Wednesday, 1 April 2020

On 8.2%

An update on my initial virus post, "Notes on the virus" (6th March 2020, updated 10th March). I had concluded that post by estimating a 3 to 5% chance of a 'severe' reaction for a healthy 50-something male allotment holder, on catching the virus. But now there is new data on this age group.

The seemingly reliable new study is from Imperial College, London and is based on data rather than computer models. The press release states a rate of hospitalisation ['severe'] of 8.2% in 50-59 year-olds known to have the virus. That's a one-in-twelve chance of needing to go to hospital.

However, this is across both non-healthy / healthy people in that age-group. Thus it's not to be imagined that all 50-59 year-olds have a uniform 8.2% chance of being hospitalised, if they catch the virus. They don't. It's far more likely that it's hospitalising mostly those in the 50-59 year-old group who already had underlying health condition (of a type that puts them in the danger category).

Further, on looking at the published paper from Imperial College one finds that their age-group data is actually from "mainland China", rather than from the UK. Thus my initial caveats about heavy drinking and smog-damaged lungs in urban China still apply. Such things would surely increase the risk. There's also the Chinese habit of conjuring with and censoring their statistics to consider. And that's presumably why the Imperial College paper can't state what the % risk is to otherwise healthy 50-59 year-olds. They presumably can't determine that because the Chinese didn't release that bit of the data, in the file on the 3,665 cases being tabulated by the report. Otherwise Imperial College would surely have reported the breakdown within the age group.

Thus "8.2%" sounds like a big juicy bit of new data, re: the risk to 50-somethings of hospitalisation. It will no doubt make for screaming newspaper headlines tomorrow, and scare half of Stoke witless. But it turns out it isn't very useful. It could actually be positive news for the healthy 50-something (i.e. there could be a negligible possibility of being hospitalised if a healthy 50-59 year-old), or it could be gloomy news (perhaps a 3-4% possibility of being hospitalised if a healthy 50-59 year-old). The latter is about the 3-5% chance I'd initially suggested in my initial post.

But, amazingly, it seems we still don't know what the actual % risks are for this age group, when properly split between the healthy/non-healthy, fat/normal, male/female and smokers/non-smokers. It should be relatively easy for such distinctions to be made, so why are we only hearing misleadingly broad statistics on this? I mean, if smokers are 80% more likely to die then we should know that ASAP and not at the end of pandemic.

Some extra background on this, from other sources on China. STAT reported that, across all age groups, having a pre-existing risk condition "increased risk of dying by 2.5 times". But that the "fatality rate in patients who reported no other health conditions was 0.9%".

No comments:

Post a Comment

Thanks! Comments get held for approval, but I hope to post yours soon!