Last week, NHS Scotland’s statistics service (Information Services Division Scotland LINK) released the latest figures for cancellations of elective operations, i.e non-emergency surgery. The figures showed a seemingly high level of cancellations due to a lack of hospital capacity and a noticeable increase from last year. As you’d expect, the opposition parties and various newspaper outlets picked up on this news and published stories reporting the figures.
As is the way in Scotland these days, the radical anti-establishment movement of Scottish independence then went into motion – seeking a way to defend the decade long governing party. Not unusually, it was Wings Over Scotland who produced a rebuttal titled Data Without Information [LINK].
The Wings piece attempts to undermine the negative NHS Scotland figures by first comparing them to 2010 figures reported in this “Deadline News” piece [LINK]. It would seem that in 2010 NHS Scotland was cancelling 30 operations a day due to lack of capacity, whereas the latest figures show 20 cancelled operations a day… so everything is fine. Things have improved under our glorious government. It’s not news. SNP good, etc.
The crux of Wings’ argument is thus:
A June 2010 report by Deadline News revealed that of over 600 cancelled operations in Scotland a week, a quarter – roughly 150 – were cancelled by hospitals for reasons of capacity. (In 2010 non-emergency operations were only conducted on weekdays, so that’s effectively 30 a day.)
He then favourably compares this to the ISD Scotland figures, reported in The Times, which show cancellations now at 20 per day.
What Stuart Campbell, the author, has either failed to tell you or failed to notice is that he is not comparing like-with-like. Whereas the 2016 statistics only include “cancellation based on capacity or non-clinical reason by hospital”, the 2010 figures reported by Deadline News include cancellations “because the patient is unfit for surgery”.
Under the current ISD Scotland reporting system, operations which are cancelled due to the patient’s ill-health are covered by a second category: cancellation based on clinical reason by hospital. During the time-series available in ISD Scotland (May ’15 to Nov ’16), this second category tends to be between 50% and 100% higher than those cancelled “based on capacity or non-clinical reason by hospital”.
If you were to include both of these statistics, you would reach an average of 53 cancelled operations per day. Which doesn’t compare very favourably with 2010’s 30 a day…
It should be noted that “cancelled for clinical reasons” also includes a second category where the patient has not been “prepared for procedure correctly”. It’s not clear if this is also included in the 2010 stats, so it’s not possible to say with absolute certainty that we’re comparing apples with apples.
What is very clear is that Wings’ claim that cancellations have “gone down by a third” is based on at best a misunderstanding, at worst sheer deception. Either way, it’s wrong.
Not to worry, though, because Stuart is content that non-clinical cancellations have only gone up by “a whopping 0.4%… a very small sample and is a micro-fluctuation only”.
Whilst Stuart is right that the percentage of non-clinical hospital cancellations has “only” increased from 1.7% of total elective operations to 2.1%, this is a significant relative increase based on 19 months of data and over 570,000 operations.
In an attempt to allow for seasonal variations, I’ve plotted the changes year-to-year for months where this is data from both 2015 and 2016, then calculated the relative change from 2015 to 2016.
Every month is showing a marked increase in cancellations from the corresponding period in the previous year, with an average relative increase of 17%. Micro-fluctuation indeed.
But it’s fine. Even that is irrelevant because NHS England is worse, isn’t it? It doesn’t matter how bad something is in Scotland so long as it’s worse in England, right? To such an end, Stuart attempts to claim “cancellations are still around 10% lower in Scotland”. He even provides a handy link to this NHS England site for statistics [LINK].
Small problem. The statistics show nothing of the sort.
Here are the NHS England stats for “the number of last-minute cancelled elective operations for non-clinical reasons” (see note 1).
The average percentage of cancellations in NHS England for the last 18 months is 1.0% of planned elective operations. In Scotland, it’s 2.0%. Or double.
So Wings’ three basic attempts to undermine these figures just don’t stack up. There is no evidence that cancellations have reduced since 2010, there has in fact been a 17% relative increase since last year and the numbers aren’t better than NHS England.
It’s obvious that there are huge problems in NHS England. It’s equally obvious that there are huge problems in NHS Scotland. No doubt one performs worse than the other across the vast array of NHS services available but surely we should at least look to recognise problems if and when they do exist. This sort of government apologism in a desperate effort to protect a desired constitutional outcome surely does no-one any favours.
Contrary to what Wings would have you believe, the press reports raise perfectly legitimate questions about the level of canceled operations*.
Data without information, eh? Might have been nice to start with just the data.
Note 1 – both the NHS England and ISD Scotland statistics have the same definition for the timeframe of cancellations included in the stats.
For ISD Scotland [LINK]:
For NHS England [LINK]:
However, the ISD Scotland metadata document states that the stats are not comparable due to “definitional differences”
It’s not clear whether these “definitional differences” relate to other areas of the statistics, given NHS England only appears to provide data for cancellations for non-clinical reasons. The definition of these also appears to be the same as ISD Scotland:
To my reading, this appears to make these specific statistics comparable, at least indicatively. If you wish to discount them, then fine.
Note 2 – throughout this piece I have tended to use the short-hand “cancellations”. To be clear, when referring to the latest ISD Scotland statistics this is intended to mean “cancellation based on capacity or non-clinical reason by hospital”. It should be obvious why and where this short-hand is used but thought it worth mentioning.
*This previously read “… were accurate and raised…”. Stuart Campbell (Wings) quite rightly pointed out that he didn’t claim they were inaccurate, ‘only presented without context’. Hence I have edited this sentence.