iiinova — where your digital self keeps evolving

Jean Fisch @jeanfisch.bsky.social · 07/10/2026

ADDENDUM: I forgot to add something: the age pattern of heat deaths While clearly, the bulk is among older people, it is interesting to see that the May and June-July heat events saw some spike of mortality among the <65 as well

010

Jean Fisch @jeanfisch.bsky.social · 07/10/2026

There are signs of this also in 2026 in England and Wales in the May wave but very little in the summer waves Until an analysis is done of the profile of the people who died during the heat wave vs before, the life expectancy of heat deaths it remains everybody's guess END

100

Jean Fisch @jeanfisch.bsky.social · 07/10/2026

Clearly, there will be a harvesting effect, ie the heat kills very frail people about to die anyway In fact, more often than not, a spike of deaths in a heat wave is followed by a trough of deaths, a sign of harvesting effect 2/

100

Jean Fisch @jeanfisch.bsky.social · 07/10/2026

A quick comment on heat deaths - I purposefully used the word "heat triggered" because that is what it is: The person died THAT day because of high heat Without high heat, the person would have lived longer - How much longer is the big question 1/

100

Jean Fisch @jeanfisch.bsky.social · 07/10/2026

Now that the summer has come to an end, it's possible to have a view on the heat mortality in England and Wales in 2026 I come to 5,800 deaths triggered by heat, an amount which is far higher than say in 2020, 2021 or 2022 (which also saw significant heat-triggered deaths)

151

Jean Fisch @jeanfisch.bsky.social · 30/09/2026

... at the height of the pandemic, the tendency to suspect covid was higher than now so this creates a bias Positivity also has a bias as it only means something in relationship of a testing regime Have you been able to find testing levels somewhere? This could help Anyway: thanks for all this!

100

Jean Fisch @jeanfisch.bsky.social · 30/09/2026

Ah thanks, now I understand what you mean (and it also clarifies some of the points I was not sure about in the methodological paper!) Yes, syndromic has an arbitrary element about it but in the case of Spain, it's done via sampling which should dampen the issue but I see that indeed that ...

100

Jean Fisch @jeanfisch.bsky.social · 29/09/2026

yes but I don't understand the implication Could you help me by rephrasing / expanding your point? That would be great!

100

Jean Fisch @jeanfisch.bsky.social · 29/09/2026

Anyway, there is no good approach except for large scale random sampling as the UK did at the height of the pandemic but the "least bad" approximate is sentinel sampling as Spain does it At least so it seems but maybe I miss something! 5/5

210

Jean Fisch @jeanfisch.bsky.social · 29/09/2026

Recently, I found this very interesting substack piece from Biobot in the US which explains that pepper virus corrections are wildly off due to trends and this explains the issue that I found in the Scottish data 4/ biobotanalytics.substack.com/p/a-deep-div...

biobotanalytics.substack.com

A deep dive into Biobot’s approach to normalizing viral abundance data

By: Maxim Imakaev, PhD, and Josh Harrison, PhD

110

Jean Fisch @jeanfisch.bsky.social · 29/09/2026

Positivity only works for absolute comparisons if one has a similar size of samples every year which was not true between 2020-2022 and later I have never thought about mixing wastewaters and positivity but wastewaters also do not work for absolute comparisons, even if pepper virus normalized 2/

120

Jean Fisch @jeanfisch.bsky.social · 29/09/2026

Thanks for all this! I don't know about the second bit, but what you describe here is a standard sentinel protocol used everywhere designed to avoid the issue of bias from random hotspots (an issue if one has only a small number of sentinel places) 1/

110

Jean Fisch @jeanfisch.bsky.social · 28/09/2026

Yes, the covid incidence is also increasing fast However, side question: In absolute terms, I get a very different picture from your's yet I simply plotted the weekly sentinel SARI and ARI incidence for covid (straight out of the sentinel data source you also shared) What's the difference?

110

Jean Fisch @jeanfisch.bsky.social · 28/09/2026

The latest measures of test positivity rates for covid in the US all signal that a peak has been reached recently So the message in my original chart may well prove correct, but the supportive chart was not, hence me erasing the whole thing END

120

Jean Fisch @jeanfisch.bsky.social · 28/09/2026

Here the picture with wastewater measures up to 12/9 and severity indicators up to 19/9 Up to that point, there was still an increase / increase to be expected based on past registration lag of deaths and hospital data However!

