The COVID-19 pandemic has highlighted how misinformation, disinformation, information overload, and information voids can pose a challenge and threat to public health. The term infodemic has been used to describe the situation in which there is an overwhelming abundance of information, including false, misleading, or unnecessary information, in digital and physical environments during an epidemic.¹ Various factors drive the creation and spread of health misinformation. Issues can arise from predatory journals or preprint papers publishing inaccurate or unverified scientific information, medical professionals spreading disinformation for financial or political gain, or the media misreporting scientific research.^2-6 The algorithmic design of social media platforms could also contribute to the dissemination of problematic content.⁷ Additionally, the absence of credible and accessible health information, often referred to as information voids, can promote misinformation.⁸ Factors such as consumers’ political views, emotions, age, and ability to process abstract information can also affect their susceptibility to misinformation.⁹

However, infodemics are not only driven by misinformation and disinformation; rather, they can influence and be influenced by the broader information ecosystem, which refers to the dynamics of how people consume, produce, interact with, and behave around information.^8,10 There are two main dimensions to the information ecosystem: the supply side, which is shaped by the quantity and quality of media (including factors such as content and reach of media outlets, internet and mobile penetration and usage, presence of social media platforms, and the legal frameworks that govern the media sector), and the demand side, which encompasses people’s behaviours around media and non-media information sources (including people’s information needs, access to information, trust in sources, information-sharing behaviours, and information literacy).¹⁰ During emergencies such as a pandemic, both the supply and demand sides of the information ecosystem can undergo rapid changes. For example, new information sources could emerge and people’s demand for information could increase due to fear or uncertainty.⁸

Infodemics pre-date the COVID-19 pandemic, but the pandemic raised awareness of their potential to negatively affect public health outcomes. Infodemics are hypothesised to hinder individuals from making informed decisions about their health and consequently lead to morbidity or mortality.¹¹ Examples of harm that have been reported in connection with the COVID-19 infodemic include methanol poisoning, improper use of medications such as hydroxychloroquine and ivermectin for COVID-19 treatment, negative attitudes toward masking, reduced vaccination intent or uptake, and stigma or discrimination against specific population groups at high risk of COVID-19.^8,12-17 Moreover, physicians have reported that medical misinformation has made it more difficult for them to provide care for their patients with COVID-19·¹⁸ It has been suggested that infodemics could also have broader effects than negative outcomes for that specific health epidemic, such as reduction of public trust in health institutions and economic burden due to increased morbidity and mortality,^11,19 including costs that take away resources from other public health activities.

In response, public health authorities, civil society organisations, and other actors have implemented various measures, ranging from fact-checking on social media to training health-care workers in interpersonal communication for addressing patient questions.²⁰ Such efforts have been collectively referred to as infodemic management, defined by WHO as the systematic use of evidence-based approaches to manage infodemics and reduce their effect on health behaviours during health emergencies.²¹ Infodemic management is a practice that is informed by diverse disciplines, including (but not limited to) epidemiology, risk communication, data science, social and behavioural sciences, media studies, and marketing.^22,23

Public health practitioners and academics have proposed various frameworks to describe the crucial components of addressing infodemics. For example, Eysenbach suggested a framework that promotes high-level strategies, such as increasing eHealth literacy and information monitoring.²⁴ Kirk Sell and colleagues similarly presented priorities for a national strategy to address health misinformation in the USA, which includes detailed descriptions of potential interventions (eg, establishing guidelines for reporting on misinformation).²⁵ WHO has proposed a comprehensive set of actions with policy-level recommendations, such as forming strategic partnerships across sectors.¹ Scales and colleagues applied epidemiological models to outline a process for managing the COVID-19 infodemic through real-time surveillance and rapid response to misinformation.²⁶

Although these frameworks are useful for public health practitioners, they collectively leave gaps that limit their applicability to the complex and evolving challenges posed by infodemics. Several frameworks provide a wide range of potential interventions but do not give clear guidance on the process for deploying different types of interventions. Other frameworks focus on emergency response and consequently do not account for upstream strategies that address underlying determinants of infodemics. Furthermore, a framework that is based on a single discipline, such as epidemiology, can be too narrow to fully capture the breadth of approaches needed.²⁷

The overall aim of this Health Policy is to propose a unifying, expansive framework that addresses these gaps in existing frameworks—namely their little focus on upstream strategies beyond emergency response, insufficient guidance on identifying and categorising different types of interventions, or narrow scope due to framing infodemic management within one discipline. We sought to present an overview of interventions that can aid public health practitioners and policy makers in conceptualising and strategising their responses to the ongoing and future infodemics, including broader applications to non-acute settings.

