Please leave any comments here.

Aim

This post provides an overview of emerging pandemic response strategies which aim to deliver the "right intervention at the right time, and to the right population", moving beyond blanket instruments such as nationwide lockdowns. It proposes for the EA biosecurity community to dedicate more resources towards research and advocacy for such interventions to be incorporated into national preparedness plans.

Background

Kevin Esvelt proposed a three-pronged approach towards safeguarding civilization from global catastrophic biological risks (GCBR): delaying the development and misuse of bioweapons, improving systems to detect emerging GCBR events, and strengthening our defences against infections by preventing GCBR events from escalating into actual global catastrophes.^[1] (Alternatively, I would have used the more conventional "Prevent, Detect, Response" framework if this post was meant for other forums).

Most of the interventions proposed by the effective altruism biosecurity community thus far focus on delaying/prevention (e.g. evals for bio-design tools, DNA synthesis screening) and detection (e.g. metagenomic surveillance). There has been some work on next-generation PPEs and far-UVC as technologies that can improve our ability to defend against infections. However, the history of smallpox eradication showed that besides these technological innovations, novel disease control strategies can also significantly improve our odds against pandemics.

(In addition, I also recall reading somewhere on the forum that someone argued that it is insufficient to focus on surveillance, but we should also look at how to improve the translation of surveillance insights into disease responses)

I will argue in the rest of the post that,

• Recent advances in disease surveillance have enabled a new generation of precision pandemic response strategies that promise to deliver the "right intervention at the right time, to the right population".
• The world would be significantly better equipped to handle GCBRs if public health agencies adopt precision pandemic response strategies as part of their disease control toolkit.
• The EA community can help to advocate for precision pandemic response strategies to be adopted into preparedness plans, provide resources for capacity building, fund implementation research, and build networks to share expertise.  

I want to be clear that I am not disagreeing that the "delay" and "detect" interventions that the EA biosecurity community are currently pursuing are valuable. I am only making the (one-sided) case that precision pandemic responses can be one of the most promising interventions. Nevertheless, comparing their value to investments into precision pandemic response falls beyond the scope of this post.

Advancements in Public Health Surveillance

In recent years, there have been significant advancements in public health surveillance which have made new approaches to disease control possible.

Predictive analytics

Traditional reactive responses typically involve mobilizing resources and implementing interventions after an outbreak has been confirmed and cases are rising. This often means playing catch-up, with public health actions initiated in response to observed morbidity and mortality.^[3]

On the other hand, predictive analytics helps public health officials stay one step ahead of potential outbreaks or surges, giving them valuable time to take pre-emptive action. Thanks to significant advances in machine learning, we now have sophisticated analytical tools that can make accurate forecasts anywhere from a few days to several weeks' notice. For instance, when retroactively applied to the COVID-19 pandemic, PandemicLLM accurately predicted disease patterns and hospitalization trends one to three weeks out, particularly excelling when the outbreak was in flux.^[4] 

Spatial Analytics

With spatial analytics, it is now possible to precisely identify geographic areas where disease events cluster significantly such that they exceed the expected rates (i.e. “hotspots”). Such analytics have been enabled by the availability of geographically referenced data such as wastewater surveillance data and mobility data derived from anonymized cell phones and credit card transaction data, which are not traditionally used in epidemiology.^[5] Geographic Information Systems (GIS) allows epidemiologists to integrate this information and identify areas of higher risk by analyzing pathways of disease spread.^[6] Geographically referenced socioeconomic and environmental data (e.g. land use, precipitation) can also be used to map vulnerable communities.

Real-Time data

With real-time surveillance systems, it is now possible to detect emerging outbreaks and changes in disease transmission patterns as they happen, rather than with the delay typical of traditional public health reporting. These systems are powered by continuously updated data streams such as emergency department visits^[7], syndromic surveillance, wastewater viral load measurements, and digital signals like search engine queries and social media mentions. Advances in data infrastructure and machine learning allow public health authorities to process and interpret these dynamic inputs in near real-time, enabling earlier warnings and faster responses. Real-time dashboards, alert systems, and predictive models further enhance decision-making by translating raw data into actionable insights for intervention planning.

These advancements in public health surveillance have made it possible for new disease control strategies.

However, it is not sufficient to simply strengthen disease surveillance capabilities. It is equally – if not more important – for disease control agencies to incorporate these insights into precision pandemic response strategies.

Precision Pandemic Response Strategies

Precision pandemic response strategies refer to infectious disease control strategies that are pre-emptive, population-targeted, and/or adaptive. These include, but are not limited, to the following:

• Pro-active infrastructure development. Disease control officials can use forecasts to anticipate future pandemic response needs, in order to respond more effectively when the time comes. For instance in March 2020 (a few months into the pandemic), the American Hospital Association predicted that hospital bed capacity would be exceeded by 270 percent for inpatient care and by about 500 percent for intensive care.^[8] Given this information, the US Army Corps of Engineers (USACE) was able to set up 32 alternative care facilities and add 15,074 hospital beds by April 2020.

