Could the death rate be calculated accurately before an epidemic is over? Naive calculations could mislead the public.
A novel conoronavirus, named 2019-nCoV1 is currently spreading around the world. It is closely related to the SARS virus, which spread in 2002-2003[Xu et al (2020)]. Therefore, people like to contrast 2019-nCoV with SARS, in terms of the way that it spread. SARS killed roughly 10% of those infected.
According to the Chinese Center for Disease Control and Prevention, as of 28th January 2020, 5993 infection cases have been confirmed, including 132 deaths [Wu et al (2020)]. It is tempting to estimate the death rate to be 2.2% (132 divided by 5993). That could be misleading.
Those infected by the virus do not die immediately.2 For example, some may die after 14 days and some after 20 days. Before everyone who eventually die with the infected disease has died, the above calculation of death rate cannot be inaccurate.
To illustrate, suppose everyone infected eventually die of the virus. Before the epidemic is over, new cases will be confirmed. Some patients will not be dead yet. So the death rate will never be estimated at 100%.
In fact, when the number of new infection cases is on the rise, the calculated death rate will be an underestimation. The underestimation will get worse and worse before the peak of infection. The 2.2% calculated above could seriously underestimate the death rate of 2019-nCoV.
When the number of new infection cases starts to fall, the calculated death rate will increase, moving towards the actual death rate. The increase of death rate calculated is merely a reflection of the way that we calculate it. But it could be (mis-)interpreted as: "the virus is getting more and more fierce".
It is worth pointing out the obvious: the death rate of a virus could rise dramatically under a massive outbreak. This is because whether a patient survives or not depends on how he/she is treated. When the number of patients exceeds the number of hospital beds, or when it exceeds the number of medical equipments available for supporting these patients, the chance of them dying could increase.
[End]
References:
