforts to safeguard the health of populations worldwide from potential threats, whether they are infectious diseases, natural hazard-related disasters, or other health emergencies (Zhang et al. 2023). In an increasingly interconnected world, where diseases can easily transcend borders and impact diverse communities, the importance of GHS has become paramount (Wennman et al. 2022). The approach to GHS involves proactive measures such as surveillance, early detection, and rapid response to outbreaks, as well as strengthening health systems and promoting international collaboration (Wennman et al. 2022). The ongoing challenges posed by emerging infectious diseases, such as the Covid-19 pandemic, underscore the need for a comprehensive and coordinated GHS strategy to protect the well-being of individuals and communities across the globe (Fakhruddin et al. 2020).
The Global Health Security Index (GHS index) serves as a comprehensive assessment tool that evaluates a country’s preparedness and capacity to prevent, detect, and respond to public health emergencies (Leichtweis et al. 2021). Developed in collaboration between the Nuclear Threat Initiative, the Johns Hopkins Center for Health Security, and the Economist Intelligence Unit, the GHS index provides a systematic framework to measure and compare the resilience of countries against various health security threats (Bell and Nuzzo 2021). This index considers factors such as healthcare infrastructure, surveillance capabilities, emergency response systems, and overall risk environment. By offering a standardized and transparent evaluation, the GHS index aims to enhance global awareness, facilitate targeted improvements in health security, and promote international cooperation to address emerging challenges and protect populations from the impact of infectious diseases and other health crises (Ji et al. 2021).
The GHS index involves six key categories, each playing a crucial role in assessing a country’s preparedness for health emergencies (Bell and Nuzzo 2021). The first category, “prevention of the emergence or release of pathogens,” evaluates a country’s efforts to minimize the risk of infectious diseases through measures such as biosecurity and biosafety. The second category, “early detection and reporting of epidemics,” focuses on a country’s ability to promptly identify and report potential health threats. “Rapid response to and mitigation of the spread of an epidemic” assesses the country’s capacity for swift and effective actions once a threat is identified. “Sufficient and robust health system to treat the sick and protect health workers” evaluates the resilience and capability of a country’s healthcare infrastructure. The fifth category, “commitment to improving national capacity, financing, and adherence to global norms” assesses a country’s dedication to ongoing improvements in health security. Last, “overall risk environment and country vulnerability to biological threats” considers the broader contextual factors that may influence a country’s vulnerability to health emergencies.
The “rapid response to and mitigation of the spread of an epidemic” category within the GHS index places a critical emphasis on a country’s ability to swiftly and effectively respond to health emergencies (Assefa et al. 2021). Indicators and subindicators in this category include various aspects of emergency preparedness and response planning, such as the existence and inclusivity of national public health emergency plans, private sector involvement, and nonpharmaceutical intervention planning (Bell and Nuzzo 2021). Additionally, attention is given to risk communication planning and the accessibility of communication infrastructure, as reflected by indicators measuring Internet and mobile phone usage, with a focus on gender inclusivity (Pereira et al. 2022). Trade and travel restrictions, vital tools in mitigating the spread of epidemics, are also assessed, encompassing both the implementation of restrictions and their impact on national and international movements (Gwenzi et al. 2022).
Existing studies on the “rapid response to and mitigation of the spread of an epidemic” category within the GHS index have predominantly focused on individual indicators, providing valuable insights into emergency preparedness, response planning, and operational capabilities (Chiossi et al. 2021). However, a notable gap exists in terms of modeling the intricate interactions among these indicators within a probabilistic network setting. This gap limits our understanding of how different aspects of rapid response relate to and influence each other, hindering the development of a holistic framework for assessing a country’s readiness in the face of an epidemic. To address this gap, the main aim of this study was to employ a Bayesian belief network (BBN) model, offering a systematic and probabilistic approach to capture the complex interdependencies among indicators in the specified category and their collective impact on the overall GHS index. By utilizing the BBN model, this research seeks to establish the relative importance of each related indicator, providing a better understanding of their contributions and interplays in shaping a country’s ability to respond effectively to health emergencies.
The subsequent sections of this article are arranged as follows: Sect. 2 offers a brief overview of the relevant literature. Section 3 presents the research methodology. Section 4 explores the results along with their implications. Section 5 outlines potential directions for future research.
2 Literature Review
This section provides a comprehensive literature review on GHS, focusing on the GHS index and its role in evaluating a country’s preparedness for health emergencies. It examines key studies that highlight the importance of rapid response and risk communication in managing health crises. Additionally, the review explores the application of BBNs in health security research, emphasizing their utility in analyzing complex interdependencies and improving health system resilience.
2.1 Global Health Security
The GHS index serves as a crucial tool in evaluating a country’s preparedness for health emergencies, providing a comprehensive framework that encompasses prevention, detection, and rapid response (Pereira et al. 2022). Various studies have underscored the significance of understanding a country’s health security capabilities, emphasizing the need for a holistic approach in a globally interconnected context (Kachali et al. 2022; Zawadzki and Montibeller 2023).
Within the GHS index, the category “rapid response to and mitigation of the spread of an epidemic” plays a key role in evaluating a country’s ability to respond swiftly to health emergencies (Nawaz et al. 2020). This category covers diverse indicators, including emergency preparedness, response planning, risk communication, and trade and travel restrictions (Lencucha and Bandara 2021). Given its importance, a more in-depth exploration of this category is warranted (Bruinen de Bruin et al. 2020).
Recent research articles have provided valuable insights into specific elements of health emergency preparedness. Chiossi et al. (2021) delved into the effectiveness of national public health emergency preparedness plans, emphasizing the importance of strategic planning in response operations. Ab Aziz et al. (2019) focused on risk communication strategies, highlighting their role in facilitating effective emergency response operations.
