Why estimating population-based case fatality rates during epidemics may be misleading

Lucas Böttcher; Mingtao Xia; Tom Chou

where the death and recovery rates, µ(τ, t|τ 1 ) and c(τ, t|τ 1 ), depend explicitly on the duration of infection at time t and implicitly on patient health and age a [15]. They may also depend explicitly on time t to reflect changes in clinical policy or available health care. For example, enhanced medical care may decrease the death rate µ, giving the individual's intrinsic physiological processes a chance to cure the patient. These intrinsic individual-based death and recovery rates do not directly depend on population-level viral transmission. Equation (1), assuming an initial condition of one particular individual who has been infected for time τ 1 at the time of positive test, can be solved using the method of characteristics. From the solution P (τ, t|τ 1 ) one can derive the probabilities of death and recovery by time t as