Coronavirus statistics: what can we trust and what should we ignore?
“On 27 March the Gov announced that 926 Covid-19 deaths had so far taken place in English hospitals, NHS England now reports that the true figure was 1,649”
That is 723 deaths out. 78% theguardian.com/world/2020/apr…
“On 27 March the Gov announced that 926 Covid-19 deaths had so far taken place in English hospitals, NHS England now reports that the true figure was 1,649”
That is 723 deaths out. 78% theguardian.com/world/2020/apr…
There is a time lag between a death and its report. Sometimes a time lag of weeks. It’d be helpful if we could understand the full extent of the delays and the reasons.
But knowing the number of deaths that occurred each day shows the true scale of the disease and the curve
But knowing the number of deaths that occurred each day shows the true scale of the disease and the curve
Different types of predictive modelling.
Important to have accurate and up to date data for effective predictions. That means large scale testing, accurate data in deaths, including deaths in the community, hospital and ICU admissions.
Important to have accurate and up to date data for effective predictions. That means large scale testing, accurate data in deaths, including deaths in the community, hospital and ICU admissions.
Keep an eye on excess death data (often a weekly comparison with the average of the last 5 years’ deaths in that week)
Due to pressures on public services lives will be lost avoidably for non COVID reasons
But other risks are reduced (eg road accidents, other respiratory deaths)
Due to pressures on public services lives will be lost avoidably for non COVID reasons
But other risks are reduced (eg road accidents, other respiratory deaths)