We chose the public health prevention model to inform the development of our framework because it provides an ideal theoretical basis for addressing the gaps in existing infodemic management frameworks. The public health prevention model is a conceptual framework that outlines distinct stages of prevention efforts aimed at addressing health issues and categorises health interventions into four levels of prevention.^28-30

These four levels of prevention are: primordial prevention, which seeks to prevent the emergence of risk factors by addressing environmental, policy, or social conditions;^28,31 primary prevention, which aims to prevent the onset of a specific condition in people who are already at risk;^28,32,33 secondary prevention, which aims to detect and treat a condition in its early or preclinical stage by implementing measures such as routine medical screening;^34,35 and tertiary prevention, which includes interventions that are designed to prevent progression of disease, minimise suffering, or alleviate symptoms to avoid subsequent disability or complications.^36,37

The public health prevention model has been used to inform a wide range of public health topics, such as cancer control, cardiovascular disease prevention, mental health promotion, and injury prevention.^38-41 Examples of the use of this model include a review of approaches to cancer prevention in low-income and middle-income settings³⁸ and a guide for nurse practitioners to plan a regimen for the prevention of cardiovascular disease in patients who are HIV positive.³⁹ In addition to the widespread use of this model in public health and preventive medicine, its different levels of prevention comprehensively capture both acute and upstream interventions. Furthermore, the model has distinct stages that can help practitioners identify appropriate interventions depending on the intended level of prevention.

For each level of prevention, we sought to apply the same underlying principles to the context of infodemics and provide specific examples of interventions that fall under each level (figure). We then present the implications of our framework, including recommendations for the public health community, and the limitations of our approach.

Primordial prevention is the initial stage of preventive measures, which aim to intervene before risk factors for disease emerge in societies or communities. This level involves interventions that address environmental, policy, or social determinants of health, such as designing built environments that promote physically active lifestyles. In the context of infodemics, primordial prevention can include interventions that help develop communities and health systems that are resilient to infodemics.

Primary prevention refers to strategies for preventing the onset of a specific condition among individuals at risk, such as vaccinating individuals to protect them from pathogens that they could be exposed to. Likewise, primary prevention in our proposed framework includes interventions that aim to prevent infodemics from occurring particularly in individuals vulnerable to misinformation or other negative outcomes of infodemics.

In public health and preventive medicine, secondary prevention involves measures such as routine medical screening, which enables early detection and treatment of a condition. Similarly, the goal of secondary prevention of infodemics is to proactively identify early warning signs of infodemics and to take swift action to contain them before they affect the wider population.

Tertiary prevention in public health and preventive medicine involves preventing disease progression or alleviating symptoms. Similarly, in our proposed framework, tertiary prevention aims to reduce the negative effect of infodemics on public health. This approach involves implementing reactive activities to slow the spread of an infodemic or mitigate the negative effect of an infodemic on people’s health or public health programmes after the infodemic has become established.

Addressing infodemics through the preventive lens of public health offers several advantages. This framework expands the scope of infodemic management beyond emergency response, while still recognising its importance, and emphasises the need to develop upstream interventions before public health emergencies occur. Furthermore, this breadth encourages public health professionals to consider developing interventions beyond only responding to misinformation, such as debunking or other communication interventions.

Taking a broad preventive approach involves the consideration of the information ecosystem. An information ecosystem approach ensures that public health interventions are effective by considering all levels of prevention, including routine activities that address infodemics. Focusing resources and interventions on only secondary and tertiary levels will probably leave gaps in addressing the full range of challenges that infodemics present and could result in inadequately addressing underlying social, cultural, and historical factors. Developing interventions to increase trust at the primordial and primary levels would help increase population resilience before public health emergencies occur and shows the importance of upstream strategies for addressing infodemics alongside acute responses. For example, higher levels of trust in health-care workers or governments appear to be associated with decreasing levels of belief in misinformation or conspiracies.^44-46

The information ecosystem approach also highlights that misinformation is just one aspect of a complex ecosystem and is not created or consumed in isolation. Factors from the demand side of an information ecosystem, such as people’s information needs,⁹⁶ literacy levels,⁹⁷ or trust in specific political leaders,⁹⁸ have been proposed to shape the emergence or circulation of misinformation. Additionally, factors from the supply side, such as the design of social media platforms^99-101 or the quality of information from public figures,¹⁰² have been suggested to play a role in influencing the spread of misinformation. Thus, although reactive interventions to counter misinformation (eg, debunking or mass communication) are crucial in some contexts, they should not be the sole focus.

