TennTradition
Defended.
- Joined
- Aug 14, 2006
- Messages
- 16,919
- Likes
- 822
Estimated Influenza Illnesses, Medical visits, Hospitalizations, and Deaths in the United States — 2018–2019 influenza season | CDC
CDC estimates that influenza was associated with more than 35.5 million illnesses, more than 16.5 million medical visits, 490,600 hospitalizations, and 34,200 deaths during the 2018–2019 influenza season. This burden was similar to estimated burden during the 2012–2013 influenza season
IHME | COVID-19 Projections
All beds needed
56,831beds
WTF are we doing?
I believe that all beds needed stat is their projection of max beds needed on peak day. Is that right? If so, we need to do some math.
Flu hospitalizations are odd. If you take the CDC official cumulative hospitalization rate for the 18-19 season it ended the season at 63.6 / 100,000. With about 328MM people that gives you 209k deaths. But their median estimate was 491k as you posted above. That is 2.35x. This must have something to do with the catchment zones they use to track flu (TN is one of those) vs. how they scale that up to general population in their final modeling.
Peak week over week cumulative hospitalization rate changes in 18-19 occurred during weeks 8-11 of 2019 (that is weeks 21-24 of total flu season). Over that time the cumulative hospitalization rate changed from 33.5 / 100,000 to 46.8 / 100,000. That is a change of 13.3 over 3 weeks, or a total of 13.3 / 100,000 x 328 MM = 43.6k hospitalizations over this 3 week peak period. Given that it seems the hospitalizations tracked by the CDC ended up being about 2.35 times lower than their final estimates for hospitalizations, I'm going to scale this up: 43.6k x 2.35 = 102.5k hospitalizations over that 3 week period.
Doing more math:
103k hospitalizations over the 3 week peak
Looks like the average hospitalization time for influenza is 5.3 days per patient
So, that is 546k hospital bed-days during the 3-week peak
21 days over the 3-weak peak
That works out to be 26,000 beds needed during the peak
That is less than half the beds that were projected by IHME for the peak hospitalization rate for CV. And that is on top of the beds that were still occupied by flu patients. I think that IHME, even after their revisions, still over-predicted hospital bed need, though.
But I think this math makes sense of the numbers. We just finished week 6 of having CV deaths. 3 of those 6 weeks have been spent in lockdown. Yet we ran 2x past the number of hospital beds needed at peak flu season. I don't want to think of what those numbers would have looked like had we waited 3 more weeks before doing anything. We are already over 25k deaths. By week 6 of having flu deaths, we were only at 119 deaths. Now, the same scaling issue exists there. We only lab-confirmed 7,171 flu deaths in the 2018-2019 season. But we estimate there were 34,200 deaths. So that is 4.8x higher deaths estimated in the season than were lab-confirmed. If we apply that same scaling factor we get 571 deaths vs. over 25,000 (but these are lab-confirmed CV+ deaths - modeled deaths would almost certainly be higher). CV came on fast and hard once we started counting deaths. I think that answers the question of what we are doing.