UK Coronavirus Update 4 May
The Evolution of the UK Epidemic
As an interested scientist, I have taken data from the UK government and provided some analysis and graphics to track the spread and impact of the outbreak.
Here I track pillar 1 (hospital) testing and mortality
In the first chart I plot the number of deaths each day relative to total deaths up to and including that day (normalised second derivative of total deaths for the maths-inclined). This matches the analysis presented here. We want to see a downward trend in this graph – that indicates that the rate of increase in deaths each day is slowing, and therefore that total deaths will begin to level off. I have added a line showing the moving average, which smoothes out the day-to-day variation. The rate of decline has been slowing over time, meaning that mortality has continued for a longer period that the initial drop in this rate suggested.
Here I plot the rate of change (normalised) of daily new cases. Again we see that while this rate is falling towards zero, the downward slope is almost flat, extending the length of the outbreak. We don’t understand why the initial rate of decrease leading up to April 13 has shallowed in the following days.
In this next new graph I plot simultaneously the development of both positive tests (left axis) and daily deaths (right axis – note different scale). We can see signs now of both cases and mortality dropping in the hospital setting (pillar 1).
The total number of people tested in pillar 1 (at hospitals) has reached 630,000. The strange increase in testing numbers on 25th April is not explained in government figures.
The total number of positive tests in pillar 1 (hospitals) has reached 153,000 today.
The proportion of positive results has now dropped strongly in recent days, now down to ~10%, as the volume of testing has increased and daily new positive cases has fallen. This is encouraging as it means we are improving our capacity to contact tracing to take advantage of testing capacity.
The daily new cases is slowly declining since the peak a month ago. The under-reporting at weekends complicates the analysis, but the decline is clear.
Testing today in hospitals (pillar 1) has now exceeded 30,000 in recent days. Many more tests are undertaken in Pillar 2, the broader community testing.
The number of new cases was down slightly today from yesterday – as the moving average in the graph below shows, we are just around 1.0 in the daily ratio to the previous day, which shows that each day the number of positive hospital tests is similar to the previous day
The total number of deaths continues to rise, and is now at 24332 in hospitals.
There were 235 deaths today reported in hospitals, lower than any reported value in April.
Note on Dates: I use the date of reporting as the date associated with the data. In fact, the test data cover part of the previous day and up to 9 am on the day the data are released, while the mortality data cover the 24 hours up to 5 pm on the previous day.