Should data be popular?

It was once so easy.. Flattening the curve was a quick and easy answer to the pandemic and was dressed up to give us comfort. So long as we survived the initial onslaught we could return to normal. So Kitty needed to lie of lock down and things would be ok

This version of science enabled Government to suppress the disease and avoid the apocalypse that had been predicted by commentators with access to footage in Italian hospitals. In the early days of the pandemic, data was popular – because it led to simple messaging we could understand.

But now the model cannot be simplified as “catchily”. “Following the science” now involves developing a 3D model rather than a quick way out of the crisis. COVID-19 is here to stay and that is very uncomfortable. So some political leaders (notably in Belarus and the United States), find popular support in claiming that they can out-bluster science.

For all Donald Trump’s foibles, he does at least articulate a lowest common denominator. His latest utterances on the science suggest that it is only useful so long as it wins him the next election. Right now , the Trump line on data is that “testing makes things worse”, which suggests “what you don’t know, don’t hurt you”. That’s a very popular thing to say (though totally idiotic).

Trump’s dismissal of data is an extreme example of what we see in UK press conferences and both in the US and UK, science isn’t so much followed as put to work.

Unsurprisingly, most people are eager to disinter-mediate the politicians and go straight to the science. We look for trusted sources like Chris Witty whose Gresham Lecture I posted yesterday

The COVID-19 Actuarial Response Group is another trusted source and one that makes sense of data in determining the direction of travel. While the scientists are building a three- D model of COVID-19, the actuaries are trying to find future trends by making sense of the recent past.

Witty and the COVID-19 actuaries are not populist, but they are increasingly popular (for not being idiotic).

A trusted source

This is a short blog taken from the website of Dan Ryan, one of the COVID-19 Actuary Response Group.

Bringing together ONS weekly registrations today with predictions based on NHS England data highlights that over the 7 week period to May 1, we saw almost 45,000 extra deaths in England, with 22,000 COVID-registered deaths inside NHS hospitals and 13,500 in nursing homes and at home.

Those numbers are humbling, but they also highlight our lack of understanding. There is a gap of 9,000 deaths that need to be understood if the right decisions are to be taken by society.

Unravelling these numbers will undoubtedly be complex, and will involve

– consideration of the impact of lockdown itself (both positive and negative)

– estimating how much the presentation and treatment of both acute and chronic illnesses has been disrupted by the fear and confusion caused by the maelstrom of COVID-19

– using electronic health records to improve our assessment of where COVID-19 was the culprit or just an onlooker.

The gap may indeed have been wider on May 1. It will be wider now. And of course, the threat of COVID is not going away any time soon.

One further insight from the ONS data as we move into the world of more frequent (socially distanced) contacts with friends and family. Nursing homes saw their terrible toll peak 2 weeks after the hospitals – when attention was elsewhere.

The most vulnerable need to be protected. We have been warned.