While individual studies have contributed to understanding specific indicators, comprehensive research on the associations among these indicators within the “rapid response to and mitigation of the spread of an epidemic” category is limited (Liu et al. 2021). Sarmiento et al. (2015) explored the interplay between private sector involvement and response planning, revealing the need for a more integrated approach to emergency preparedness efforts. Existing studies offer valuable insights into isolated components, but a comprehensive understanding of how these elements interact remains limited.
To address the identified gap, this study adopted a BBN model. Bayesian belief networks offer a robust approach to capture complex interdependencies among indicators, providing a probabilistic framework to analyze their relationships (Al Nuairi et al. 2023). The use of a BBN in this context is crucial for establishing a better understanding of how various components within the “rapid response to and mitigation of the spread of an epidemic” category influence each other.
2.2 Bayesian Belief Networks
Bayesian belief networks have gained prominence as a powerful modeling tool in various domains, offering a probabilistic framework to represent and analyze complex relationships among variables (Hosseini and Ivanov 2020). Their ability to capture uncertainty and model dependencies makes BBNs well-suited for applications in health security, where complex relationships and uncertainties often characterize the dynamics of emergencies and response mechanisms (Cox et al. 2016).
Previous studies have explored the utility of BBNs in public health contexts, demonstrating their effectiveness in modeling disease spread, risk assessment, and decision-making processes (Liao et al. 2017; Liao et al. 2018). For instance, Ayre et al. (2014) employed BBNs to predict the spread of infectious diseases, providing valuable insights into potential transmission routes and critical intervention points. In health security, BBNs have been increasingly applied to assess and enhance preparedness for emergencies, including epidemics and pandemics (Vizanko et al. 2024). For instance, Qazi et al. (2022) used BBNs to evaluate the relative importance of various factors influencing Covid-19, providing a systematic approach for decision makers to prioritize resources and response efforts.
Bayesian belief networks excel in capturing the complex interactions among various factors influencing health security (Qazi et al. 2021). Studies by Qazi and Simsekler (2020) and Motwani et al. (2022) highlighted the use of BBNs to model relationships between healthcare infrastructure, communication strategies, and response effectiveness during health emergencies. These studies show the versatility of BBNs in providing a holistic understanding of health security dynamics.
The integration of BBNs into decision support systems has been a focus of recent research (Al Nuairi et al. 2023). By incorporating diverse data sources and expert knowledge, BBNs offer decision makers a comprehensive tool for scenario analysis and risk assessment (Simsekler and Qazi 2022). Vizanko et al. (2024) utilized BBNs to develop decision support systems for health emergency management, aiding in the identification of optimal response strategies.
Despite their potential, challenges exist in the application of BBNs in health security. Ensuring the accuracy of prior probabilities, handling large-scale networks, and addressing dynamic changes in variables are among the challenges (Qazi 2024). Recent advancements, such as hybrid models combining machine learning and BBNs aim to address these issues and enhance the accuracy of health security assessments (Qazi et al. 2021).
The GHS index provides a comprehensive framework for evaluating health security, and BBNs offer a promising avenue to enhance its assessment capabilities. By modeling the interactions among indicators within specific categories, such as the “rapid response to and mitigation of the spread of an epidemic,” BBNs can provide a better understanding of a country’s readiness and contribute to refining the GHS index. Moreover, BBNs can empower decision makers to prioritize interventions, allocate resources efficiently, and enhance overall preparedness and response strategies.
3 Research Methodology
To achieve the research objective, a research methodology was implemented that leverages BBNs to model the probabilistic relationships inherent in indicators falling under the “rapid response to and mitigation of the spread of an epidemic” category of the GHS index for the year 2021. This section offers a comprehensive overview of the utilized data, explains the fundamental principles of BBNs, and outlines the procedural steps crucial to the modeling process.
In this investigation, our primary data came from the GHS index for 2021, authored by Bell and Nuzzo (2021). This index is divided into six categories, further subdivided into 37 indicators, 96 subindicators, and a total of 171 questions. The data for the GHS index were gathered from various countries, institutions, and publicly available sources, including the World Health Organization, World Organization for Animal Health, World Bank, Food and Agriculture Organization of the United Nations, and academic publications. Research for the 2021 edition of the GHS index was carried out by the Economist Impact from August 2020 to June 2021. Over 900 researchers and analysts worldwide contributed to this effort, collecting data from primary legal documents, government publications, academic sources, and various websites. The team acknowledged the impact of the Covid-19 pandemic on data availability, especially noting the capacities developed by countries during the validation period from August to September 2021. This extensive assessment, spanning 195 countries, grounds our analysis.
Country scores, ranging from 0 to 100, were computed as weighted averages of the six categories. Each category was evaluated on a scale of 0 to 100, with 100 indicating the most favorable health security conditions and 0 indicating the least favorable conditions. However, it is essential to note that a score of 100 does not imply flawless national health security, nor does a score of 0 denote a complete absence of capacity. Instead, these scores represent the maximum or minimum achievable scores based on the GHS index criteria. The normalization process for each category involved considering the sums of its underlying indicators and subindicators, with consistent weights applied. The default ranking weights were described as neutral or uniform. Descriptive statistics and correlation results of the dataset used in this study are provided in Tables 1 and 2, respectively.
The average GHS index value across 195 countries stands at 38.91, indicating low performance overall. The “access to communications infrastructure” indicator shows the highest mean value of 65.70. Conversely, “exercising response plans” registers the lowest mean value of 21.09. Both “rapid response and mitigation” and “emergency preparedness and response planning” display significant correlations with the GHS index. Within the “rapid response and mitigation” category, “linking public health and securities” exhibits a particularly strong correlation. However, the “trade and travel restrictions” indicator shows a weak negative correlation with other variables.
