Re-upping a prior thread on joint work on heterogeneity and #Covid19 spread in light of today's @nytimes piece on heterogeneity and herd immunity thresholds; in which we found a way to *derive* why the homogeneous assumption in SIR models is extremal. https://twitter.com/joshuasweitz/status/1283838810728488960

The @apoorva_nyc piece investigates drivers of herd immunity thresholds and impacts of prior intensive spread for understanding near-/long-term risk: https://www.nytimes.com/2020/08/17/health/coronavirus-herd-immunity.html

Despite the controversy of the topic; it is critical to be able to do two things at once.

1. Talk frankly about heterogeneity and extreme assumptions inherent in SIR models w/out feedback or variability...

As shown in this figure, variability in susceptibility implies a bias to the process of infection that changes the force of infection in SIR models.

2. We should be able to do that and still discuss the limitation of these simple models with respect to policy (and in doing so, recognize that many simple models must be expanded into detailed feedback models in practice, where reality is inevitably more complicated).

In the end, adding heterogeneity asks more of the field; but that is a process that will benefit all.