Past, present, and future: a situational analysis of infectious disease modelling in Thailand

Summary

Infectious disease modelling (IDM) is a useful tool supporting evidence to inform policies on disease outbreaks. Understanding situation, existing capacities and needs will enable countries to prepare and use the evidence derived from IDM for future outbreaks. This report maps Thailan's IDM landscape, identifies key stakeholders, and provides recommendations to develop a supportive ecosystem. We found that there is a moderate capacity to conduct and use IDM in Thailand. Users of IDM are spread across ministries and government level, while IDM evidence suppliers operate in departments in a few universities. Key challenges concern availability and quality of data, human resource capacity, integration of initiatives and communication mechanisms between evidence users and providers, and sustainable funding for IDM activities. Investing in human and data infrastructure, including IDM ecosystem development, could enhance Thailand's capacity to synthesise and use evidence for future outbreak preparedness, while also contributing to regional efforts in health security and outbreak response.

Funding

This study was supported by a grant from the Rockefeller Foundation [2022 ARO 004] and the National Science, Research and Innovation Fund, Thailand Science Research and Innovation (TSRI).

Introduction

Infectious disease modelling (IDM) refers to the mathematical description of the spread of infection.¹ Infectious diseases in this report refers to common infectious diseases (whether in humans or animals).² The report primarily focuses on transmission modelling in public health, which helps understand the disease natural history and evaluate the impact of interventions (e.g., vaccination, social distancing, facemask wearing, or lockdowns), depending on the choice of mathematical model.³ Many countries adopted this type of modelling to inform their current and future decision-making in terms of predicting, preventing, and controlling the spread of the disease.^4,5 Examples of the use of such modelling include predicting disease trends, evaluating impacts of safety and control measures imposed, and exploring uncertainties when certain policy decisions were to change. During the coronavirus disease 2019 (COVID-19) pandemic, these models were used to inform policies to prevent the spread of the disease around the world.⁶

The knowledge on the use, practice, and development of IDM in the Southeast Asian region is relatively limited, even though the region has been the epicentre of several outbreaks.7, 8, 9 In Thailand, there is a growing number of studies that have applied IDM,10, 11, 12 however, there is limited documented literature on its use in policy. Increasing attention has been paid to creating an enabling environment for the routine use of IDM in policy; conducting a situation analysis of existing capacities has been identified as a starting point for developing a coherent approach for the routine use of IDM for policy.¹³

In this context, this report aims to describe and understand the current capacity and the situation of IDM in Thailand. Adapted from a conceptual framework previously used for Health Technology Assessment (HTA),¹⁴ we aim to (i) understand the landscape of IDM, including current gaps and challenges for conducting IDM analyses in Thailand; (ii) identify the current roles and profiles of stakeholders involved (evidence providers, users, etc.) in Thailand; and (iii) provide recommendations to build an ecosystem for synthesising and using evidence from IDM, based on Thailand's experience.

Relevant information were synthesised from the document review, organisational survey, and key informant interviews (KIIs) (see detailed methods in Supplementary 1 and relevant data in Supplementary 2). Insights cover key elements of the need, demand, and supply of IDM. The current gaps to address and opportunities to leverage were also presented to formulate key recommendations for relevant stakeholders in Thailand.

The need for and importance of IDM in Thailand

We found that IDM can be useful as a tool to inform policy. In particular, several stakeholders elaborated that modelling work can be used to understand the progress of outbreaks, predict, and prepare for pandemic responses in the future. Although IDM is a powerful tool, it was also recognised that it should be used with caution, especially in a context which has limited infrastructure and poor-quality data. These can affect the accuracy of a model and its reliability for policy implications.

Key infectious diseases of priority in Thailand were identified, for which IDM would be useful in designing policy responses. These include emerging diseases (Table 1) such as the COVID-19, human immunodeficiency viruses (HIV), respiratory diseases such as influenza, as well as diseases with high R₀ values (the basic reproduction number/rate, indicating the contagiousness and transmissibility of infectious pathogens). Diseases transmitted between people and animals, including vector borne diseases, were also emphasised, conveying the need to develop an understanding of disease incidence in both humans and animals, for example rabies, as this is relatively common in the epidemiological landscape of infectious diseases in Thailand.¹⁵

Table 1.