In our study, we used BBNs, which are graphical models that represent probabilistic relationships between variables. At their core, BBNs consist of nodes representing variables of interest and directed edges indicating probabilistic dependencies between them. These dependencies are quantified using conditional probability tables (CPTs), which specify the probability distribution of each node given its parents in the network (Albreiki et al. 2024). For instance, consider a BBN modeling the spread of a contagious disease. Nodes could represent variables like “infection status,” “symptoms exhibited,” and “contact history.” The edges between these nodes capture how the presence of symptoms depends on infection status and contact history, among other factors.
Probabilistic reasoning in BBNs involves updating beliefs about variables based on observed evidence. This is done using Bayes’ theorem, which enables the calculation of posterior probabilities given prior probabilities and observed data. In our disease example, if we observe certain symptoms in a patient, we can update the probabilities of various diseases based on their likelihood of causing those symptoms.
The strength of BBNs lies in their ability to handle uncertainty and model complex systems with multiple interacting variables. They provide a structured framework for combining empirical data, expert knowledge, and probabilistic reasoning to make informed decisions (Simsek et al. 2021). In our study, we leveraged these characteristics to model the dependencies among various indicators of the “rapid response and mitigation” category of the GHS index, ultimately aiding in prioritization and decision making in public health contexts. Bayesian belief networks prove to be a powerful instrument for engaging in probabilistic modeling, demonstrating particular efficacy in capturing and quantifying uncertainties and interdependencies within complex systems (Cooper et al. 2023). Specifically, within the context of the GHS index, BBNs can offer a visual depiction of the probabilistic relationships among its indicators, enabling a thorough examination of their complex connections.
A significant strength of BBNs lies in their ability to model uncertain and dynamic relationships, setting them apart from deterministic models, as highlighted by Al Nuairi et al. (2023). Bayesian belief networks adopt a probabilistic reasoning approach, recognizing the uncertainty inherent in real-world scenarios. This becomes particularly vital when dealing with complex systems like GHS, where variables may be influenced by multiple factors and display probabilistic interdependencies. However, BBNs have certain limitations. Constructing a BBN necessitates big data to establish relationships among variables (Werner et al. 2017). Furthermore, the computational complexity of BBNs may escalate with the increasing number of variables and potential interactions, requiring substantial consideration during model development, as noted by Qazi and Khan (2021). Despite these challenges, the advantages of leveraging BBNs in capturing probabilistic dependencies and uncertainties outweigh the drawbacks (Hossain et al. 2019).
In comparing BBNs with other techniques commonly adopted in the context of GHS, BBNs offer distinct advantages over various methodologies. For instance, regression models may struggle with identifying nonlinear dependencies, while BBNs offer a flexible graphical structure capable of representing both linear and nonlinear associations. Furthermore, while machine learning techniques like neural networks and tree-based algorithms excel in pattern recognition and prediction, they often lack interpretability and may not capture causal relationships as effectively as BBNs (Simsekler et al. 2021). Additionally, BBNs outperform system dynamics models in their ability to handle uncertainty and incorporate expert knowledge through probabilistic reasoning, a crucial aspect in the context of GHS where data may be incomplete or imprecise. Overall, the adaptability, interpretability, and robustness of BBNs make them a suitable methodology in our study, ensuring a comprehensive understanding of the dependencies among indicators and enhancing the effectiveness of prioritization efforts in public health security.
Bayesian belief networks serve as a robust tool in the context of GHS. At their core, BBNs offer a framework for representing and reasoning under uncertainty, aligning well with the inherently unpredictable nature of infectious disease outbreaks and other health emergencies (Cook et al. 2023). Leveraging probabilistic graphical models, BBNs allow for the integration of diverse data sources, including epidemiological, environmental, and social factors, facilitating a comprehensive understanding of the complex dynamics at play during a health crisis. Moreover, BBNs enable scenario analysis and decision support, supporting policymakers and public health officials to anticipate potential outcomes and prioritize interventions effectively. By explicitly modeling causal relationships and statistical dependencies among variables, BBNs provide a theoretical foundation for proactive and adaptive response strategies, enabling stakeholders to swiftly adjust tactics as situations evolve. Thus, the incorporation of BBNs into GHS frameworks not only enhances the analytical capabilities but also reinforces the theoretical underpinnings of rapid response and mitigation efforts, supporting global health resilience in the face of emerging threats.
The initiation of the BBN model’s development involved preprocessing GHS index data to address missing values and outliers, ensuring overall data quality. Subsequently, a uniform-width discretization scheme categorized the selected variables into three performance states—low, medium, and high. This discretization approach, recognized for its superior performance (Al Nuairi et al. 2023), illustrated individual variable performance, with low denoting the worst state and high representing the best.
For the structural learning of the BBN model, the application of the augmented naive Bayes (ANB) algorithm was pivotal, providing flexibility by relaxing constraints found in other algorithms (Qazi 2024). The GeNIe softwareFootnote 1 facilitated the development of the BBN model, determining conditional probability values through the maximum likelihood estimation method (BayesFusion 2017). Two models were constructed, each incorporating the same indicators (predictor variables), but with a different target variable (that is, GHS index and “rapid response and mitigation”).
Validation of the two BBN models employed a k-fold cross-validation technique, dividing the dataset into k equally-sized folds, as suggested by Marcot and Hanea (2021). The iterative training and testing process, repeated k times, offered a robust assessment of the model’s generalization to unseen data. The averaged results provided a dependable estimate of the BBN’s overall performance, mitigating the impact of data variability. The models showcased an accuracy of 83.3% and 82.3% for predicting the two extreme states of “rapid response and mitigation” and the GHS index, respectively. To evaluate the model’s robustness, sensitivity analysis identified influential variables and assessed the impact of uncertainties on overall predictions.
