‘June heatwave killed 440 a day at its peak, say climate scientists’ was the Guardian’s splash this morning, while the BBC reported that ‘more than 2,700 people may have died in exceptional May and June heatwaves’. The stories are based on a paper from Imperial College London, the London School of Hygiene and Tropical Medicine and the Met Office, which estimates 2,736 ‘excess heat-related deaths’ across the May and June heatwaves in England and Wales, with 553 heat deaths from 21 to 29 May and 2,183 from 18 to 28 June. Dr Clair Barnes, the report’s lead author, said: ‘These are big numbers and we don’t want to see this many people dying.’ Of course we don’t. But there’s a problem with the reporting: the estimates aren’t based on actual counts of deaths this year.
As the paper’s authors admit: ‘The rapid nature of this analysis precluded access to observed death counts for England and Wales.’ Usually, attributing deaths to heat means looking at how many people die in hot periods compared to a baseline. Instead, the paper’s authors looked at the temperatures the country experienced during the heatwaves and estimated the expected number of deaths caused by heat based on the relationship between the two from 2000-2019. They admit limitations of their approach: they assume a constant mortality rate throughout the year. They don’t account for adaptation to heat over time, nor the effect of heat on deaths changing throughout the year, as the first heatwaves of the year tend to have bigger health effects than later ones as people acclimatise.
The trouble is that models like this can be very wrong. Last summer – the warmest on record – modelling predicted 3,039 heat-associated deaths in England, based on the historical association between heat and deaths. But when the UK Health Security Agency looked at actual death registrations, it calculated that there had been 1,504 heat deaths, only half as many as predicted.
Of course, heatwaves do kill. When the Office for National Statistics and the UKHSA counted actual death registrations for the summer of 2022, the year the UK first recorded 40°C temperatures, they found an estimated 2,985 excess deaths in England associated with five heat episodes. Those figures are based on deaths that actually happened. But as Dr Barnes seems to admit, giving an accurate figure is not the point – advocacy is. She says: ‘If by putting out these estimates we highlight to people how dangerous it is and they change their behaviour next time there’s a heatwave and our estimates turn out to be high, I will be thrilled.’
Reducing deaths is a worthwhile aim. And when actual mortality figures are published, they will likely show excess deaths from the heat – perhaps even more than the modellers predict. But right now, contrary to what the headlines suggest, the figure of 2,700 heat deaths is a guess, not a body count.
Comments