Priority infectious diseases for Thailand identified through the survey responses.

Disease	Reason for prioritisation
COVID-19	The outbreak has not yet been fully contained. Knowledge products and resources addressing this disease can play a crucial role in enhancing preparedness for future emerging diseases.
Human immunodeficiency viruses (HIV) Respiratory syncytial virus (RSV)	There is an ever-increasing number of people suffering from these diseases.
Respiratory tract infections	This is a group of diseases that occur regularly and are affected by seasons and time changes.
Endemic disease (e.g., influenza, Malaria, Dengue)	It is a group of diseases that occur regularly and are affected by seasonal changes.
Zika virus disease	It is a disease in which infected people rarely show symptoms. As a result, the data entering the surveillance system is substantially less than the actual number, making it seem like it is not a problem. One issue is that if a pregnant woman is infected, it can cause the foetus to have microcephaly and developmental delays.
Tuberculosis (TB)	It has not been completely eradicated and has a high burden on the Thai population and society.

Drivers for the use of IDM in Thailand

Several factors have contributed to the increased use of IDM in Thailand. Especially, high burden and severity of diseases such as COVID-19 led to the establishment of temporary mechanisms for requesting evidence from IDM to predict outbreaks and suggest appropriate policy measures. For example, the creation of specific Working Groups (e.g., the Centre for COVID-19 Situation Administration: CCSA) under the Order of the Prime Minister No. 6/2563 increased the demand of epidemiological evidence as part of outbreak control efforts.¹⁶

Prior to the COVID-19 pandemic, a gradual increase in technical personnel, expertise, and technological development over times was identified to be one of the factors facilitating the local development and use of modelling.^17,18 Participating in more trainings/workshops and conducting international research collaborations has strengthened technical capacity in this sphere (see also in Supplement 2).

Interestingly, IDM has been used relatively more widely as a tool for generating evidence to inform policy recommendations for animal health, compared to its use for human health. Animal and farming industry data are routinely collected for developing livestock standards and certification and are therefore more readily available for use. Additionally, livestock and farming data are perceived less sensitive as compared to human data. This might, to a certain degree, support the practice and use of IDM in the field of animal health in Thailand.

Insights derived from IDM can be employed for either informing policy recommendations (policy-oriented use) or academic use. The choice between these applications depends on the requests from the end-users of the models and the initial objectives of the modelling studies.

The users of IDM were predominantly at the national level. These include the Department of Disease Control (DDC), Ministry of Public Health (MOPH); the Department of Livestock Development (DLD) and the Department of Fisheries, Ministry of Agriculture and Cooperatives (MOAC); the Department of National Parks (DNP), Wildlife and Plant Conservation and the Department of Marine and Coastal Resources, Ministry of Natural Resources and Environment (MONRE); the Ministry of Interior (MOI); the Ministry of Foreign Affairs (MFA); and the Ministry of Tourism and Sports (MOTS).

However, at the provincial level, local Public Health Agencies or Offices were also identified as potential users. IDM could be used for surveillance and early control of local outbreaks. Moreover, non-governmental organisations (NGOs) and private sector organisations dealing with epidemiology were also highlighted as additional potential users. However, it is still unclear how this group of stakeholders use insights from IDM.

The policy questions that decision-makers typically sought answers were typically to support effective preparedness and response to outbreaks. For example, these questions included estimating the potential number of affected populations in the present and subsequent years, developing strategic plans for population movement during outbreaks, and ensuring the availability of essential goods and supplies while maintaining necessary safety measures. Anticipating these questions in advance is imperative as to facilitate modellers in developing timely and user-aligned models. These questions can be categorised into three groups, as presented in Table 2.

Table 2.

Common policy questions to support outbreak management.