4 Results and Discussion
This section presents the results and discussion of the study, focusing on the analysis of two BBN models. The first model prioritizes indicators influencing the “rapid response and mitigation” category within the GHS index. The second model examines how various indicators contribute to the overall health security as reflected by the GHS index. The section outlines the key findings from both models, highlighting the relative importance of individual indicators, their interconnections, and the implications for enhancing global health security response. The section also highlights the study’s contributions in advancing the understanding of health security dynamics and provides practical insights for policymakers and practitioners.
4.1 Bayesian Belief Network Model for Prioritizing Indicators Influencing “Rapid Response and Mitigation”
This subsection examines a BBN model developed to prioritize indicators influencing the “rapid response and mitigation” category within the GHS index. It presents key findings regarding the critical indicators identified by the model, their relative importance, their mutual information shared with the “rapid response and mitigation” category, and sensitivity analysis results.
4.1.1 Salient Findings
Figure 1 depicts the BBN model, presenting seven indicators associated with the “rapid response to and mitigation of the spread of an epidemic” category. All seven predictor variables are directly tied to the aforementioned category. Moreover, “linking public health and security authorities” is directly linked to multiple indicators, including “trade and travel restrictions,” “access to communications infrastructure,” “emergency preparedness and response planning,” “exercising response plans,” and “emergency response operation.” Around 18% of the examined countries demonstrate a relatively high level of performance in “rapid response and mitigation,” while 16% exhibit low performance. The indicators “linking public health and security authorities” and “emergency preparedness and response planning” display a relatively high likelihood of being in a low-performance state. The seven indicators exhibit a wide variation across their probability distributions. Moreover, the BBN model depicts the interdependent nature of the indicators.
The BBN model depicted in Fig. 1 was analyzed for countries showing strong performance in the “rapid response and mitigation” category, as illustrated in Fig. 2. Among these countries, there is a high probability, exceeding 90%, associated with high-performance states for indicators like “access to communications infrastructure” and “linking public health and security authorities.” Conversely, the indicator “trade and travel restrictions” appears comparatively less impactful in this particular context. These significant differences reveal the relative importance of the seven indicators in establishing high performance in the “rapid response and mitigation” category of the GHS index.
Delving into the details of the BBN model illustrated in Fig. 1, a more in-depth analysis targeted countries demonstrating suboptimal performance in the “rapid response and mitigation” category’, as highlighted in Fig. 3. Among these countries, a significant likelihood, surpassing 90%, was observed in connection to the low-performance state for indicators such as “linking public health and security authorities” and “emergency preparedness and response planning.” Therefore, low performance in the “rapid response and mitigation” category is mainly attributed to relatively low performance in the “linking public health and security authorities” and “emergency preparedness and response planning” indicators. Noteworthy is the observation that “access to communications infrastructure” and “trade and travel restrictions” did not display a strong correlation with the low-performance state within the “rapid response and mitigation” category.
4.1.2 Relative Importance of Individual Indicators
The analysis of the BBN model, as illustrated in Fig. 2, was conducted to determine the prioritization of indicators contributing to enhanced performance within the “rapid response and mitigation” category (see Fig. 4). In these countries, a substantial probability of approximately 80% was evident, indicating the likelihood of achieving high-performance levels in pivotal indicators such as “access to communications infrastructure,” “linking public health and security authorities,” and “risk communication.” Conversely, the likelihood of attaining elevated performance in indicators like “trade and travel restrictions” and “exercising response plans” was notably diminished, below 10%.
The BBN model, as illustrated in Fig. 3, was analyzed to prioritize the indicators contributing to low performance in the “rapid response and mitigation” category (see Fig. 5). In the context of these countries, a noteworthy probability, surpassing 90%, was evident in connection to achieving low performance in pivotal indicators such as “linking public health and security authorities” and “emergency preparedness and response planning”. Conversely, the likelihood of achieving low performance in indicators like “trade and travel restrictions” and “access to communications infrastructure” was notably diminished, below 30%. Figures 4 and 5 represent significant variations in the relative importance of the seven indicators, implying the importance of exploring tailored interventions based on national performance in the “rapid response and mitigation” category.
4.1.3 Mutual Value of Information
Within BBNs, the notion of the mutual value of information pertains to the mutual influence observed among interconnected variables within the network, as detailed by Zhou et al. (2023). The acquisition of information about one variable can exert a notable impact on the probability distribution and understanding of associated variables. This reciprocal influence serves to support prediction accuracy, enhance inference capabilities, streamline decision-making processes, and facilitate sensitivity analysis. The evaluation of the mutual value of information between the “rapid response and mitigation” category and individual indicators was carried out using the Hugin software, and the findings are presented in Fig. 6. Notably, among these indicators, “linking public health and security authorities” stood out as the most prominent factor, followed by “risk communication,” while “trade and travel restrictions” and “exercising response plans” exhibited a relatively lower significance in providing information for predicting the target variable. These findings provide the relative importance of the seven indicators in predicting the precise value of the “rapid response and mitigation” category of the GHS index.
4.1.4 Sensitivity Analysis
Sensitivity analysis was performed by evaluating the impact of individual indicators on the “rapid response and mitigation” category, specifically in terms of extreme performance states. An examination of the BBN model depicted in Fig. 1 involved investigating the increase in the probability of achieving a high-performance state in the “rapid response and mitigation” category as individual indicators transitioned from low to high performance (see Fig. 7). Remarkably, “linking public health and security authorities” emerged as the most crucial indicator, demonstrating a notable probability improvement of 71%. In contrast, the indicator “trade and travel restrictions” was identified as the least significant in this assessment. These findings imply the relative importance of the seven indicators in improving performance in the “rapid response and mitigation” category of the GHS index. The relative importance probabilities can be used to establish an optimal allocation of resources to the seven indicators.
4.2 Bayesian Belief Network Model for Prioritizing Indicators Influencing the Global Health Security Index
This subsection examines a BBN model aimed at prioritizing indicators that impact the GHS index. It discusses key findings that emphasize the critical indicators identified by the model, their significance within the index, insights into how these indicators share mutual information with the index, and the outcomes of sensitivity analysis.
