UK Lockdown and cancer - a look at the data

From frasernelson.substack.com

By Fraser Nelson

Yes, diagnoses fell. But data so far shows no evidence of a death surge

One of my criticisms of lockdown was that people would not present for cancer screening - told to ‘save’ the NHS by not using it - and that we’d end up with more late presentations and, ergo, more deaths. In Jul20, a modelling study in the Lancet Oncology estimated that diagnostic delays would cause between 3,300 and 3,600 additional deaths. Bowel cancer alone was projected to see up to a 17pc rise in deaths.

Six years on, what does the data say? Cancer data lag badly, so the latest finalised year is 2023 (published Oct25). I did a lot of data work during lockdown outlining potential risks. And while lockdown did lead to lasting damage in economics and social trends - school truancy, debt surge etc - it seems a surge in cancer deaths will not be one of its side effects. Here’s why:-

Why surging is not, actually, surging

You may have seen headlines that cancer is at an all-time high with a new diagnosis every 80 seconds etc. That’s technically true - more than 400,000 people a year are now diagnosed across the UK, a record - but that’s because we’re a healthier society with more people living longer. Older people are more likely to get cancer. Here’s the scary unadjusted graph:-

But oncologists adjust for age, to compare like with like and work out a comparable trend. The results are below: two not-comparable series but both showing flat rate.

In men, detected cancers haver barely moved in a generation; in women it’s up modestly, which literature ascribes to better breast-cancer monitoring. Per person, your risk of cancer is not soaring. For both, the latest trend suggests it’s lower than 2013. There is no sign of serious progress, either - which is disappointing given advances in medical science and how much money is being spent on cancer research. Cancer bucks a trend of falling, general age-standardised mortality. Below shows a century of progress both here and in the USA.

The lockdown cancer disruption

A Piccadilly Lights poster reminding people to test for lumps or any cancer scares during lockdown - to counter the “stay at home - save lives” message put out by No10 over protests by NHS England

The original lockdown point - that people would stop coming forward - was, alas, borne out. Look again at that first chart: after fifty years of steady climbing, there’s a single sharp notch downward in 2020. Cancer registrations in England fell 12pc that year, to 289,000 - roughly 38,000 cancers that, in a normal year, would have been found and weren’t. The fall was steepest where you’d expect: in the spring of 2020, breast diagnoses dropped by about half, lung and prostate by a third. Screening had stopped; people stayed home. And those cancers were caught later - the share diagnosed early, at stage 1 or 2, slipped from 54.9% in 2019 to 52.3% in 2020. On the mechanism, the lockdown sceptics were right.

The cancer detection bounceback

Lockdowns were still going on in 2021, but cancer registrations sprang back to their pre-pandemic level. By 2022 and 2023 they resumed their record highs. Early diagnosis recovered too, and then some: by 2022 the share caught at stage 1 or 2 reached 56.8%, the best on record. The Covid dip was a single bad year, not a lasting scar.

There is no sign of the extra deaths. The age-standardised cancer death rate in England has kept falling, right through the pandemic, to the lowest level ever recorded. For men it dropped from 338 per 100,000 in 2013 to 294 in 2023; for women, from 231 to 208. Cancer mortality is now about 29pc below its 1989 peak. Whatever the diagnosis delays cost, they haven’t produced a visible wave of excess deaths. yet.

Could extra deaths arrive in 2024 and 2025?

For obvious reasons. cancer deaths lag diagnosis by years, so if the 2020 delays are going to kill, some of it would surface in data we don’t have yet - 2024 and 2025 registrations and deaths, due around 2026-27. This isn’t a closed case.

But there are grounds for cautious optimism:-

• The backlog cleared fast: the cancers missed in 2020 were mostly found in 2021 and 2022, not left to fester for years.

• The lockdown diagnosis dip was small and had fully reversed by 2022.

• The mortality rate up to 2023, already three years past the disruption, shows no upward flicker at all.

• And 2023 is not premature for deadly cancers that make up about half the death toll: lung, pancreatic, oesophageal. They kill quite quickly. Any serious impact from that lot as a result of the 2020 diagnosis dip would already have shown up by 2021-2023. They didn’t.

Of course many cancers (breast, prostate) play out over five and ten years, and a modest excess could yet be hiding inside a falling trend. Or, perhaps, preventing what would otherwise have been a fall. The 2024-25 data will have more clues, and I’ll return to it when it lands in October. But from what data we have so far, we can the that the lockdown diagnosis dip was real and inexcusable, the recovery was faster than expected and the cancer-death surge that was warned about has not so far presented in the data. There are, now, reasonable grounds to hope that it never will.

https://frasernelson.substack.com/p/lockdown-and-cancer-a-look-at-the