COVID-19, artificial intelligence and the advantages of multi-method modeling


Dr Lauren Neal, AI health practice leader at research and consulting firm Booz Allen, advocates taking a multi-method approach to modeling the dynamics of COVID-19 disease in artificial intelligence.

She believes a multi-method approach helps to better understand COVID-19 and other infectious diseases – how they spread and impact communities, with the aim of being better prepared for future threats to public health.

She also believes that a “virtual laboratory” can be used to investigate a wide range of what-if scenarios and easily adapted to future high-impact public health threats.

IT health news sat down with Neal to talk about these approaches and how AI can help deal with the COVID-19 pandemic.

Q. Regarding artificial intelligence and COVID-19, how is a multi-method approach to modeling the dynamics of COVID-19 disease better than other approaches?

A. We have a long history of using simulation modeling to deepen our understanding of complex infectious diseases as well as their development, spread dynamics, and potential treatments. Examples include models for zoonotic diseases such as Zika, Ebola, West Nile virus, SARS, MERS, and the recent COVID-19.

Two modeling techniques, systems dynamics (SD) and agent-based modeling (ABM), have been used frequently in recent years to study the complex nature of infectious diseases despite their limitations. For example, SD works at a high level of abstraction by compartmentalizing the human population into different stages of disease such as susceptible (S), infected (I), and cured (R), among others, while assuming that everyone behaves in the same way in each compartment.

ABMs tend to overcome this limitation by tracking each individual member of the population and simulating granular profiles of individual interactions and movements within the population. However, with this high level of model fidelity comes a handful of tradeoffs, including high computational costs for large populations as well as increased model uncertainty due to a myriad of model assumptions.

We believe that choosing effectively between modeling methods is a matter of minimizing tradeoffs in the process of model creation, verification and validation. The idea of ​​multi-method modeling is to integrate different modeling methods to overcome the limitations of individual methods and get the most out of each.

Booz Allen’s multi-method model for COVID-19 combines the advantages of SD and ABM, allowing the simulation of spatially explicit scenarios representing future states of disease transmission within different local communities and testing policies risk management across a wide range of scenarios using “what-if” analysis.

Q. What is a virtual lab and how can it be used to investigate public health threats?

A. Historically, randomized controlled trials, cohort studies, and case-control studies were commonly used methods to study the epidemiology of public health threats as well as potential intervention options to mitigate the risks. However, performing large-scale trials and studies to achieve generalizability and sufficient statistical power is quite difficult, time-consuming and expensive.

Therefore, a comparable, reliable and easy-to-use planning tool is needed to assess interventions and their impacts. A virtual lab is a special type of simulation model that can be used to represent the dynamics of the spread of COVID-19 within a community and facilitate “what if” simulations that explicitly represent the uncertainty of data and assumptions. on the risk factors associated with the onset of the disease in the community.

A virtual laboratory is a risk-free environment, in which ideas about intervention strategies for a particular public health threat (e.g. social distancing, partial containment, and vaccination, among others) can be tested systematically without the need for it. time, costs and risks associated with experiments conducted in a real environment.

Virtual labs can have many uses and present many opportunities for innovation, but it is their ability to provide real-time information, enable forecasting, and provide decision support for live operations that is most immediately accessible. With these capabilities, community, state, and federal public health decision-makers can be more effective, improve efficiency, and save money while protecting lives.

Q. What is a Multi-Criteria Decision Analysis Framework (MCAD) and how is it used with artificial intelligence and COVID-19?

A. Decision-making regarding the implementation of public health interventions can sometimes be heuristic, and it can be argued that decisions based on a single criterion do not take into account important information about other relevant related outcomes. In the management of the COVID-19 pandemic, several compelling stories seem to have played an important role in the decision-making processes regarding the response and risk management measures to be implemented.

During the pandemic, public authorities had to make decisions based on uncertain quantitative evidence and expert scientific evidence (e.g. possible future scenarios), assessments of health system capacity (e.g. intensive care beds) and on the expected public adoption of more or less restrictive measures such as social distancing and lockdowns as well as the reopening of local communities and businesses.

When it draws on real-time data exploited using artificial intelligence and machine learning techniques, as well as disease predictive dynamics based on simulation modeling, multicriteria decision analysis (MCDA) can help decision makers make data-driven decisions based on multiple, sometimes conflicting criteria in a transparent and systematic manner.

For example, Booz Allen used an MCDA framework that took into account local decision criteria such as new daily infections, decrease in new daily deaths, new hospitalizations and use of intensive care beds to systematically analyze the simulated forecasts. obtained from our multi-method model and generate risk maps for each person. States.

These risk maps could potentially be used by public health decision makers to target available infection surveillance and control measures based on perceived levels of COVID-19 risk in local communities.

Q. How does all of this apply to the work of senior managers and caregivers in healthcare provider organizations on the front lines of the pandemic?

A. The COVID-19 pandemic has brought us unprecedented and ever-changing challenges since its onset. We’ve gone to great lengths to address these challenges using a suite of data-driven tools, including artificial intelligence and simulation modeling.

While early efforts have focused on epidemiological modeling of the spread of COVID-19 at global, national and state levels, the pandemic has raised many more localized challenges that our data-driven approaches can also address.

For example, the rapid onset of the COVID-19 crisis has shown increased demand and risk for healthcare provider organizations due to unpredictable and ever-changing circumstances. Simulation modeling and virtual labs can be used to proactively manage risks to health organizations during the current pandemic and future large-scale public health threats.

We can study a wide range of scenarios to improve our preparedness by optimizing hospital workflow structures, developing new processes, managing staff levels, purchasing equipment, managing beds and ensuring consistency. medical management of patients, among others.

In this way, a virtual lab can be used both as a learning tool (for example, to better understand how a hospital as well as frontline healthcare providers operate in the context of a COVID-19 outbreak. in the local community) and as an assessment tool (for example, testing complex scenarios such as optimal patient throughput for an emergency department).

Virtual labs can effectively support executive-level decisions made at the organizational level of healthcare providers to build capacity and manage scarce resources for effective care of critically ill patients, while testing scenarios to assess capacity the capacity of the health system to cope with expectations and unforeseen events. demands during the pandemic.

Twitter: @SiwickiHealthIT
Email the author: [email protected]
Healthcare IT News is a publication of HIMSS Media.


About Author

Comments are closed.