Category/group	Description	Example of relevant questions
I. Situation Assessment	This is to understand potential future outbreak trends based on historical patterns and recurring conditions. It is particularly relevant for common and seasonal infectious diseases.	•How long is the outbreak expected to last? •How quickly will the outbreak spread in the absence of public health interventions? •How many people are currently infected, and how many are likely to be infected over time?
II. Policy Impact & Intervention Assessment	This type of question addresses decision-making challenges with potential economic and social implications. It typically involves comparing two or more policy options to determine which interventions are most effective or context appropriate in controlling outbreaks. This approach is especially relevant for high-priority diseases or those with a significant disease burden.	•Should travel bans remain in place during COVID-19? •What are the impacts of border closures on public safety and the economy? •Which population groups should be prioritised for vaccination? •To what extent can social distancing help slow the spread of the disease?
III. After-Action-Review (AAR) Assessment	This group is similar to the Policy Impact & Intervention Assessment but takes a retrospective approach. It focuses on analysing past scenarios to evaluate the effectiveness of measures that were implemented. These questions help synthesise lessons learned and guide decisions on whether similar measures should be adopted in future outbreaks.	•What was the level of public compliance with measures such as mask mandates, social distancing, and quarantine? •Did policy measures lead to unintended social or economic consequences (e.g., mental health challenges, income loss) that influenced overall outbreak control? •How would a delay of one to two weeks in implementing key measures (e.g., isolation, contact tracing) have affected peak case numbers and outbreak duration?

Workflow for translating evidence from IDM into policy use

Although there is no legal requirement for using IDM in policy, a common mechanism for evidence translation is to set up a statutory sub-committee dedicated to overseeing the management of a specific disease. As part of the process, academics and modellers are usually recruited in a team to prepare relevant inputs and evidence for decision-makers, when implementing public health measures. There are principally two approaches to translating evidence to policy.

The process for using modelling to manage diseases related to human health typically followed a top-down approach. With a case example from the DDC, model development for policy use involves either leveraging in-house capacity when feasible or adapting existing models from other countries to respond to urgent policy questions. Indeed, urgent requests and less complicated models were usually developed in-house and could be part of routine analysis, while more complicated models often involved stakeholder and expert consultations. As illustrated in Fig. 1, the process typically begins when decision-makers request evidence support to inform key policy questions. These questions, if not adequately specified, would need to be refined and translated into modelling/research questions. Once models are developed, the results are shared with relevant teams/users. The duration for work execution depends on the complexity of the models. Expert consultations might be required for areas where databases are not available, and verification is needed, prior to presenting the results to relevant users.

Fig. 1.

Overview of the process flow for modelling work addressing diseases of human health. Light yellow indicates needs where modelling work could help address. Dark blue represents the demand side or users of evidence from modelling. Purple denotes the supply side, including entities involved in evidence synthesis and provision. Orange highlights interactions with the public and/or broader stakeholders. ∗This can be a sub-committee established for certain outbreaks.

The process of using modelling can also follow a bottom-up approach as seen in the case of animal-focused IDM. An example can be drawn from the DLD. The need for evidence and research questions typically arises from the real-world experience of veterinarians, practitioners, and implementors (see Fig. 2). Following this, relevant resources, including funding and human resources with the necessary expertise, need to be secured. Once these resources are available, discussions between funders, resource persons, and technical staff are sought to ensure alignment of objectives and research plans. Upon agreement, the study is conducted, and the output is produced. The research output is then disseminated to relevant users such as policymakers. Typically, the DLD conducts the study first before presenting it to policymakers, allowing them to visualise the implications of the findings.

Fig. 2.

Overview of the process flow for modelling work addressing diseases of animal health. Light yellow indicates needs where modelling work could help address. Dark blue represents the demand side or users of evidence from modelling. Purple denotes the supply side, including entities involved in evidence synthesis and provision. Orange highlights interactions with the public and/or broader stakeholders.

In the context of articles published in English up to the year 2023, a full-text search in PubMed resulted in 383 articles. There was an increasing trend in the number of published studies in epidemiological and mathematical modelling relating to Thailand, from 1970 to 2023 (Fig. 3). However, only 96 studies specifically addressed infectious diseases using epidemiological or mathematical modelling techniques (Supplement 2). Among all included articles, the majority were conducted in the Thai context and were either led or co-led by Thai researchers. Approximately 90% of the studies focused on infectious diseases affecting human health, while those addressing animal health accounted for 6%. The remaining articles (4%) either focused on plant diseases or did not specify diseases (e.g., methodological introductions or practical explanations).

Fig. 3.

A graph illustrating the overall number of dynamic modelling studies published in international databases between 1970 and 2023 (the search teams available inSupplementary 1).

