Lies, Gosh Darn Lies, and Not Enough Good Statistics: Why Epidemic Model Parameter Estimation Fails

Daniel E Platt; Laxmi E Parida; Pierre Zalloua

More so, the parameters that govern these epidemic models tend to reflect physiological rates of how the disease expresses itself in individuals, as well as effects that are moderated by demic characteristics. Examples are age structure in the population, which impacts both asymptomatic cases 11, 26 and severity of disease 8 . Identification of asymptomatic cases has been problematic since testing protocols tended to require symptoms, or contacts with known infected people. One case in California went untested for 10 days because she had no known contacts. Cases that advance to severe or critical depend on other factors, such as treatment modalities prior to development of advanced symptoms. The rate of transmission depends on physiological parameters as well as normal social distance and social distancing response to an epidemic, how public institutions such as schools are run, how grocery shopping interactions are handled, whether known infections are isolated and other factors specific to each community. Given how widely these parameters may vary from population to population, how they vary: how they depend on the geographically specific dominating SARS-COV-2 lineages dominant within a given geography 5, 6 , and how they depend on behavioral, social, age structure, and other factors of a population, it is worth seeking whether and how these factors relate to the expressed epidemic model rate parameters as phenotypes.