120

Jean Fisch @jeanfisch.bsky.social · 28/09/2026

ERRATUM: I just realize that the hospital and death data on this chart was "raw" ie not corrected for late reporting The data corrected for late reporting still signals so far in fact a continued increase, not a peak So I killed the thread - Sorry for the confusion

131

Jean Fisch @jeanfisch.bsky.social · 28/09/2026

Yes, but I think the small size of the country (vs the US) flattens any signal and makes it more erratic To me the remarkable thing is the absence of wave in 2026 In fact, this absence has led New Zealand to drop covid from its SARI hospitalization reporting completely: It doesn't happen

110

Jean Fisch @jeanfisch.bsky.social · 28/09/2026

Look at the flu waves: Unlike what one sees in western Europe, flu is not an annual event in Taiwan, but can hit the country several times during the year and at non-seasonally set moments If anyone has an explanation or material explaining this, I am interested! 2/2

012

Jean Fisch @jeanfisch.bsky.social · 28/09/2026

Also in Taiwan, the covid wave looks comparatively smaller than last year and is now receeding Let me add an interesting point on flu below though 1/2

121

Jean Fisch @jeanfisch.bsky.social · 27/09/2026

I have this chart which relies now on ABS weekly deaths (lag of 4-6 months) and monthly surveillance from ACDC It is remarkable how flat covid has been this winter in Australia (confirmed by sentinel hospital data)

110

Jean Fisch @jeanfisch.bsky.social · 08/09/2026

Just for clarity: / Estimates are mine / based on official stats / using the UKHSA approach, ie look at short term excess in heat weeks vs. the level expected from death evolution in the surrounding weeks (cf. here the example of Switzerland)

040

Jean Fisch @jeanfisch.bsky.social · 08/09/2026

Switzerland continued to experience completely off the chart temperatures and this for weeks on end early August What is interesting, from a mortality analysis, is that, obviously, the June episode did not "deplete" a pool of frail people so heat is more than harvesting 2/2

140

Jean Fisch @jeanfisch.bsky.social · 08/09/2026

Interestingly, Switzerland got hit strongly in June, as much of western Europe north of the Alps / Pyrenees ... but then got hit equally strongly (and much more severely) as its neighbours in the 1st weeks of August The reason is simple 1/2

141

Jean Fisch @jeanfisch.bsky.social · 04/09/2026

Paris only ever came briefly out of tropical nights during the summer ... and surprise-surprise, it saw massive amounts of heat-triggered deaths The levels are almost hard to apprehend

040

Jean Fisch @jeanfisch.bsky.social · 04/09/2026

Ah, I just see @actuarybyday.bsky.social also pointed out that there was heat-triggered mortality in the week ending mid-August in England and Wales He adds more perspective over the season as well so worth reading! bsky.app/profile/actu...

020

Jean Fisch @jeanfisch.bsky.social · 04/09/2026

Other update on heat: England and Wales seem to have had another heat-triggered mortality episode during the week ending mid-August and this despite having already gone through two other such episodes this summer

130

Jean Fisch @jeanfisch.bsky.social · 04/09/2026

Note: Heat-triggered deaths here is any excess mortality vs surrounding weeks in weeks with high heat (UKHSA approach) This is different to @santepubliquefrance.fr which derives heat deaths vs a statistically expected level of deaths from previous years, not short-term around the heat weeks 3/3

020

Jean Fisch @jeanfisch.bsky.social · 04/09/2026

That there was abnormally high (heat-triggered) excess in France can be seen from the comparison of actuals vs. expected range from pre-pandemic trends This range includes a baseload of heat deaths and yet there was continuous excess "on top" in the summer 2026 2/

110

Jean Fisch @jeanfisch.bsky.social · 04/09/2026

Quick update on 2026 heat mortality in France as @insee.fr released deaths up to Jul There was more heat excess in the aftermaths of "end of June" as initially expected and there was continuous heat triggered deaths up to mid-August at least I now come to 11-12k in total 1/

150

Jean Fisch @jeanfisch.bsky.social · 02/09/2026

NB2: / My estimates cover mortality up to week ending 23/8 for the "big" countries, ie DE, FR, SP, PT and IT (partially reverse engineering from z-scores using past correction factors) / So the total mortality is already "quite complete", hence my guestimate of ~50k for the full-year

010

Jean Fisch @jeanfisch.bsky.social · 02/09/2026

NB: I estimate the deaths using the UKHSA approach, ie look at excess deaths in "heat weeks" vs the surrounding weeks ... and I corrected during pandemic years for covid (which messed up things in particular in 2022) So it's a "broadly correct" estimate not a detailed analysis

110

Jean Fisch @jeanfisch.bsky.social · 02/09/2026

Quick head up: 2026 will see extreme heat in western Europe and high mortality (probably ~50k) triggered by it, esp in FR, Benelux and GER However, it is unlikely that the mortality will exceed that of 2022, because 2022 was also extreme in Iberia and Italy (which 2026 was less)

130

Jean Fisch @jeanfisch.bsky.social · 02/09/2026

I was updating my temperature data for France which I model based on 7 cities including Marseille 2026 is really out of the norm there as well (despite the fact that it did not see the super-high heat that much of France saw at the end of June)

065

Jean Fisch @jeanfisch.bsky.social · 30/08/2026

Ah, I see what you meant now!