There are different types of dynamic models,² though the ones most commonly developed and used in Thailand were compartmental models. The compartmental models were typically Susceptible-Infectious-Recovered (SIR) or Susceptible-Exposed-Infectious-Recovered (SEIR) models. To a lesser extent, spatial models and agent-based models were also used.

Various types of funding sources support modelling work in Thailand, depending on the area of focus. Among the studies reviewed, 35% were funded by domestic sources, 41% by international organisations, and 7% received co-funding from both domestic and international bodies. However, 17% of the studies did not disclose any funding information or reported receiving no funding support.

For modelling studies concerning animal health, funding often comes from international organisations, as compared to internal sources such as the government or local funding agencies. For example, specialised agencies of the United Nations (UN), Food and Agriculture Organisation (FAO) were identified as the main funders for animal-health focused studies, as well as the United States Agency for International Development (USAID). However, with regards to human health, funding support can be from both local and international sources. International funders to note are the U.S. Centres for Disease Control and Prevention (CDC) and the Rockefeller Foundation. Some examples of local sources include the MOPH and its affiliated funding agencies (e.g., Thailand Health Systems Research Institute or HSRI), universities, and the Ministry of Higher Education, Science, Research and Innovation (MHESI). The list of funders can be found in Supplement 2.

Although several domestic funding sources were identified, concerns were raised regarding the short duration of these grants. Short-term funding was noted as a significant constraint, given that complex dynamic modelling typically requires extended timeframes for development, calibration, and validation. It is also noteworthy that various agencies may have different expectations regarding the outcomes of modelling studies. As such, there may be a perception by some that modelling studies are less tangible (in terms of producing physical output), compared to the development of novel technologies. This can create additional hurdles for researchers when seeking support from funders.

The primary contributors of evidence derived from IDM in Thailand were predominantly affiliated with academia. Notably, local universities and independent or international research agencies, such as the Mahidol-Oxford Research Unit (MORU) and International Health Policy Programme (IHPP) also play a key role in conducting modelling studies. Additionally, the Division of Epidemiology within the Thai DDC also has the internal capacity to conduct essential analyses, particularly when rapid responses to policy question are required.

Interestingly, to a certain extent, recommendations informed by modelling studies from international agencies such as the World Health Organisation (WHO) and the CDC also influenced some policies in Thailand.^19,20 For example, the prioritisation of COVID-19 vaccines for health professionals during the initial phase of the vaccine rollout was in line with the WHO Strategic Advisory Group of Experts (SAGE) recommendations.²¹

Several universities in Thailand offer taught courses which involve elements of epidemiology and dynamic modelling. These courses mostly sit within the Faculty of Sciences (e.g., physics, mathematics), the Faculty of Medicine and/or Public Health. Most of these universities are located in greater Bangkok and nearby provinces, with a lesser extent sited at the provincial level. These include institutions such as Mahidol University (MU), Chulalongkorn University (CU),²² King Mongkut's Institute of Technology Ladkrabang (KMITL), King Mongkut's University of Technology Thonburi (KMUTT), King Mongkut's University of Technology North Bangkok (KMUTNB), Silpakorn University (SU), Prince of Songkla University (PSU), Khon Kaen University (KKU), Chiang Mai University (CMU). While a reasonable number of universities offer relevant academic programs, the composition of local expert teams convened to address policy questions tends to be less diverse in institutional representation. Although it seems that IDM has health-related implications, the involvement of relevant experts often extends beyond those from the health sector. For example, local expert teams frequently include modellers with backgrounds in physics and mathematics.

Challenges, gaps, and opportunities to the advancement of IDM use and development

One of the most prominent challenges for IDM in Thailand is the tendency of relevant stakeholders or key actors in the field to work in silos. The absence of a formal mechanism or network connecting researchers and modellers in the country contributes to communication gaps, leading to duplicated efforts and inefficient allocation of resources needed for studies. Moreover, without the institutionalisation of IDM, engagement between users and modellers tends to occur on an ad hoc basis, highlighting the need for more systematic integration and collaboration.

