Thomas House 热爱科学 Profile picture
Professor of Mathematical Sciences. Views own & open for revision. Mainly COVID plus other modelling & data science in health.
Adam Kleczkowski Profile picture 1 subscribed
Jun 3, 2023 8 tweets 2 min read
This study has been getting some attention. I happen to work quite a bit on household transmission, so just a few thoughts.

jamanetwork.com/journals/jaman… 1. They don't test virologically, but rather look at fever - while hospitalisation with other viruses became low during COVID, we can't completely rule out other sources of fever.
Nov 6, 2022 6 tweets 2 min read
OK, so there is a current "controversy" (mainly restricted to here) about two possible explanations for the large RSV wave in the US:
(1) Delayed first infections due to immunity debt / gap;
(2) Damaged immune systems from COVID.
As you'd expect, discussion is heated! /1 But, the great thing about science is we can test things independently of rhetoric, so let's see what these explanations *predict*:
(1) Large out of season RSV waves happen after a year with low rates, then things largely return to normal independently of COVID prevalence; /2
Nov 2, 2022 6 tweets 1 min read
OK - so I see various people here who don't believe that things like RSV and flu will be worse after the pandemic due to immunity debt. What this skepticism usually stems from is the difference between *individual* and *population* level impacts. /1 From an individual point of view, once you're infected with something you have specific immunity to that pathogen, but you don't "need" to experience viral infection for your immune system to work. Generally as an individual you want to avoid infection with respiratory viruses./2
Oct 30, 2022 13 tweets 3 min read
OK - so this point but with more maths. It seems outrageous (as you can see from the responses to it) - we all know people who are more or less intelligent, and see that in scores on tests. The issue is that while intelligence is real, we can't measure it for a given human /1 with anything like the accuracy implied by a 3 significant figure "IQ" score. So sure - some people need support at work, school or home primarily because they have a learning disability and they would consistently score very low marks on "IQ tests". Also, we clearly /2
Aug 20, 2022 5 tweets 1 min read
R0 for COVID isn't 18, but it's very high - let's say it's 9 for the sake of argument. When people talk about continued masking etc., what value of R0 are they expecting to arise from that? Even incredibly optimistic numbers like 6 don't actually >

theguardian.com/commentisfree/… > look very different from R0 of 9 in terms of hospital demand. It's not helpful to describe a massive, incredibly serious problem of hospital overload in the face of a new demand, but conclude that community masking could play much of a role in reducing that. >
Apr 17, 2022 6 tweets 2 min read
So, here is how most scientists think about pandemics pre-vaccine, as well as how not to think about them. The x-axis is time, the y-axis is prevalence of infection, and LD stands for periods of "lockdown", NPIs or your preferred term for such measures. 1/6 Top left is a rough picture of what happened in most of the world - in this picture, we have medium-length LDs triggered by infections being at a level the politicians deem unacceptable, mitigating the worst of the pandemic. 2/6
Apr 16, 2022 5 tweets 1 min read
Calling out bad behaviour is not personal abuse. In fact, usually the more accurate the call out, the more indignant the response. In particular, explicit, deliberate use of the scientific peer-reviewed publishing process with the primary aim of advocacy for a particular political outcome is not appropriate - I really think this is a red line.
Dec 16, 2021 4 tweets 1 min read
I'm happy to report that after two years of pandemic science slog, my papers are mainly desk rejected, emergency funding is all ending soon, I've won no awards, can't tell people about most of the contributions I've made because confidentiality, get regular threats from > > both "sides" (iSAGE tend to try to get you sacked and Nuremberg 2 want to hang you), but you know, "Government scientist" so don't feel bad about any pile ons. As we try to get the postdocs facing redundancy soon and PhD students struggling with theses disrupted >