[dragontamer]

The pipeline is cases -> hospitalizations -> deaths, with a week or 2-weeks between each stage.

Only 20% of cases turn into hospitalizations, and only a small percentage of hospitalizations turn into deaths... but that's the general pipeline. The exact % changes depending on population dynamics (who gets sick, and where), but the biggest contributing factor is the number of people in the previous stage of the pipeline. (If you double cases, then a few weeks later, hospitalizations will almost certainly double).

If a case-spike happens, the idea is to make a decision BEFORE hospitalization spikes up a few weeks later. We want to be ahead of the wave, not behind it.

[slimed]

!remindme 2 weeks

[christkv]

Well I'm saying the Madrid numbers are not bearing out the prognosis you are outlining. It could be younger people are getting ill and that elderly people are better isolated but it's hard to know because they don't provide deaths by age groups.

[dragontamer]

My spanish isn't the best, but with assistance from Google Translate, I'm able to line up the death-charts with the cases charts.

https://imgur.com/cJw6dTk.png

If I'm reading that spanish correctly: its "Number of daily cases" (casos diarios) on the top, with "number of deaths" (Muertos) on the bottom, both in Madrid alone.

----------

The first wave is clearly wreaked by undertesting the population: Spain probably just didn't have enough tests to really count the cases back then. But for the 2nd wave (going until today), the death numbers clearly went up a few weeks after the case-count went up.

The case-count has begun to decline in Madrid, I think they're "over the hump" (knock on wood...). So looking forward, they're in a good spot.

But you can clearly see: when cases go up, deaths go up soon afterwards. When cases go down, the deaths go down soon afterwards.