Data scarcity remains a key challenge in resource-constrained setting, including Thailand. This extends not only to the availability of data, but also to the accessibility of secure, high-quality databases. Several datasets were perceived as particularly difficult to access. Foremost among these were demographic data (e.g., population, birth/death rates specified by population group), including mobility data tracking the movement of travellers entering and leaving the country. These parameters are often dispersed across multiple data custodians, further complicating access. Following closely were epidemiological data, covering surveillance information such as the number of populations infected, recovered, or deceased. These also included clinical data such as disease's incubation periods, the efficacy of treatment or disease control measures, and other healthcare data which involve the duration of treatment or hospitalisation, and respectively the capacity of healthcare facilities to accommodate patients.

It is worth mentioning that Thailand has implemented the Personal Data Protection Act B.E. 2562 (2019) or PDPA. However, this may not necessarily pose additional challenges, as the data required for model development are typically not at the personal level. Instead, a more immediate concern revolves around ensuring and maintaining the quality of the available data.

Thailand's proficiency in dynamic modelling has shown a gradual improvement. This is evidenced by the growing number of peer-reviewed publications, aligning with a global trend presented elsewhere,²³ and the existence of numerous ongoing training initiatives. For example, the DLD delivers their annual training on epidemiology and modelling relating to animal health as part of the personnel development. Similarly, the DDC has integrated epidemiology and elements of IDM into its Field Epidemiology Training Program (FETP) courses, illustrating a commitment to skill development in this critical domain.

Nevertheless, there remains a need to continue harnessing technical capacity, raising awareness, and promoting literacy among local stakeholders. Academia was perceived as the top priority for technical training, including effective communication strategies, due to the high demand for good-quality evidence to inform decision-makers. Evidence users, particularly decision-makers, were ranked as the second highest priority. Essentially, decision-makers ought to understand how to interpret and appropriately use evidence from IDM, especially when faced with uncertainty. This knowledge could help them make better informed decisions. Additionally, university students were highlighted as another key group of stakeholders that would benefit from relevant training. Exposing them to the concept of modelling would raise awareness and enhance literacy, contributing to future learning and human resource development.

While Thailand has witnessed many infectious disease outbreaks, these events could be converted into opportunities, to build the infrastructure and strengthen Thailand's position in understanding of infectious diseases. This does present Thailand with unique opportunities such as becoming a repository site for disease data and surveillance, enabling in-depth studies of infectious diseases. Moreover, the increased occurrence of outbreaks can attract and engage more international experts, building collaboration and knowledge exchange in the field. Thus, Thailand has the potential to position itself as a hub for expertise and data in infectious disease research, contributing to global efforts in understanding and combating such health challenges.

Another highlighted opportunity lies in an increase in research collaboration with international experts which Thailand has developed over the years. According to the document review conducted for this report, the UK and the US were the top two countries of affiliated organisations of lead/co-authors in the identified publications (Supplement 1). This is in line with a global research activity on mathematical modelling of disease transmission and control, where both countries were also among the top five actively publishing research on IDM, and that the UK featured more than 70% of its publications with contributions from international collaborators.²³

Recognising the potentials of such international collaborations, it is crucial to explore ways to structure them to ensure mutual benefit for all partners. Establishing platforms and databases that facilitate the creation of a knowledge pool within the country and promote engagement is also imperative. This approach not only enhances the effectiveness of collaborative efforts but also contributes to sustained knowledge exchange in the field of infectious disease research.

This report highlights the need for a concerted effort to institutionalise the use of IDM. There are a range of suppliers of such evidence, and there are avenues to use this type of evidence. However, there remain challenges in establishing formal institutional processes for integrating modelling outputs into policy, as well as in strengthening national capacity for both conducting and interpreting IDM. To address these gaps, a collaborative decision-making framework that actively engages relevant stakeholders should be considered.²⁴ During COVID-19, this was trialled through a study identifying optimal vaccination strategies that combined mathematical modelling with economic evaluations, demonstrating the value of integrating disciplines. Furthermore, as shown in this report, IDM is employed across both health and non-health sectors, including the livestock industry. Developing a more coordinated system for the use of IDM evidence across sectors would help address existing fragmentation and siloed operations, ultimately supporting more coherent policy responses.^25,26

As modelling work gains traction, both in supporting health decision-making and in facilitating academic teaching and development,²⁷ creating infrastructure for modelling—particularly in ensuring timely and reliable data—will be important. The growing demand for more complex models highlights the need for continued advancements in computational power, integration of emerging technologies, and the availability of high-quality datasets.^27,28 Indeed, regardless of a model's complexity, the integration of information and data from diverse and heterogeneous sources remains a persistent challenge.²⁹ Therefore, ensuring the accessibility and reliability of data is paramount for improving model accuracy and properly accounting for uncertainty.

