Search results

Its not just that they are false positives its also the lack of balance in the population causes massive numbers of false positives for even quite an accurate test. If a population of 100,000 people where 3% have the disease are tested with a test with a 5% false positive rate then that is 4,850...

This article is worrying on two counts. Firstly it seems to provide evidence that the UK government is following a herd immunity strategy although not admitting to it in public. This seems to fit with there in action of tracking. Secondly, the quotes from the the Home office's scientific adviser...

Or was it a bad test. The WHO is giving the isolate 14 days after the end of symptoms as they see a risk people are still infectious after recovery I'm not sure what data that was based on.

I was assuming that if they did a test, trace and quarantine system coupled with the lock down then that would be a relatively short time (6-8 weeks) till they could relax things but keeping the test, trace etc in place afterwards. But I'm fully expecting the lockdown to last a year or more as...

The Faroe Islands have apparently tested a higher percent of the population than anywhere else! (10% of the 60000 population) then they are doing track and quarantine with 184 people having tested positive...

I think someone should. There is another thread discussing some of this. I wonder if it is something CureMe could do maybe in collaboration with Kings (The Twins Study group who have a symptom tracking app and not the normal BPS groups). CureME have an understanding of the diagnostic...

I think we need to be concerned with the James Lind process but this means participating in it. It isn't coming from the BPS groups so it has a good chance of giving sensible results but it could be high-jacked if we aren't careful. It could be an important thing to influence the NIHR and MRC...

It could be I think it is more than a simple regression fit. They talk about the different controls in place and also the load on ICU so I assume there is a level of simulation within the model to follow forward but it could be based on current death, spread and ICU rates. Also the UK figures...

I'm wondering if the model or results are just wrong for the UK. But it isn't clear to me what the different assumptions they are making are.

Some model predictions for deaths are reported here http://www.healthdata.org/covid/updates Predictions for the UK seem very high (66,000) in relation to other countries (say 20,000 for Italy) but I'm not sure why.

The UK ONS (Office of national statistics) have been collecting stats at a slower rate than announcements but probably with a higher degree of accuracy. They are looking at death rates from death certificates rather than hospital reports and they are doing so in batches. It seems to show a...

This report is suggesting an ICU in london is running short on other things then ventilators including dialysis machines. Suggesting other organs are being damaged as well as respiratory issues https://www.ft.com/content/e26524a5-c868-451c-a7d7-a91627a1722c

Some have expressed concern that the death figures aren't very timely in that they are saying it can take a few days for them to be reported from the hospital. The recovered figure seems to have been constant for a while so my guess would be its not being collected/reported. It does make me...

Issues are now being raised with the Hydroxychloroquine-COVID-19 study I think due to methodological failures in the rush to publish https://retractionwatch.com/2020/04/06/hydroxychlorine-covid-19-study-did-not-meet-publishing-societys-expected-standard/

It seems to me that some in the medical profession don't get basic notions of risk. He could be right but by the sounds of it he doesn't have a strong evidence base for his beliefs. If he is wrong and we followed his advice could lead to increased virus spread; if he is right then not following...

An interesting tweet from Vincent Rananiello about prediction of who will have respiratory failure. Which refers to this paper https://www.medrxiv.org/content/10.1101/2020.04.01.20047381v1.full.pdf

If anyone is interested in proximity tracing that preserves privacy then there is a report out by a number of cryptographers on how this could be done. I've not read it but the people involved are good. https://github.com/DP-3T/documents/blob/master/DP3T%20-%20Data%20Protection%20and%20Security.pdf

Kings collage have been developing a Covid symptom tracking App. It comes from the group that does the twins studies. One of the people involved in the Twins study (Francis Williams) presented at the recent CMRC conference which made me wonder if this app could be made into a long term tracking...

Judging by a tweet from Costello there may be issues with this report

I wonder if we have an issue because too many of the scientists who get the governments ear are those who are good at the politics and go into administration rather than those who are good scientists (who want to spend there time doing the research).