Join Hafta-Ichi to Research the article “England lockdown was needed but graphs were a mess, says statistics guru | Coronavirus”
England needed dramatic action to reduce the spread of coronavirus, despite the government’s messy presentation of data to justify the latest lockdown, according to an expert.
Prof Sir David Spiegelhalter, a non-executive board member at the UK Statistics Authority, said one slide shown at last Saturday’s press conference suggesting a possible peak of 4,000 deaths was “really ghastly”.
He said the data was out of date and had never been meant to be part of any formal document.
Spiegelhalter told BBC Radio 4’s Today programme: “It has been a mess, it really has. All those graphs that got put up at the press conference last Saturday, the projections were out of date at the time, they’re definitely out of date.
“That one [the slide about 4,000 deaths] was really ghastly – that was out of date when shown. It was never meant to be part of any formal document. It was leaked early and then it was part of the briefing to MPs.”
He said projections did have some “validity” but needed to be “taken with extreme caution”, as they could often be out of date by the time they were shown.
“Another projection of short-term deaths up until 6 December, and there’s been accusations of a conspiracy, because it’s changed between the press conference and what’s been released now. The slides have changed for technical reasons, but it makes it look a bit less dramatic.”
The top-end projection of the revised graph did not exceed the peak of the first wave of coronavirus deaths, he said. However, “if you read the small print, there was only a 50/50 chance it could be in or outside that range”.
The statistician and chair of the Winton centre for risk and evidence communication at the University of Cambridge said projections were models of possible futures based on numerous assumptions and were usually out of date by the time they were shown. “So they have got some value but they have to be taken with extreme caution,” he added.
Spiegelhalter was asked about comments from Prof Tim Spector at King’s College London – which has been monitoring the symptoms and test results of millions of people through its Covid app – who said on Friday that the peak was likely to have already passed.
“We can’t rely simply on confirmed cases or daily deaths without putting them into context,” Spector said. “Hospital admissions are rising as expected, but deaths are still average for the season.”
Spiegelhalter said: “If this is going to go down, it is going to go down very slowly unless some dramatic action is taken, which has been taken. The point is we are getting about 20,000-25,000 positive tests a day, that feeds through to about 1,500 hospitalisations a day, about 250-300 deaths a day and these are broadly stable but going up a bit – the deaths in hospitals and hospitalisations are going up slowly – and we are coming into winter.
“Those sorts of levels, even if they stay very stable and below the first peak of the virus, unless they start dropping, we are stuck with those for months and it seems to me and others that that’s not going to be sustainable in terms of what the health service can deal with … so it can stay open for everyone else.”
The Office for National Statistics has indicated that the steep rise in new infections was levelling off in England and stabilising at about 50,000 a day, with an estimated 618,700 people having the disease during the most recent week to 31 October, up from 568,100 the previous week.
Source: The Guardian
Keyword: England lockdown was needed but graphs were a mess, says statistics guru | Coronavirus