The number of covid cases in the US is spiking but the death count isn't. This mystery has spawned many explanations, most of them optimistic ("it's young people who are recovering," "our therapies are better," etc). But the real explanation is simpler, and it's sad.
1/
1/
As epidemiologist @EpiEllie explains, it's almost certainly just "lead time bias."
That's when you test more people, including presymptomatic people, and thus discover the disease earlier than before.
2/
That's when you test more people, including presymptomatic people, and thus discover the disease earlier than before.
2/
That means we learn people are sick earlier, which means that the time between detection and death gets longer - not because people are surviving longer from the onset of symptoms, but because we're detecting sick people before they exhibit symptoms.
3/
3/
Lead-time bias emerges whenever we ramp up testing: routine mammograms and colonoscopies appeared to change the course of related cancers, but what was really going on was earlier, presymptomatic identification of cancers.
4/
4/
And while it's true that we measured earlier cases from the first symptoms, we didn't know what some symptoms were (loss of smell, for example) and we relied on self-reporting by gravely ill people, which isn't as good as actual tests.
5/
5/
That's called "recall bias." Sometimes we'd ask family members, who only knew about the symptoms that were severe enough to warrant mentioning ("proxy respondent bias").
6/
6/
Bottom line: "When you start identifying people at earlier stages of a disease, it looks like they survive longer (or have the disease longer) compared to when you identify based on severe symptoms." -Ellie Murray
eof/
eof/
