select search filters
briefings
roundups & rapid reactions
before the headlines
Fiona fox's blog

expert comment on the number hospital-based COVID-19 deaths on the 25th March

Prof Sheila Bird has analysed the data on the number hospital-based COVID-19 deaths on the 25th March.

 

Prof Sheila Bird, Formerly Programme Leader, MRC Biostatistics Unit, University of Cambridge, said:

“The table below shows that at least 179 COVID-19 deaths occurred in English hospitals on 25th March 2020; and still counting!

“ The media spotlight would have been on the 107 COVID-19 deaths in English hospitals that were reported on 26th March but only 16 of these had occurred by 5pm on the preceding day: 25th March.

1. Notice that, initially, ratio of death-counts on 2nd versus 1st report-date for the four death-dates (25th to 28th March) was [73+97+69+80]/ [16+38+29+41] = 319/124 or 2.6.

2. For the next two death-dates (29+30 March), this ratio increased to 277/72 or 3.8.

3. There is then an abrupt perturbation on 31st March. The reason is that, for the report-date of 2 April, the modal count of 159 was 3 days back rather than 2 days back as hitherto.

“These remarks warn that the reporting delay distribution is not particularly stable, as yet, so that we may need to allow at least 10 days for the reporting-in of COVID-19 deaths in English hospitals before the toll for a particular death-date is almost completely reported-in.

“Journalists, please remember: the count of deaths on a particular report-date includes deaths that have occurred during the previous 10 days and earlier still. Transparency about reporting-delays enables hospitals to work out local solutions to ensure awareness sooner rather than later about suspect COVID-19 deaths. Hospitals will respond but they can’t solve hidden problems.

“If I believed, which I don’t, that my 8-day total of 179 COVID-19 deaths on 25th March 2020 was the ‘final’ count for 25th March, then I might argue that since the 5-day total of 138 was 77% of the ‘final’ total of 179, then for 26th March I’d ‘estimate’ that 204/0.77 = 265 ( uncertainty: 245 to 288) would be close to the mark for 26th March – and substantially closer than the count of 168 which was reported on 27th March 2020 and presumed to have pertained, for the most part, to 26th March 2020.

“Let’s see what more the next three days’ reporting tells us about COVID-19 deaths on 25th March 2020 in England’s hospitals.”

 

All our previous output on this subject can be seen at this weblink: www.sciencemediacentre.org/tag/covid-19

 

Declared interests

None received.

in this section

filter RoundUps by year

search by tag