Thread by @michaelzlin

Just don't get it. Why is a 2-week lag between cases and hospitalizations so hard for some people (MDs no less) to understand? We've only been studying that every day for the last 15 months.

To be specific: Israel wave lagging UK wave by ~2 weeks. Too early for champagne.

The original image (chart labels were cropped away in the above screenshots). I mean I've got the bubbly chilling like everyone else, but I don't see anything to celebrate here.

Yeah, cases triple May 24 to June 14. Hospitalizations triple June 7 to June 28. "It's different this time."

Sure, the non-vaccinated may be younger and healthier, but there will be hospitalizations and not everyone is going to regain their former levels of function.

Israel reached a minimum in case #'s about 14d ago. We can expect hospitalizations to start going up but very very slowly because cases are just now reaching ~100s per day. Expect about 1% hospitalization rate (half young unvaxxed, half older vaxxed)

https://twitter.com/carlquintanilla/status/1409683741757280257

By contrast, UK was at a minimum in cases 6 weeks ago, and then saw tens of thousands of cases. So we've had 4 weeks to observe the resulting hundreds of new hospitalizations, which are clearly seen above. Perhaps lower hospitalization rate than before, but it's not nothing.

Yet today we get this article: "Delta is not driving a hospitalization surge in England." What did we just see above? The writer may not consider 100s of hospitalizations as a surge, but it's irresponsible and wrong to imply that Delta is not a threat
nytimes.com/2021/07/01/hea…

Actually if you look closely at the UK curves, an increase in daily cases by 5000 led to an increase in hospital admissions by 500. So that's a full 10% of cases, which is just as bad as ever. I wonder if there is severe undercounting of cases going on.

Correction about Israel: I'd expect 3% hospitalization rate (derived from 1% in the young unvaxxed and 5% in the old vaxxed, and assuming 50/50 split of cases between them, very roughly speaking). Should not try to post when sleepy. So expect just a few hospitalizations next week

Follow-up to this old thread: I analyzed the UK data as of 7/7 in a separate thread. Sure enough there are hospitalizations, at a lower per-case rate than before, but not that much lower (2.5–3.3% of cases vs 7% in January)

https://twitter.com/michaelzlin/status/1413044278096367618

And now the follow-up about Israel. Sadly, but not surprisingly, the Topol-Gandhi celebration (see thread start) was indeed premature.

Weekly hospital admissions have risen by to 67 since 6/20. Rise starts somewhere between 6/20 and 6/27

We see cases start rising on 6/15. That's entirely consistent with hospitalizations rising between 6/20 and 6/27 (lag depends on if cases recorded as infection date or diagnosed date).

Actually if you look at hospitalization trends, Israel was on its way to 0 admissions this week until Delta ruined the party. That means this week's 67 new admissions are basically all from the recent Delta wave (there are ~0 patients being admitted with pre-Delta infections)

So then we can see how many cases gave rise to these 67 admissions. Because the dips lag by 5 days, we'll sum up cases from the week ending 5 days before the last weekly admission number (7/4), which is 6/24 to 6/30. That number is 1552. So hospitalization rate is 67/1552 = 4.3%

I had predicted 3% which was just a wild guess. In the Jan 2021 peak, there were 2000 hospital admissions in 1 week vs 58000 cases, or 3.4%. So the recent 4.3% rate is not lower than pre-Delta.

The 4.3% rate is not different enough from 3.4% to really analyze, but if you wanted to, you could wonder if it's because Delta is more severe (Scotland said you're 2x more likely to land in the hospital) or if people more likely to have severe disease are being diagnosed better.