An exact method for quantifying the reliability of end-of-epidemic declarations in real time

In this study the researchers develop and test a novel and exact data-driven method for optimising the timing of end-of-epidemic declarations. Their approach converts observations of infected cases up to any given time into a prediction of the likelihood that the epidemic is over at that time. Using this method, the authors quantify the reliability of end-of-epidemic declarations in real time, under ideal case surveillance, showing that it can depend strongly on past infection numbers. They then prove that failing to compensate for practical issues such as the time-varying under-reporting and importing of cases necessarily results in premature and delayed declarations, respectively.