Capacity building through short-term (e.g., workshops), medium-term (e.g., secondments across institutions) and long-term (e.g., graduate programmes in country) in terms of technical aspects should also be established. This may also include strengthening communication skills of modellers and other relevant stakeholders.²² International research collaboration is particularly essential to elevate such capacities, as it can improve research quality, productivity, and academic impact.³⁰

Professionalisation of the field should also be considered to attract and retain talent in the sector, particularly given the diverse academic backgrounds of researchers in modelling, many of whom come from disciplines such as mathematics, computer science, and physics. Developing a strategic mechanism to recruit and integrate these talents is essential for the long-term sustainability and growth of the modelling ecosystem, with significant implications for human resource and capacity development.

Collaborations with partners, within and outside the country would also be beneficial to build local capacity. Regional initiatives such as the Modelling for Infectious Diseases in Southeast Asia (MIDSEA),³¹ Strengthening Preparedness in the Asia–Pacific Region through Knowledge (SPARK),³² Policy Relevant Infectious disease Simulation and Mathematical Modelling (PRISM),³³ and the International Consortium of Modellers (COMO)³⁴ could serve as networks linking modeller, providing platforms for collaboration and training in supporting policy decisions, and ultimately enabling a more coordinated approach to disease surveillance and control.³¹ In November 2024, an international conference on IDM was held for the first time, attracting modellers from across Asia as well as other regions, highlighting regional networks.³⁵ The Global South Leaders in Epidemic Analytics and Response Network (GS LEARN), supported by the Coalition for Epidemic Preparedness Innovations (CEPI), also seeks to build capacity in modelling and will be an important for regional engagement on IDM.³⁶ Notably, as the ASEAN Centre for Public Health Emergencies and Emerging Diseases (ACPHEED) is being established, with Indonesia, Thailand and Vietnam as co-hosts, there is potential to strengthen the capacity domestically and contribute to regional level efforts for building capacity for IDM and informing policy, linking the initiatives highlighted above.^37,38

Conclusions

This report identified the current roles and profiles of local stakeholders involved in the IDM, and there are several lessons learned in the context of Thailand. There is a need for greater investment in developing the IDM ecosystem locally that extends beyond the health sector to solidify the foundation for using evidence to inform the policy response during infectious disease outbreaks. The study offers an example for conducting such landscape analyses in countries to better understand the baseline capacity and actors involved in IDM. Equipped with a better understanding of the current status, countries can work together to articulare strategies to strengthen the use of IDM in preparing for future outbreaks.

Contributors

SD, MS, CR conceptualised and conceived the study. MS, CR, and PM were involved in the data acquisition, data analysis and all authors were involved in interpretation. MS and CR led the writing of the manuscript, and all authors provided critical inputs to the draft.

Data sharing statement

The authors confirm that the data supporting the finding of this report are available within its Supplementary materials. Additional raw data that support the findings of this report are available from the corresponding author, upon reasonable request.

Declaration of interests

Authors declare no conflict of interests.

The Health Intervention and Technology Assessment Program Foundation (HITAP) is a research unit in Thailand and supports evidence-informed priority-setting and decision-making for healthcare. HITAP is funded by national and international public funding agencies. HITAP is supported by the Health Systems Research Institute (HSRI), the Thai Health Promotion Foundation (ThaiHealth), the World Health Organisation (WHO), the Access and Delivery Partnership, which is hosted by the United Nations Development Programme and funded by the Government of Japan, the Rockefeller Foundation, the National Science, Research and Innovation Fund, Thailand Science Research and Innovation (TSRI), among others. The funders had no role in the study design, data collection, data analysis, data interpretation, or writing of the manuscript. The findings, interpretations and conclusions expressed in this article do not necessarily reflect the views of the funding agencies.