4.2.1 Salient Findings
Figure 8 illustrates the BBN model, showcasing seven indicators connected to the GHS index. Each of the seven predictor variables was directly associated with the index. Additionally, “linking public health and security authorities” is directly associated with “access to communications infrastructure.” The indicators “access to communications infrastructure” and “risk communication” exhibited a relatively high probability of being in a state of high performance. Unlike the BBN model shown in Fig. 1, Fig. 8 exhibits a network with fewer dependencies among the seven indicators.
Examining the BBN model as presented in Fig. 8, the focus is on countries demonstrating robust performance in the GHS index (refer to Fig. 9). Among these countries, there is a significant 95% likelihood associated with a high-performance state concerning “access to communications infrastructure.” Conversely, the indicators “trade and travel restrictions” and “exercising response plans” seem to have comparatively less influence in this specific scenario.
Exploring the intricacies of the BBN model presented in Fig. 8, a thorough examination focused on countries exhibiting less-than-optimal performance in the GHS index, as indicated in Fig. 10. In these countries, a notable 95% probability was identified in relation to the low-performance state for “linking public health and security authorities.” It is noteworthy that “access to communications infrastructure” and “risk communication” did not exhibit a strong correlation with the low-performance state of the GHS index.
4.2.2 Relative Importance of Individual Subindicators
The examination of the BBN model, depicted in Fig. 9, aimed to establish the prioritization of indicators contributing to the improved performance of the GHS index (refer to Fig. 11). In these countries, a significant probability of at least 80% was observed, indicating the likelihood of attaining high-performance levels in key indicators such as “access to communications infrastructure,” “risk communication,” and “linking public health and security authorities.” On the other hand, indicators like “trade and travel restrictions” and “exercising response plans” were relatively less crucial in this context.
The examination of the BBN model, depicted in Fig. 10, aimed to prioritize indicators contributing to the low performance of the GHS index (refer to Fig. 12). In the context of these countries, “linking public health and security authorities” emerged as the most critical indicator, followed by “emergency preparedness and response planning.” Conversely, the likelihood of experiencing low performance in indicators such as “exercising response plans,” “trade and travel restrictions,” “risk communication,” and “access to communications infrastructure” was notably diminished, below 30%. Figures 11 and 12 reveal two different ranking schemes based on the performance in the GHS index. For instance, the “risk communication” indicator drove high performance in health security, whereas low performance was mainly attributed to relatively suboptimal performance in the “linking public health and security authorities” and “emergency preparedness and response planning” indicators.
4.2.3 Mutual Value of Information
The assessment of the mutual information value between the GHS index and individual indicators was conducted using the Hugin software, and the results are depicted in Fig. 13. Significantly, among these indicators, “emergency preparedness and response planning” emerged as the most prominent factor, followed by “access to communications infrastructure,” whereas the indicator “exercising response plans” demonstrated a comparatively lower significance in providing information for predicting the target variable. This analysis can be used to prioritize the seven indicators for predicting the precise value of the GHS index.
4.2.4 Sensitivity Analysis
Conducting sensitivity analysis involved assessing the impact of individual indicators on the GHS index, specifically concerning extreme performance states. An examination of the BBN model illustrated in Fig. 8 entailed exploring the increase in the probability of achieving a high-performance state in the GHS index as individual indicators transitioned from low to high performance (see Fig. 14). “Exercising response plans” emerged as the most crucial indicator, closely followed by “emergency preparedness and response planning” and “linking public health and security authorities.” In contrast, the indicators “access to communications infrastructure” and “trade and travel restrictions” were identified as the least significant in this evaluation.
4.3 Discussion and Implications
The analysis revealed a relatively low significance of “trade and travel restrictions” in the “rapid response to and mitigation of the spread of an epidemic” category, contradicting previous studies that emphasized the role of such restrictions (Lee et al. 2020). This finding may be attributed to the dynamic nature of the pandemic, where factors like “linking public health and security authorities” gained importance in real-time responses. This highlights the need for an adaptive approach in selecting epidemic response strategies.
In terms of the mutual value of information, the assessment identified “linking public health and security authorities” as the most influential factor, aligning with the network nature of health security (Bell and Nuzzo 2021). The prominence of “risk communication” further highlighted the complex dynamics of information flow in health security (Husnayain et al. 2020).
Sensitivity analysis underscored the pivotal role of “linking public health and security authorities” for the “rapid response to and mitigation of the spread of an epidemic” category. The limited significance of “trade and travel restrictions” revealed through sensitivity analysis contributed to the literature, suggesting the necessity for a robust approach in epidemic response.
Moving to the GHS index, the examination of countries with robust performance highlighted the overwhelming influence of “access to communications infrastructure,” reinforcing the importance of effective communication channels (Razavi et al. 2020). Conversely, the relatively lower impact of “trade and travel restrictions” suggested a need for reevaluation in the context of health security. The prioritization of indicators for the GHS index further confirmed the significance of “access to communications infrastructure,” “risk communication,” and “linking public health and security authorities”. This aligned with existing literature emphasizing the pivotal role of these indicators in health security (Khatri et al. 2023). The mutual value of information assessment for the GHS index identified “emergency preparedness and response planning” as the most prominent factor, emphasizing the importance of preparedness in achieving a robust health security index. The sensitivity analysis reinforced the critical role of “exercising response plans,” underscoring the importance of proactive planning in health security.
The theoretical implications of the study’s findings are significant in advancing our understanding of the dynamics within the “rapid response to and mitigation of the spread of an epidemic” category. The unique result regarding the relatively low significance of “trade and travel restrictions” challenges existing theoretical frameworks that emphasize these measures as central to the epidemic response. This suggests a reevaluation of traditional assumptions and highlights the need for a better understanding of the factors that influence rapid response strategies. The study underscored the dynamic and context-dependent nature of GHS, urging scholars to incorporate real-time dynamics and evolving priorities into theoretical models.
