COVID19 statistical analysis

UK COVID-19 forecasts from WHO reports up until  Day 91: 22/4/2020 Independent forecasts and charts: Cases: Pessimistic trend curve 386,804 Optimistic trend curve 292,973 Mathematical Model 342,337 Deaths:  Most optimistic 27,793 Most pessimistic 50,355 New charts uneil Day 93 but basic trend and forecasts have not changed in any significant way:

UK COVID-19 forecasts from WHO reports up until  Day 91: 22/4/2020 Independent forecasts and charts: Cases: Pessimistic trend curve 386,804 Optimistic trend curve 292,973 Mathematical Model 342,337 Deaths:  Most optimistic 27,793 Most pessimistic 50,355

The peaks of daily new cases have been reached and passed in most European countries now, at slightly different times. Her is a table showing the effect of the lockdown / physical distancing strategies and the time taken to reach a peak in new cases. What will also be very important is the strength of the declines after the peaks. It's OK to claim that things are improving when the R0 reproduciton number of the disease gets under 1 (past the peak) but it needs to get as close to zero as possible and weak policies will take much time to achieve this.

UK and Italy seem to be converging towards similar outcomes if you look at this style of chart which I find very useful: daily rate of change versus existing cases: Cases: Deaths: Spain and France added for comparison: See also 

14 April 2020 New forecasts for Italy based on different interpretations of the actual daily WHO data trends: Italy pessimistic cases 233,208 Italy optimistic cases 212,271 Italy log cases 242,202 Italy deaths  pessimistic  29,209 Italy deaths  optimistic  27,275 The pessimistic and optimistic lines are converging as might be expected as the forecasts become less volatile as new data points are added to the various trend lines.

Estimates of ICU / Critical Care Bed Usage in UK, updated 08 April, plus forecast curves:

UK forecasts as of 10 April total cases: 853,751 (pessimistic from WHO data) 452,856 (optimistic from WHO data), 412,913 (mathematical model). Deaths: 161,838 (from WHO data). The UK forecasts are shrinking but I don't fully trist this shrinkage of the numbers right now. I have seen this kind of volatility in the forecasts before and they often go too optimistic for a while. You need to take some kind of average over time. Anyway, here is the chart:

COVID-19 exit strategy 2 years alternating lockdown and release Here is the Imperial College research link and also an interview with another researcher. https://www.youtube.com/watch?v=R1wkV2IzR_U I wrote in a comment that Professor Gunther Shonrich is saying that herd immunity is basically inevitable so good luck. I think it probably is because the thing is so contagious. We are all going to have to take our chance with it, especially next winter. He is talking about things like the Imperial College study pages 10 & 12, note the 2/3 lockdown, 1/3 release diagram on p12:  https://www.imperial.ac.uk/media/imperial-college/medicine/sph/ide/gida-fellowships/Imperial-College-COVID19-NPI-modelling-16-03-2020.pdf The study has these figures. A picture paints 1000 words. They are in reverse order from that shown iin tht eoriginal paper: Alternating lockdowns and not-lockdowns until end of 2021 possible: How the epidemic curves are modified by some interventions but yo...

Italy and UK case and death forecasts 20200408: Charts:

Here are the curves for Italy, Spain and France. Italy peaked a little while ago, Spain next and possibly France after that, though their numbers are a bit more volatile from day to day: Current forecasts for Italy:  Cumulative cases: 164,665 to 239,162. Cumulative deaths: 30,364-32,954 I am very pleased with my forecasting tools especially in terms of their ability to calculate the peak numbers of new daily cases quite accurately in advance. One has to be a little bit careful to change the start points of forecasts in line with the new data and choose a point that is on the forecast equation line each time it is done, so it not an "automatic" process. Explanations on charts: Italy: Italy deaths: Spain: Forecast of total cumulative cases: 338,614 or perhaps fewer because I have not yet constructed an "optimistic" forecast as I did for Italy. I don't have the time time or resources to do everything! France trend has improved and the tot...

ICU/CC Intensive care critical bed usage in the UK Weekly ICNARC reports Intensive Care National Audit & Research Centre Reported critical bed used versus my forecasts from curve below using trend to 6 April. All forecasts based on WHO case data reports 53-78: This is the forecast curve (most optimistic, given WHO reports 53-78): (c) D. Bellamy 2020. Revision late night 08 April 2020 from forecast curves (range between optimistic and pessimistic forecast curves): 19&20 March ranges   108-124 and 142-152  26&27 March ranges   433-513  and 505-612 03&04 April ranges   1525-1673  and  1724 -1883 10&11 April ranges   3388-3428 and  3670-3770 17&18 April ranges   5310-5894 and 5569-6268. My curves have risen a little more steeply than the actual figures, so far - but it is early in the trend and forecasts are realistic compared to actual numbers from ICNARC.

Rest of World COVID-19 forecasts (Global total minus China#s original epidemic, because trend cannot be calculated with China included because their epidemic was earlier in time and needs to be treated as a separate system). These forecasts include some variations on the analysis for 31 March. There was a subsequent drop for the 5th April forecast due to some points that lowered the trend line. I have taken optimistic and pessimistic trend lines, which include different data sets. The optimistic set is biased more towards recent data (WHO report #59) whereas the pessimistic forecast (includes data back to WHO report #38). I have taken the geometric means of the pessimistic and optimistic forecasts to give an idea of that the mid-line forecast could be. Interestingly, the case fatality rate of the confirmed cases is around 5-14% which matches well with Italy for example (or the UK currently) in that 7-9% region, presumably because almost all mild cases are not tested: Author: D ...

I said on one of my YouTube videos that the day we see the Queen give a speech that we really are in trouble. Shortly after, our Prime Minister Boris Johnson was taken to Intensive Care at St. Thomas' Hospital in London with more serious symptoms of COVID-19 and by the next day he was in their Intensive Care Unit and has been said to be treated with standard oxygen supply. Our hopes and prayers are for Boris to get better and for all other patients of this deadly disease to get better. It seems to act like a poison on the body of those who become very sick and there is some research that is looking at how the disease may latch on to haemoglobin in blood to impair the transport of oxygen. It is strange that so many patients who develop Acute Respiratory Distress Syndrome (ARDS) or related sever symptoms have bilateral opacities (in both lungs at the same time and all over the lungs) whereas typical pneumonias are often confined to one lung or even one or some areas whilst not af...