Super small sample, but that’s immensely encouraging.
KevinC
5674
After working under Trump for this long he might find COVID-19 a better alternative.
Wait, that’s not a photoshop?
I don’t even know if super small covers how super small it is. But the odds ratios are significant enough that the p-value is 0.001. This obviously calls for replications in larger patient cohorts, and I fear some regression to the mean, but from what I’ve read this is a cheap drug that could be cranked out.
Mr.GRIM
5678
Worth noting, Chinese Studies have found the same thing. So the sample size is not so small as you might think. Like this one from a month ago. https://www.ncbi.nlm.nih.gov/pubmed/32074550
KevinC
5679
That press conference… this is like a nightmare scenario where Idiocracy had a baby with Outbreak and it became reality.
Its the worst I’ve seen from him.
What’s also notable to me (and I am most certainly NOT smart about anything having anything to do with virology) is the notable reduction in viral load at Day 3. And why I say that is that it’s possible to make assumptions on this data that some of the good outcomes are the result of the disease running its course.
But that movement in viral load from day 2 to day 3 suggests to me – as a graph reader and stats dude – that there’s some motive force creating that delta between it and the control group.
The talk I had my notes for this morning is now up for everyone to watch. We had a pretty good discussion in lab, and I want to point out something about the R0 values calculated.
Here’s the data in question with the time:
Notice that one X on about feb 1st that’s way above the other points? There’s a question from a viewer at the end of the talk on if that is an outlier or not. If it’s an outlier, than it really changes the slope of the like and the R0 of social distancing to be less than 1. The speaker insists it is NOT an outlier, so should be included - something about them shifting patients from stay at home to the large stadium hospitals.
Sometimes a single measurement can have a huge impact on what your plan is going forward. I consider this my grain of salt for the talk.
UK is closing pubs, restaurants, theatres and nightclubs after tonight. To be reviewed “each month”.
Johnson asked those to close “as soon as they reasonably can” tonight. They won’t re-open.
Jesus. This press conference is brutal.
I’d love to see a bigger data set with confidence intervals. One of the things we’re not measuring (yet, but my lab has already worked this out) is a quantitative measurement of viral load. One might imagine if you’re knocking down the viral load by some factor per day with treatment (like 1/2) then people who started off with lower loads will fall below the limit of detection. Then, as people who have higher loads are treated longer, they also get knocked out.
I’d also like another plot that was person by person, to see that people who were negative one day did not test positive the next. They have a supplemental table that could be converted into that plot. I picked out the first 6 rows to highlight row6 bouncing back and forth, but it looks like between positive and ND, which I guess means “No data” (I had originally thought that was a negative, but it’s just the test failing I guess.)
newbrof
5687
(no need to be sorry, it is acutally a Demons Souls reference), there is no place like Boletaria in Germany, I would live there otherwise
The UK government will also cover 80% the salary of employees furloughed because of the Corona Virus, up to £2,500 a month. They will pay the employers directly to help them retain their employees.
Also, no VAT whatsoever due this quarter.
Trump just keeps getting angry at almost every question, its a horrible look, for someone who should be conveying confidence and calm.
Suck it Libs! Coronavirus is your fault!
What could go wrong with that plan…?