The study’s emphasis on BBNs and their application in modeling complex relationships has theoretical implications for the field of GHS. The choice of BBNs acknowledges the complexities and uncertainties inherent in the system, aligning with contemporary perspectives that advocate for probabilistic modeling approaches (Qazi et al. 2022). This contributes to the ongoing discourse on methodological advancements in health security research, urging researchers to adopt models that capture the dynamic interplay of various factors.
From a managerial standpoint, the study provided valuable insights for policymakers and health security practitioners. The unique finding regarding the limited impact of “trade and travel restrictions” in the “rapid response to and mitigation of the spread of an epidemic” category challenged the managerial emphasis on these measures. Policymakers may need to reconsider the allocation of resources and efforts, directing them toward more influential factors like “linking public health and security authorities” and “emergency preparedness and response planning.” This shift in focus could enhance the efficiency and effectiveness of health security strategies, aligning them with the dynamic realities of global health crises. Moreover, the prioritization of indicators within both the “rapid response to and mitigation of the spread of an epidemic” category and the GHS index offered managerial insights. For instance, the importance of “access to communications infrastructure” and “linking public health and security authorities” underscored the critical role of communication networks in facilitating effective health security responses. Policymakers and practitioners should prioritize investments in strengthening communication infrastructures and fostering collaboration between health and security authorities.
The study’s managerial implications extend to the sensitivity analysis results, which highlight the critical role of proactive planning, as evidenced by “emergency preparedness and response planning” and “exercising response plans”. Policymakers should emphasize the development and regular testing of response plans to ensure preparedness for various scenarios, contributing to a more resilient health security framework.
4.4 Contribution
This study made several noteworthy contributions to the field of GHS research. First and foremost, it advanced the understanding of the complex dynamics involved in the “rapid response to and mitigation of the spread of an epidemic” category. The finding regarding the limited impact of “trade and travel restrictions” contributes to existing theoretical frameworks. By shedding light on the evolving priorities in epidemic response, this study encourages scholars to reevaluate traditional approaches and incorporate real-time, context-dependent factors into their theoretical models. This nuanced understanding contributes to the theoretical richness of the field, encouraging a more adaptive and informed approach to GHS.
Second, the study introduced BBNs as a powerful methodological tool for modeling the complex relationships within the “rapid response to and mitigation of the spread of an epidemic” category of the GHS index. The application of BBNs provided a visual and probabilistic representation of the interconnected variables, allowing for a comprehensive examination of their influences. This methodological choice aligned with contemporary perspectives advocating for probabilistic modeling approaches, contributing to the methodological diversity within health security research. The study showed the strengths of BBNs in capturing uncertainties and dynamic relationships, providing researchers with a valuable tool for future investigations into complex global health systems.
Last, the managerial implications derived from the study’s findings offered practical guidance for policymakers and health security practitioners. The prioritization of indicators, the reassessment of the significance of certain measures, and the emphasis on communication networks and proactive planning contribute to actionable insights. Policymakers can use these contributions to refine strategies, allocate resources more effectively, and strengthen preparedness for future health crises.
5 Conclusion
This study provided valuable insights into the dynamics of the “rapid response to and mitigation of the spread of an epidemic” category within the GHS index. The revelation regarding the limited impact of “trade and travel restrictions” calls for a reevaluation of traditional approaches, emphasizing the necessity for adaptive, context-aware strategies in epidemic response. This study contributed to the theoretical understanding of GHS, advocating for more dynamic models that incorporate real-time factors. However, this study has certain limitations. The effectiveness of the BBN model is dependent on data availability and reliability, introducing potential uncertainties. Additionally, while BBNs capture complex relationships, they may oversimplify certain aspects of the complex health security landscape. The study’s findings are specific to the year 2021, highlighting the need for continuous research efforts to capture the evolving nature of global health challenges.
Future research in GHS should build upon the foundations laid by this study. An in-depth exploration into the unique results, such as the limited impact of “trade and travel restrictions,” may be considered in future studies. Comparative studies across different temporal and geographical contexts can provide a more comprehensive understanding of evolving health security priorities. Advances in modeling methodologies, including the integration of BBN with artificial intelligence algorithms, may offer further opportunities to enhance predictive accuracy and capture dynamic interdependencies more comprehensively. Furthermore, future research could delve into the practical implementation and effectiveness of specific interventions suggested by this study, such as communication infrastructure improvements and proactive planning, offering actionable insights for policymakers and practitioners.
The insights from this study offer valuable guidance for future applications, particularly at the regional or national level. By employing the proposed approach, customized models can be developed to prioritize indicators within the “rapid response to and mitigation of the spread of an epidemic” category specific to countries or states. Using GHS index data, regions or countries can identify their unique priorities within this category, emphasizing indicators such as “emergency preparedness and response planning” or “access to communications infrastructure” based on regional needs and vulnerabilities. Moreover, by developing adaptive strategies grounded in real-time dynamics, policymakers at the regional or national level can enhance their response to global health crises.
Notes
GeNIe 2.0. The Decision Systems Laboratory, GeNIe and SMILE.
References
Ab Aziz, N.F., F.W. Akashah, and A. Abdul Aziz. 2019. Conceptual framework for risk communication between emergency response team and management team at healthcare facilities: A Malaysian perspective. International Journal of Disaster Risk Reduction 41: Article 101282.
Al Nuairi, A., M.C.E. Simsekler, A. Qazi, and A. Sleptchenko. 2023. A data-driven Bayesian belief network model for exploring patient experience drivers in healthcare sector. Annals of Operations Research. https://doi.org/10.1007/s10479-023-05437-9.