Enacting this framework would require a paradigm shift within public health and would probably require policy-level or systems-level attention for allocating technical, financial, and human resources to the management of information ecosystems. Coordinated effort and collaboration between various actors would be necessary for interventions to be consistently and comprehensively designed and implemented, especially at the primordial and primary prevention levels. Civil society organisations, educational systems, media organisations, and other community-based organisations all play an important role in addressing underlying vulnerabilities of communities in the context of information ecosystems. Moreover, robust partner networks can be used across multiple levels of prevention; for example, a strong network of fact-checking organisations can be instrumental in developing trust, identifying information voids, or disseminating health information to diverse communities. Additionally, it is essential to identify, work directly with, and address the needs of community members who are particularly susceptible to infodemics, such as communities with language barriers or historically marginalised communities with low trust in health authorities. A practical approach for public health authorities and practitioners to help in achieving this is by prioritising community-centred interventions, exemplified by the Bridging Research, Accurate Information and Dialogue model⁵⁰ mentioned earlier.

It could also be necessary to conduct assessments within health systems to identify organisational barriers that impede the implementation of management strategies of information ecosystems. Health authorities should continue to review and evaluate their management strategies for the COVID-19 infodemic, including risk communication approaches. This evaluation could yield insights that improve the implementation of interventions to manage information ecosystems, especially in the context of rebuilding trust. Specifically, it is important to reflect on whether risk communication by health authorities could affect public confusion or misinterpretation during health emergencies. Examples of such challenges might include difficult-to-understand policy guidance documents, changing stances on virus transmission, and inconsistent advice regarding masking.^103-105

However, an information ecosystem approach could also result in ethical pitfalls. Within preventive public health, quaternary prevention has been defined as protecting individuals from medical interventions that are likely to cause more harm than good. Some publications have defined quaternary prevention even more broadly, extending beyond the prevention of overdiagnosis or overtreatment, and positioning it as an ethical approach that applies to other levels of disease prevention.^106,107 This approach is especially crucial when considering the potentially unethical or harmful consequences of infodemic management interventions,¹⁰⁸ such as concerns about privacy, the effect of content moderation on free speech, and the monitoring of closed social media. The concept of quaternary prevention appears to be new and less established in the context of infodemic management interventions. Therefore, we did not include it in the proposed framework, but examples such as ongoing efforts to create ethical guidelines for social listening¹⁰⁹ can be classified as quaternary prevention and could be added as another layer.

Our framework can also be potentially expanded by considering regulatory and legal interventions specifically tailored to address disinformation. For example, legislators, medical boards, or professional associations could implement disciplinary measures for health professionals who spread disinformation, such as revoking their certification or licenses.^110,111 In Europe, the law enforcement agency Europol has removed online links to advertisements or marketplaces promoting fake COVID-19 pharmaceutical products.¹¹² These strategies are typically not within the purview of public health and were therefore not included in our framework. However, public health professionals could explore areas such as counter-disinformation governance and cyber security, while also considering the ethical implications of applying these interventions to public health.¹¹³

We acknowledge several limitations to our approach. First, there are intrinsic limitations of a narrative review, such as selection bias. For example, our review of interventions was limited to literature published in English, which could have resulted in the omission of relevant interventions and studies discussed in other languages. Second, although we have provided examples of interventions within each level of prevention, we were not able to comment on the effectiveness of the interventions in a systematic way. Further research is needed to evaluate the effectiveness of the interventions and determine which ones are most effective in different contexts.^11,23 Third, our proposed framework does not outline specific timings for starting various interventions. We acknowledge the importance of considering timing when preventing and responding to infodemics in practice, especially in settings where it is not feasible to continuously do all four levels of prevention. Decisions about when to start interventions are contextual and require consideration of factors such as the available resources and the nature of the current infodemic-related issues. Fourth, some interventions have elements that can encompass multiple levels of prevention, making it less clear-cut to categorise them into a single level. For instance, community engagement can be applied to activities to develop trust or public communication. Therefore, the prevention framework should serve as a guide rather than a rigid categorisation system.

Ultimately, the framework we propose highlights a gap in the existing literature on infodemiology and a need for research to better establish the chains of causality between the levels of prevention. Associations between risk factors, information exposure, health behaviours, and health outcomes must be better understood. Researchers must investigate the relative effects of different components of the information ecosystem, including not only misinformation and disinformation but also information overloads and information voids. Developing indicators for measuring the effect of exposure to misinformation, disinformation, information voids, and information overloads is also crucial, particularly regarding their effects on the intent and uptake of health behaviours. Such indicators can also promote formative research to identify communities or individuals susceptible to infodemics, which is particularly important for effective development of interventions at the primary prevention level. In the same way we approach any challenge in public health, combining a preventive and holistic approach requires a better understanding of the overall ecosystem, and only then can we be prepared for future infodemics.