Number 10 admits mistake over Covid-19 death-toll graphs

Published: 6 November 2020

Post Desk : Downing Street has admitted there was a mistake in graphs that showed deaths during the second wave of coronavirus could soon surge beyond the first peak.

Graphs used during Boris Johnson’s press conference on Saturday showed predictions of 1,500 deaths a day by early December.

The Government has since found “an error” in the graphs that made the numbers too high, and amended them to reduce the worst case scenario to just under 1,000 deaths a day.

A spokesman for the Prime Minister said: “The SPI-M [Scientific Pandemic Influenza Modelling group] central medium term projections remain the same for admissions for the next six weeks.

“But we accept there was a mistake in plotting the confidence intervals which is the blue shaded areas which we corrected as soon as identified. But there was no error in the underlying analysis.”

Asked if he accepted the graph was misleading, the spokesman added: “The consensus remains that without action we would breach the first wave of hospital admissions and deaths in a matter of weeks.”
The Telegraph broke the story overnight, leaving a Government minister facing a grilling from broadcasters this morning.

Business minister Nadhim Zahawi said it was right to share data with the public despite a death-toll forecast used to justify the second national lockdown having to be revised down.

Mr Zahawi told LBC: “There was different models and it’s right to share as much data as possible.

“The more data that we share the more scrutiny there is. That is a good thing in a democracy.

“We never said that this is a prediction of outcome, these are models that different groups are putting together. We never said these are absolute numbers.”

He said lockdown did not sit comfortably with him but that ministers had to act, with hospitalisations soaring to more than 12,000.

He added: “There is no way you can be a bystander and say oh that doesn’t matter because some of the modelling may have been wrong.

“No. I would say to you, yes I come up and I put my hand up and say yes of course we’ve put out some models we could always make them better. It’s right to share the data.”

It comes after it emerged that separate modelling — showing a worst case scenario of 4,000 deaths a day by the end of December — was based on outdated figures. These were also amended.

Tory ex-chief whip Mark Harper wrote: “How many more graphs from incorrect overstated figures have been used to try to justify a damaging, one size fits all national lockdown?”

Lib-Dem Munira Wilson said the Government should apologise and clarify when incorrect data has been shared.

She added: “Obviously one major data point which was key in their argument has now been revised and disputed.

“Clearly there were other indicators that were concerning. But if you are transparent and show your working-out actually people can much more legitimately, scrutinise, challenge and discuss so that we can make sure we are reaching the right conclusions and taking the public with us. Sadly at every stage they have failed to do so.”

Conservative MP Steve Baker told the Standard: “On the face of it this revision is a dramatic blow to the Government’s case. But it must be said the underlying forecast for bed use is still, I think, the same.”

He said standards of Government needed to be reformed and dramatically improved, particularly the communication of data. Referring to the 4,000-deaths dossier, he added: “Who on earth instructed scientists with integrity to wave incorrect death forecasts at MPs in the hope it would persuade them?”

A government source said there was no error in any of the “underlying analysis”.

A spokesman said: “The main consensus projection remains unaltered. The data still clearly shows, and the consensus remains, that without intervention we are likely to breach the first wave of hospital admissions and deaths in a matter of weeks.”