Albreiki, S., M.C.E. Simsekler, A. Qazi, and A. Bouabid. 2024. Assessment of the organizational factors in incident management practices in healthcare: A tree augmented naive Bayes model. PLOS ONE 19: Article e0299485.
Assefa, Y., P.S. Hill, C.F. Gilks, W.V. Damme, R.V.D. Pas, S. Woldeyohannes, and S. Reid. 2021. Global health security and universal health coverage: Understanding convergences and divergences for a synergistic response. PLOS ONE 15: Article e0244555.
Ayre, K.K., C.A. Caldwell, J. Stinson, and W.G. Landis. 2014. Analysis of regional scale risk of whirling disease in populations of Colorado and Rio Grande cutthroat trout using a Bayesian belief network model. Risk Analysis 34: 1589–1605.
BayesFusion. 2017. GeNIe modeler-user manual. https://support.bayesfusion.com/docs/GeNIe/. Accessed 24 Jul 2017.
Bell, J.A., and J.B. Nuzzo. 2021. Global Health Security Index: Advancing collective action and accountability amid global crisis. www.GHSIndex.org. Accessed 25 Nov 2023.
Bruinen de Bruin, Y., A.-S. Lequarre, J. Mccourt, P. Clevestig, F. Pigazzani, M. Zare Jeddi, C. Colosio, and M. Goulart. 2020. Initial impacts of global risk mitigation measures taken during the combatting of the COVID-19 pandemic. Safety Science 128: Article 104773.
Chiossi, S., S. Tsolova, and M. Ciotti. 2021. Assessing public health emergency preparedness: A scoping review on recent tools and methods. International Journal of Disaster Risk Reduction 56: Article 102104.
Cook, J.D., D.M. Williams, D.P. Walsh, and T.J. Hefley. 2023. Bayesian forecasting of disease spread with little or no local data. Scientific Reports 13: Article 8137.
Cooper, A., F. Mazzeo, P. Waterson, M.S. Young, and D. Louis. 2023. The use of Bayesian belief networks (BBNs) to probe deeper into railway safety management systems – Two studies from Great Britain and Italy. Applied Ergonomics 109: Article 103968.
Cox, R., C.W. Revie, D. Hurnik, and J. Sanchez. 2016. Use of Bayesian belief network techniques to explore the interaction of biosecurity practices on the probability of porcine disease occurrence in Canada. Preventive Veterinary Medicine 131: 20–30.
Fakhruddin, B., K. Blanchard, and D. Ragupathy. 2020. Are we there yet? The transition from response to recovery for the COVID-19 pandemic. Progress in Disaster Science 7: Article 100102.
Gwenzi, W., E.C. Skirmuntt, T. Musvuugwa, C. Teta, D. Halabowski, and P. Rzymski. 2022. Grappling with (re)-emerging infectious zoonoses: Risk assessment, mitigation framework, and future directions. International Journal of Disaster Risk Reduction 82: Article 103350.
Hossain, N.U.I., F. Nur, S. Hosseini, R. Jaradat, M. Marufuzzaman, and S.M. Puryear. 2019. A Bayesian network based approach for modeling and assessing resilience: A case study of a full service deep water port. Reliability Engineering & System Safety 189: 378–396.
Hosseini, S., and D. Ivanov. 2020. Bayesian networks for supply chain risk, resilience and ripple effect analysis: A literature review. Expert Systems with Applications 161: Article 113649.
Husnayain, A., A. Fuad, and E.C.-Y. Su. 2020. Applications of Google Search Trends for risk communication in infectious disease management: A case study of the COVID-19 outbreak in Taiwan. International Journal of Infectious Diseases 95: 221–223.
Ji, Y., J. Shao, B. Tao, H. Song, Z. Li, and J. Wang. 2021. Are we ready to deal with a global COVID-19 pandemic? Rethinking countries’ capacity based on the Global Health Security Index. International Journal of Infectious Diseases 106: 289–294.
Kachali, H., I. Haavisto, R.-L. Leskelä, A. Väljä, and M. Nuutinen. 2022. Are preparedness indices reflective of pandemic preparedness? A COVID-19 reality check. International Journal of Disaster Risk Reduction 77: Article 103074.
Khatri, R.B., A. Endalamaw, D. Erku, E. Wolka, F. Nigatu, A. Zewdie, and Y. Assefa. 2023. Preparedness, impacts, and responses of public health emergencies towards health security: Qualitative synthesis of evidence. Archives of Public Health 81: Article 208.
Lee, K., C.Z. Worsnop, K.A. Grépin, and A. Kamradt-Scott. 2020. Global coordination on cross-border travel and trade measures crucial to COVID-19 response. The Lancet 395: 1593–1595.
Leichtweis, B.G., L. De Faria Silva, F.L. Da Silva, and L.A. Peternelli. 2021. How the global health security index and environment factor influence the spread of COVID-19: A country level analysis. One Health 12: Article 100235.
Lencucha, R., and S. Bandara. 2021. Trust, risk, and the challenge of information sharing during a health emergency. Globalization and Health 17: Article 21.
Liao, Y., B. Xu, X. Liu, J. Wang, S. Hu, W. Huang, K. Luo, and L. Gao. 2018. Using a Bayesian belief network model for early warning of death and severe risk of HFMD in Hunan Province, China. Stochastic Environmental Research and Risk Assessment 32: 1531–1544.
Liao, Y., B. Xu, J. Wang, and X. Liu. 2017. A new method for assessing the risk of infectious disease outbreak. Scientific Reports 7: Article 40084.
Liu, L.-Y., W.-N. Wu, and D.A. Mcentire. 2021. Six Cs of pandemic emergency management: A case study of Taiwan’s initial response to the COVID-19 pandemic. International Journal of Disaster Risk Reduction 64: Article 102516.