• Strategic deployment of testing sites and resources. Officials can also use these forecasts to strategically deploy resources (such as medical supplies, PPEs, staffing, temporary quarantine facilities, vaccines, or other countermeasures) to communities of greatest need (however defined), that have insufficient resources, or to optimize for other outcomes. For instance, researchers at South Carolina developed a GIS-based multicriteria decision model to guide the placement of COVID-19 testing sites, incorporating considerations like accessibility, vulnerable populations, and daily changing COVID-19 case fluctuations.^[9] Another example is “Get Us PPE” which used spatial analysis to optimize PPE distribution by minimizing transportation miles required while incorporating donor and recipient demands.^[10] 

• Nuanced social distancing measures. Nuanced social distancing measures – such as limits on gatherings, promoting telework, or business closures – can be applied to target communities which are likely to have a higher risk of disease transmission or towards higher risk activities. For example, in Portland, Oregon, GIS was used to identify areas where vulnerable populations, such as people experiencing homelessness, congregate.^[11] This allowed authorities to target the placement of sanitation stations and push out targeted communications about social distancing and hygiene measures. Such nuanced interventions are a much less costly alternative to blanket interventions like national lockdowns.

• Adaptive intervention protocols. Adaptive protocols represent a flexible approach to disease control, characterized by a real-time feedback loop that allows health authorities to adjust measures in response to evolving epidemiological data. This adaptive approach contrasts with static, long-term interventions by enabling policies to be tightened or loosened based on predefined triggers and continuous monitoring. Modelling studies based on data collected from Minnesota state explored intermittent "light switch" approaches, where social distancing measures were turned "on" and "off" based on threshold triggers, such as weekly hospitalization rates. Adaptive response scenarios were shown to substantially reduce infection circulation and prevent health care capacity from being exceeded.^[12]

What can EA do to help?

With recent cuts in budgets to USAID, international NGOs, and the gradual fading of COVID-19 in the global public’s memory, there is not enough systematic effort to incorporate precision pandemic response strategies into existing preparedness plans. There is also limited research to develop information systems to effectively translate data from surveillance into actionable insights to guide precision pandemic responses.

The EA community can help to:

1. Advocate for precision pandemic strategies to be adopted as part of preparedness plans
2. Fund implementation research, and
3. Grow networks and communities of practice to share expertise.  

My preliminary assessment is that such interventions can be highly impactful, and while they may not be highly neglected, they are quite tractable. It may be worth further investigation to quantitatively estimate its potential impact and compare it against other biosecurity interventions.

References

1. ^{^}

    Esvelt, K.. (2022). Delay, Detect, Defend: Preparing for a Future in which Thousands Can Release New Pandemics. Geneva Paper 29/22.

2. ^{^}

    Sharma, V., Sharma, R., & Singh, B. B. (2024). Etymologia: Ring Vaccination. Emerging Infectious Diseases, 30(2), 279. https://doi.org/10.3201/eid3002.221909.

3. ^{^}

    Hosen, Nadirsyah, and Nurussyariah Hammado, 'Indonesia’s Response to the Pandemic: Too Little, Too Late?', in Victor V. Ramraj (ed.), Covid-19 in Asia: Law and Policy Contexts (New York, 2021; online edn, Oxford Academic, 21 Jan. 2021), https://doi.org/10.1093/oso/9780197553831.003.0021, accessed 28 June 2025.

^{^}

 Du, H., Zhao, Y., Zhao, J. et al. (2025). Advancing real-time infectious disease forecasting using large language models. Nat Comput Sci 5, 467–480. https://doi.org/10.1038/s43588-025-00798-6

^{^}

 A Five-Year Retrospective: GIS in the Fight Against COVID-19, accessed June 27, 2025, https://www.esri.com/about/newsroom/blog/covid-gis-five-year-look-back

^{^}

 Smith, C. M., Le Comber, S. C., Fry, H., Bull, M., Leach, S., & Hayward, A. C. (2015). Spatial methods for infectious disease outbreak investigations: systematic literature review. Euro surveillance : bulletin Europeen sur les maladies transmissibles = European communicable disease bulletin, 20(39), 10.2807/1560-7917.ES.2015.20.39.30026. https://doi.org/10.2807/1560-7917.ES.2015.20.39.30026

^{^}

 Hughes, H.E., Edeghere, O., O’Brien, S.J. et al. Emergency department syndromic surveillance systems: a systematic review. BMC Public Health 20, 1891 (2020). https://doi.org/10.1186/s12889-020-09949-y

^{^}

  A Five-Year Retrospective: GIS in the Fight Against COVID-19, accessed June 27, 2025, https://www.esri.com/about/newsroom/blog/covid-gis-five-year-look-back

^{^}

 Carbajales‐Dale, P., et al. (2023). Using GIS to improve public health emergency response in rural areas during the COVID‐19 crisis: A case study of South Carolina, US. Transactions in GIS, 27(4), 975-995.

^{^}

 Bala, R., Sarangee, K. R., He, S., & Jin, G. (2022). Get us PPE: a self-organizing platform ecosystem for supply chain optimization during covid-19. Sustainability, 14(6), 3175.

^{^}

Location intelligence enhances COVID-19 Collaboration, accessed June 27, 2025, https://www.esri.com/arcgis-blog/products/arcgis-pro/analytics/location-intelligence-enhances-covid-19-collaboration

^{^}

 Sanstead, Erinn C et al. “Adaptive COVID-19 Mitigation Strategies: Tradeoffs between Trigger Thresholds, Response Timing, and Effectiveness.” MDM policy & practice vol. 8,2 23814683231202716. 11 Oct. 2023, doi:10.1177/23814683231202716

Show all footnotes