If you have any problems related to the accessibility of any content (or if you want to request that a specific publication be accessible), please contact us at firstname.lastname@example.org.
Approximating SIR-type structured-population models using data-driven homogeneous-population analogs
AdvisorHurtado, Paul J
AltmetricsView Usage Statistics
Infectious disease models are crucial tools for policymakers and public health experts when proposing mitigation strategies and allocating scarce healthcare system resources. One of the most common model frameworks used in these applications is the SIR-type family of state transition models. At their core, these models assume a homogeneous population of individuals who are partitioned into several states based on their interaction with the disease, e.g., susceptible, infectious, and recovered. However, this assumption of homogeneity in the population is not always justified. Taking the COVID-19 pandemic as an example, the rates of hospitalization and death can differ by multiple orders of magnitude across age groups. This heterogeneity indicates that partitioning the population into sub-populations may provide more accurate modeling. But this strategy comes with its own drawbacks of increased data needs, model overparameterization, and increased computational complexity. In this talk, I will introduce a modeling approach that strikes a practical balance between the simplistic homogeneous population models and their structured-population model counterparts. I will discuss their implementation, and illustrate how they can outperform homogeneous population models without all of the added computational and statistical costs incurred by including an explicit sub-population structure.