Marcot, B.G., and A.M. Hanea. 2021. What is an optimal value of k in k-fold cross-validation in discrete Bayesian network analysis?. Computational Statistics 36: 2009–2031.
Motwani, A., P.K. Shukla, and M. Pawar. 2022. Ubiquitous and smart healthcare monitoring frameworks based on machine learning: A comprehensive review. Artificial Intelligence in Medicine 134: Article 102431.
Nawaz, A., X. Su, M.Q. Barkat, S. Asghar, A. Asad, F. Basit, S. Iqbal, H. Zahoor, and S.A. Raheel Shah. 2020. Epidemic spread and its management through governance and leadership response influencing the arising challenges around COVID-19 in Pakistan – A lesson learnt for low income countries with limited resource. Frontiers in Public Health 8: Article 573431.
Pereira, J., P. Contreras, D.C. Morais, and P. Arroyo-López. 2022. Multi-criteria ordered clustering of countries in the Global Health Security Index. Socio-Economic Planning Sciences 84: Article 101331.
Qazi, A. 2024. Exploring the impact of global competitiveness pillars on sustainable development. Environmental Impact Assessment Review 105: Article 107404.
Qazi, A., and M.S. Khan. 2021. Exploring probabilistic network-based modeling of multidimensional factors associated with country risk. Risk Analysis 41: 911–928.
Qazi, A., and M.C.E. Simsekler. 2020. Assessment of humanitarian crises and disaster risk exposure using data-driven Bayesian networks. International Journal of Disaster Risk Reduction 52: Article 101938.
Qazi, A., M.C.E. Simsekler, and M. Akram. 2021. Efficacy of early warning systems in assessing country-level risk exposure to COVID-19. Geomatics, Natural Hazards and Risk 12: 2352–2366.
Qazi, A., M.C.E. Simsekler, and B. Gaudenzi. 2022. Prioritizing multidimensional interdependent factors influencing COVID-19 risk. Risk Analysis 42: 143–161.
Razavi, A., N. Erondu, and E. Okereke. 2020. The Global Health Security Index: What value does it add? BMJ Glob Health 5(4): Article e002477.
Sarmiento, J.P., G. Hoberman, M. Ilcheva, A. Asgary, A.M. Majano, S. Poggione, and L.R. Duran. 2015. Private sector and disaster risk reduction: The cases of Bogota, Miami, Kingston, San Jose, Santiago, and Vancouver. International Journal of Disaster Risk Reduction 14: 225–237.
Simsek, S., A. Dag, T. Tiahrt, and A. Oztekin. 2021. A Bayesian belief network-based probabilistic mechanism to determine patient no-show risk categories. Omega 100: Article 102296.
Simsekler, M.C.E., and A. Qazi. 2022. Adoption of a data-driven Bayesian belief network investigating organizational factors that influence patient safety. Risk Analysis 42: 1277–1293.
Simsekler, M.C.E., C. Rodrigues, A. Qazi, S. Ellahham, and A. Ozonoff. 2021. A comparative study of patient and staff safety evaluation using tree-based machine learning algorithms. Reliability Engineering & System Safety 208: Article 107416.
Vizanko, B., L. Kadinski, A. Ostfeld, and E.Z. Berglund. 2024. Social distancing, water demand changes, and quality of drinking water during the COVID-19 pandemic. Sustainable Cities and Society 102: Article 105210.
Wennman, I., C. Jacobson, E. Carlström, A. Hyltander, and A. Khorram-Manesh. 2022. Organizational changes needed in disasters and public health emergencies: A qualitative study among managers at a major hospital. International Journal of Disaster Risk Science 13(4): 481–494.
Werner, C., T. Bedford, R.M. Cooke, A.M. Hanea, and O. Morales-Nápoles. 2017. Expert judgement for dependence in probabilistic modelling: A systematic literature review and future research directions. European Journal of Operational Research 258: 801–819.
Zawadzki, M., and G. Montibeller. 2023. A framework for supporting health capability-based planning: Identifying and structuring health capabilities. Risk Analysis 43: 78–96.
Zhang, X.-X., Y.-Z. Jin, Y.-H. Lu, L.-L. Huang, C.-X. Wu, S. Lv, Z. Chen, H. Xiang, and X.-N. Zhou. 2023. Infectious disease control: From health security strengthening to health systems improvement at global level. Global Health Research and Policy 8: Article 38.
Zhou, Z., X. Yu, Z. Zhu, D. Zhou, and H. Qi. 2023. Development and application of a Bayesian network-based model for systematically reducing safety risks in the commercial air transportation system. Safety Science 157: Article 105942.
Acknowledgments
The work in this paper was supported, in part, by the Faculty Research Grant (FRG23-E-B91) from the American University of Sharjah. This paper represents the opinions of the authors and does not mean to represent the position or opinions of the American University of Sharjah.
Author information
Authors and Affiliations
Corresponding author
Rights and permissions
Open Access This article is licensed under a Creative Commons Attribution 4.0 International License, which permits use, sharing, adaptation, distribution and reproduction in any medium or format, as long as you give appropriate credit to the original author(s) and the source, provide a link to the Creative Commons licence, and indicate if changes were made. The images or other third party material in this article are included in the article's Creative Commons licence, unless indicated otherwise in a credit line to the material. If material is not included in the article's Creative Commons licence and your intended use is not permitted by statutory regulation or exceeds the permitted use, you will need to obtain permission directly from the copyright holder. To view a copy of this licence, visit http://creativecommons.org/licenses/by/4.0/.
About this article
Cite this article
Qazi, A., Simsekler, M.C.E. & Al-Mhdawi, M.K.S. Prioritizing Indicators for Rapid Response in Global Health Security: A Bayesian Network Approach. Int J Disaster Risk Sci (2024). https://doi.org/10.1007/s13753-024-00570-w
Accepted:
Published:
DOI: https://doi.org/10.1007/s13753-024-00570-w