3 ms·
> More testing will prove this out by increasing the denominator by 10x. Source? You're ignoring the fact that the number of deaths lags behind the number of p
by panda-giddiness 7y ago
> More testing will prove this out by increasing the denominator by 10x.
Source? You're ignoring the fact that the number of deaths lags behind the number of people infected, which means the numerator is also being underestimated. Assume people either will/will not die three days from diagnosis. Then the naive case fatality rate is (for small t)
CFR_naive = [D exp(t - 3)] / [I exp(t)]
with D and I appropriate constants and t the number of days. The corrected CFR becomes
CFR = CFR_naive * exp(3)
That is, in this simple model, the true CFR is underestimated by approx 20.
- eanzenberg 7y agoThe numerator is unaffected by testing. When you die, you die, you can’t hide it and can’t hide the cause. If anything, testing may slightly decrease the numerator as at risk individuals may get treatment sooner.
- panda-giddiness 7y agoThe numerator is underestimated in a way separate from testing. People don't die immediately after they're diagnosed. They die some time after, during which the number of cases increases. Since the naive CFR is just the total number of deaths divided by the total number infected (assuming perfect knowledge of both -- the latter is also underestimated, for reasons you mentioned), this will always underestimate the true CFR until the spread reaches a steady state.