010

Jean Fisch @jeanfisch.bsky.social · 30/08/2026

Well, yes ... I do take that assumption ... You make me think: To me it's obvious ("better be rich and healthy than poor and ugly") so what do I miss?

110

Jean Fisch @jeanfisch.bsky.social · 30/08/2026

PS: Any alternative explanation for the abrupt decrease post 1964 in the US would need to explain why it only happened as of 1969 in the UK and why it was only gradual there (Hint: England introduced measles vax in 1968 and uptake did not get beyond 60% until MMR in 1988)

050

Jean Fisch @jeanfisch.bsky.social · 30/08/2026

Measles: a look at the rates since 1950 - Vaccine got mandatory in 1963 so unless there is an alternative explanation for the abrupt reduction in 1964-68, the public health assumption has to be that / a highly effective vaccine "killed" the disease / vax should be encouraged no?

2111

Jean Fisch @jeanfisch.bsky.social · 29/08/2026

As measles deaths are a thing in the US, here an annual view by age band (I also show deaths with measles as contributing cause: Per definition, a contributing cause is one where the medic believes that, without out, the person would not have died at that moment so it's a cause of death)

042

Jean Fisch @jeanfisch.bsky.social · 28/08/2026

So just for clarity: / These are deaths per death certificate as compiled in CDC Wonder / I extracted those with B05 (Measles) on the certificate as UCOD or MCOD Either UCOD or MCOD means that, had the person not had measles, it would not have died according to the medic

030

Jean Fisch @jeanfisch.bsky.social · 28/08/2026

One thing surprises me in the measles situation in the US / I read everywhere: "no measles deaths so far in 2026" (pre Pennsylvania) / Yet when I tickle CDC wonder, it says: "two deaths (1 in MI / 1 in OH) where measles contributed to the death in 2026" What do I miss?

120

Jean Fisch @jeanfisch.bsky.social · 27/08/2026

The remarkable thing is that it's not uniform: West-Flanders (the part along the sea) hardly saw any of the drama Why? Well / It was cooler than 100km inland and it is a richer (higher GDP) / but it has also the highest % of older people so there are factors going in both directions... 2/2

030

Jean Fisch @jeanfisch.bsky.social · 27/08/2026

Belgium released its mortality data up to end of July What remains intriguing, is that there is no "obvious" sign of mortality deficit after the heat episode among the <65 I would really like to understand what the profile is of these deaths ... Let's hope someone makes a study 1/2

130

Jean Fisch @jeanfisch.bsky.social · 19/08/2026

Germany seemed to have seen a second peak of mortality triggered by heat this summer but there is no sign in the French data despite continued high temperatures (NB: deaths past June are ballpark correct own estimates using the French national mortality surveillance)

042

Jean Fisch @jeanfisch.bsky.social · 18/08/2026

Interestingly, the heat wave of early August led to a second peak of mortality in parts of Germany, but one which is much smaller Is it due to - weakened people for whom the 2nd wave was too much? - situational deaths? Without more detailed info, we will never know

062

Jean Fisch @jeanfisch.bsky.social · 16/08/2026

Link to the retracted paper: bmjpublichealth.bmj.com/content/bmjp...

bmjpublichealth.bmj.com

030

Jean Fisch @jeanfisch.bsky.social · 16/08/2026

So, technically @bmj.com is more than correct: the argumentation is anything but balanced But BMJ is also fully responsible for the mess here: It should NEVER EVER have published this opinionated blabla underpinned by selected arguments and references as a research paper in the first place END

142

Jean Fisch @jeanfisch.bsky.social · 16/08/2026

... / Reports of persistence of mRNA for weeks after administration in the body / Batch heterogeneity of mRNA toxicity was found in Denmark (no mention that it was refuted) / Simultaneous onset of excess and vaccination in Germany (no mention that it is not the case elsewhere) 2/

120

Jean Fisch @jeanfisch.bsky.social · 16/08/2026

If you wonder about the paper on excess mortality that got retracted: Here its argumentation why vaccines could be a cause / adverse effects are under-reported / there is no consensus that vaccines do not cause more harm (citation - an opinion piece by a French "alt thinker") ... 1/

150