Local modeling

Learning from Covid-19 requires a modeling renaissance

“Flatten the curve” was a rallying cry at the start of 2020 as Covid-19 began to sweep the world. Despite limited understanding of the virus and how it is transmitted, public health officials made one point clear: Reducing transmission was the surest way to deprive Covid-19 of the oxygen it needed to thrive. to keep.

Top disease experts were quickly able to reasonably model and predict the early behavior of Covid-19. In just two months after the first recorded infection in the United States, public health officials had effectively offered Epidemiology 101 to a class of more than 325 million people. These models were powerful teaching tools during a time of intense uncertainty, offering insight into how quickly the virus is spreading, the likelihood of a deadly infection, and what a surge of cases might look like.

Almost two years later, however, modeling no longer seems so prominent in the toolkit used to determine how we discuss or anticipate Covid learning in the future.


This change reflects how experts’ understanding of the virus has evolved alongside the development of new knowledge about how best to create effective models. Models are limited and bound by the information and variables on which they are based. The more complete and accurate the data, the more detailed and accurate a model becomes – in other words, the better the data, the better the model.

With limited information at the start of Covid-19, experts had relatively few variables to help predict how the virus would spread and what could be done to contain it. Despite everything, modeling was an effective solution for disseminating the most relevant information to as many people as possible. But as scientists have come to understand the genetic makeup of SARS-CoV-2, the virus that causes Covid-19, and predictions have been weighed against the actual effectiveness of prevention methods such as the role of wearing masks, modeling has become more applicable. in highly specialized segments of the population.


The results of this approach – that without a vaccine, the virus would spread out of control – clashed with the political and logistical realities of containment policies across the United States, which tested the feasibility of prolonged lockdowns and mandatory social distancing. . It became increasingly clear that there was no magic bullet to stop Covid-19, so the focus on national modeling was replaced by decentralized and highly localized level-based policies. of infection in a given community.

With the emergence of the Omicron variant driving new infections around the world, there is even more urgency to revise traditional epidemiological thinking through the lens of the rise and fall of modeling. How can this trend improve our response to Covid and its variants, and better prepare public health officials for the rapid spread of unknown viruses in the future?

Revisiting the value of modeling

The datasets and specialized variables that have limited the effectiveness of modeling at the national scale show the value it can bring to setting a new operational standard for communities, businesses and governments. Introducing information such as vaccination rates or established prevention policies can help local constituencies predict behaviors and even reap economic benefits.

The key, however, is to improve how and where specialized data is deployed and studied in safe environments. Rather than relying on traditional tools, such as randomized controlled trials, cohort studies or case-control studies – which may each be subject to their own set of regulatory conditions, particularly when used to test pharmaceuticals and vaccine response – solutions driven by artificial intelligence can produce highly personalized models at an unprecedented rate using the same initial data. These virtual labs can simulate any number of variables, improving the scale at which modeling can quickly provide reliable information. By experimenting digitally, public health officials can bypass the slow process of in-person studies and create models that can inform real-world actions, thereby saving lives.

For example, a virtual lab environment can safely measure the impact of different preventative policies, such as social distancing, confinement, or notification of vaccination rates, on the spread of a specific virus strain. By cross-referencing known impact variables (policies) with a newly introduced variant, scientists can determine the optimal path for disease containment in near real-time.

The shift from modeling to highly digitized and AI-driven solutions will play an important role in mapping the future of Covid-19 behavior and better preparing experts for future pandemics. Given the very nature of AI – self-improvement through healthy data flows and analysis – virtual simulation mapping creates the foundation to apply any parameter imaginable. The more we understand the different contagion factors, the easier it becomes to predict how quickly a pathogen can spread or what measures will be most effective in preventing infection.

That’s not to say we should rely solely on AI: there’s always potential for bias when AI is involved, and it’s important that humans work side-by-side with the technology to avoid erroneous conclusions. . But as new threats emerge, AI’s use of historical data can help the scientific community anticipate potential infection rates and their overall impact.

In a virtual lab, for example, experts can compare individual activities such as grocery shopping or attending a sporting event in specific population groups. It can help answer questions like, “If a virus is airborne, how quickly could it spread among a group of adults in a movie theater?” Understanding how a virus behaves in specific contexts allows local communities to design policies tailored to their needs, particularly if they contain an above-average at-risk population, such as the elderly. This hyperspecific modeling technique can provide public health officials and government leaders with the flexibility they need to make the appropriate decisions to protect their communities.

The future of modeling will forever be shaped by Covid-19: redeploy national infrastructure to equip local communities with the best information possible to pursue the most effective action. Digital solutions, such as virtual labs and simulations, have the potential to take this work even further. Virtual modeling is the ideal domain for experimentation, so experts can ask the right questions, identify the right signals, predict the most likely outcomes, and plan accordingly.

Amir Mokhtari is Chief Scientist of Booz Allen Hamilton’s Strategic Innovation Group.