For the third time we've seen a flattening in the number of new cases for a few days only for it to spike up again on April 5th. I would have hoped to see some slow down by now due to lock-down procedures starting to kick in.
There was a sudden increase in the rate of fatalities last week - well above model predictions and well in excess of the rate predicted by the reported level of new cases. I suspect we are starting to see a wave of patients dying who have not been included in the diagnosed figures and may suggest it is from nursing homes and self-isolating people (see France below). I am going to let the model run for a few more days and will re-calibrate if the current trends continue.
At the current rate we will see 1000 deaths per day soon and 10000 total deaths by the end of the week.
France:
The situation in France has become rather confusing in the last few days due to large numbers of cases and deaths from nursing home suddenly being reported. May of these cases do not have specific dates attached, so cause a spike, whereas in all fairness the values probably relate to several weeks.
Without these additional data the trends were holding well and we were starting to see a slow down in cases, however, now we need to adopt a watching brief and see how the situation pans out over the coming week.
Italy:
The rate of growth in the level of new cases has been slowing in Italy for some weeks. We have seen a plateau in new cases and are starting to see early indications of a decline.
In the last few days we have seen the daily deaths being slightly lower than expected (approx 100 people per day) but we can expect these numbers to remain steady in the 500-700 range for some time due to the large amounts of infected patients in the system.
Expect total deaths to reach 20000 in about a week's time.
Spain:
New cases in Spain look to have flattened out and are hopefully starting to decline, but ideally I would like to see a few more day's data to be certain.
Peak daily new cases were of the order of 7700 per day - these will take some time to work through the health system and would be expected to lead to fatalities rates in the range of 800-1000 patients per day.
Current fatality rates have been lower than expected for the last few days, so I will re-calibrate the model here in a few day's time if the current trend continues.
Expect cumulative deaths to approach 20000 in about a week's time.
Germany:
Germany is included here for completeness. Due to wide-scale testing and a different reporting standard for deaths, the data is following a very different pattern to the rest of the world. The rate of new cases is starting to flatten off, but the large number of new patients on a daily basis suggests a potential ramp in fatalities in the coming weeks may be possible.
However, predictions for this country are extremely difficult and it will be interesting to see how the situation in Germany evolves.
Sweden:
I've included Sweden as a control case to monitor rates of infection following different government advice. At the moment there is a fairly rapid compound growth in new cases and it could take several weeks before we see the impact of new advice on the level of new cases.
The differing growth rates for new cases and fatalities on the log chart suggest there are a significant number of deaths occurring withing the un-diagnosed (or late diagnosed) population.
Expect total deaths to reach about 800 by the end of the week with the possibility of 100 deaths a day shortly afterwards.
USA:
The rate of new cases in the US is only just showing signs of slowing down - although the latest daily figures looks suspiciously low. (Weekend/Monday figures are often low in some countries but this is not usually the case in the US).
Earlier testing of patients appears to be driving a long lag time between new cases and fatalities with a long lag time used in the modelling process.
Expect the daily deaths rates to continue climbing this week due to the large number of new cases being added in the last two weeks - daily deaths could exceed 2000 per day by the end of the week, with 20000 total deaths in the same time-frame.
New York City:
New York City is well ahead of the trends observed for the total US with the daily rate of new cases fairly flat. The regular troughs seen in the daily rates correspond to Saturday and Sunday values, so the most recent dip (on a Saturday) is not a sign of a reduction in new cases. Hopefully we will see a turn-down in new cases over the coming days.
Daily deaths for last week were higher than expected, which has been observed in a few countries when the patient load is very high, expect the daily death rate to fall back to the 150-200 range in the coming week and total deaths to approach 3000.
Data Sourcing: worldometers and wikipedia
No comments:
Post a Comment