A novel cohort analysis approach to determining the case fatality rate of COVID-19 and other infectious diseases

Charit S Narayanan

Best fit 114 We ran 125,000 simulations, modifying the onset-to-death interval, the CFR, and the 115 CFR growth rate. The CFR was kept in the range of 0.5% to 20.0%; the slope was held 116 between 0·005 and 0·20; and the onset-to-death interval was simulated between 0.1 and 117 20·0. We picked these ranges after some modeling exploration that convinced us that 118 the solution would certainly be within these parameters. We calculated the coefficient 119 of determination (R 2 ) for all of these simulations and identified the value for which the 120 fit was the best. We show results from two cases: one where we are relatively early in 121 the disease phase and another that is later. We studied these two cases to understand 122 how the progression of the outbreak affected the results. The difference in decrease in case fatality rates could be due to a number of factors 135 including the tactics employed by the Chinese government. These two cases 136 demonstrate the strong predictive power of the cohort-based methodology. Thus, using 137 a combination of known cohort sizes and the number of deaths on each day, we are able 138 to predict mortality very well.