Official death data doesn’t tell us whether junior doctors’ strikes cost lives
19 April 2023
What was claimed
Official data shows that strikes are one of the reasons why excess deaths rose in England in March 2023.
Official data shows a rise in excess deaths in England at this time, but this does not mean that strikes were necessarily a factor. A similar rise occurred at the same time in Wales, where there were no strikes.
Walkouts are among the reasons excess deaths tripled, according to official data
Any increase in excess deaths is a concern but it is wholly wrong to say that the strikes are the root cause
An article on the front page of the Telegraph on Friday 14 April claimed that strikes by junior doctors in March were part of the reason for a rise in the number of “excess deaths” observed at the time.
Excess deaths is a statistical term used to describe the difference between the number of deaths observed and the number that would normally be expected.
However, the British Medical Association (BMA), a trade union that represents the doctors, said in a press statement that strikes were not “the root cause”. In a tweet, the BMA also said “there was no change in mortality trends during strike action”.
The story was also reported by articles in the Daily Mail and Express, while an article in the Mirror claimed that the strikes did not cause a rise in excess deaths.
In fact, from the data available, it is not possible to say whether the strikes in England on 13-15 March had any effect on deaths at the time.
Although the number of excess deaths rose compared with the weeks before, it rose to a level that has often been seen in recent months. In England, the number of deaths also remained at a similar level in the fortnight after the strike period that the Telegraph considered.
Meanwhile in Wales, the number of excess deaths rose in a similar way at the same time, even though there was no doctors’ strike. The Telegraph noted this when quoting the BMA later on in its article, and the BMA also mentioned it in its press release.
Professor David Spiegelhalter, a statistician at Cambridge University, told Full Fact: “There are many possible explanations for the rise in excess deaths at this time, and no particular reason to attribute it to the doctors’ strike. Correlation does not imply causation.
“Compared to the long waiting lists and record A&E delays, I find it difficult to believe that a short strike could have such a large effect.”
For its article, the Telegraph analysed recent mortality data from the Office for National Statistics (ONS).
This data also includes a recent “five-year average” for the number of deaths registered each week (excluding 2020 because of the extreme number of Covid-19 deaths early in that year). This average is often used as a baseline that the number of “excess” deaths is said to differ from.
The Telegraph seems to have used this data to compare the number of excess deaths registered in England in the two weeks up to 24 March with the two-week period immediately before it. The article says it chose to look at a two-week period including the week after the strikes to account for the likely delay in death registrations. Deaths in England and Wales are normally registered within five days, but there can be longer delays, particularly if a death is referred to a coroner.
This comparison does show, as the article claimed, that the number of deaths in the period including the strikes was about 11% higher than expected, whereas the number in the two weeks beforehand was about 4% higher than expected.
This rise was the basis for the Telegraph’s claim that “excess deaths almost tripled” after the strikes.
Why did deaths rise?
Observing data is often much simpler than explaining it.
In this case, deaths could have risen because of the strikes, or because of something else, or because of a mixture or factors, or just by chance.
The Telegraph includes a comment from an expert acknowledging that strikes could have contributed to the rise in deaths. The article also includes the expert opinion: “Other factors could contribute to excess deaths, such as flu cases, hospitalisations and cold weather.”
In fact, we do not know whether the strikes were a factor. They might have been, but this data does not prove it.
The Telegraph itself also quotes an expert who says: “In a period of such consistently high excess death and incredible strain on the system, attributing direct responsibility is impossible.”
We contacted the Telegraph to ask them whether they agreed that the strikes were not definitely a factor. At the time of writing they have not responded. Since we contacted them, the subheading in the online version has been altered and now reads: “Walkouts among possible reasons excess fatalities have tripled, experts say”.
Could there be other explanations?
Doctors deal with life and death every day, so it is reasonable to ask whether more people might die when doctors go on strike.
However, weeks with deaths 11% above the five-year average have not been unusual recently.
In the year before these strikes, the ONS had already recorded weeks in England with excess deaths abovethatlevel 23 times.
Data published after the Telegraph article appeared also shows that in the fortnight after the strike period, the level of excess deaths in England remained at about 10%, even though junior doctors were back at work.
What about Wales?
The same ONS data from Wales, where doctors did not go on strike, suggests that other factors might explain the rise the Telegraph observed.
Deaths in Wales during England’s strike-affected weeks were about 14% above average—compared with a level about 1% below average in the comparison weeks beforehand.
In other words, there was an even more pronounced spike in excess deaths in Wales at the same time, which cannot be explained by strikes.
The Telegraph did not report this in detail, although it did say that the BMA council chairman, Professor Philip Bamfield, made this argument.
To be clear, the comparison with Wales does not prove that strikes did not contribute to excess deaths in England. It is possible that both countries experienced spikes in deaths at the same time for different reasons, or for many reasons.
However, the Welsh data does show that a similar spike could have happened in England even if the junior doctors’ strikes had nothing to do with it.