It’s been an interesting day. First in a very good way – I attended a very interesting event ‘Journalists and the new health system’, met some very good people and ended up on the panel, and had a lot of fun.
The next part was no less interesting but somewhat less fun, at least initially. An email was forwarded to me that has been sent by Private Eye – a fairly aggressive and accusatory one (although things did mellow a little after a few responses had been exchanged) which asked whether I (and another writer who has blogged on Mid Staffs) was being paid by the Labour party to write my articles on Mid Staffs and HSMRs, and accusing me ( and the other writer) of attacking people who are ‘trying to expose poor care‘.
Both are nonsense (although I have certainly criticised the Eye’s columnist(s) for lazily assuming that the headlines about HSMRs were true instead of checking the facts first). But rather than me write on them at length, I’m going to show you the initial email, my response, and then the emails that followed, so you can judge for yourself.
I’ve deleted the details of the other parties to preserve their privacy, but the content of the emails is, as far as I’m concerned, fair game. Emphases are mine, and my responses are in blue:
From: xxxxx xxxxxx <email@example.com>
Date: 26 March 2013 10:57:17 AM GMT
Subject: Private Eye
Dear Ms xxxxx
I am writing for Private Eye and have noted both yours and Steve Pleb Walker’s tweets. Please could you tell me whether your tweeted opinions on HSMR are Labour funded – how much funding you receive from your local party – and what scientific basis you have for making any of the claims you do? Further are you linked in any way to the BMA?
It would be really helpful if you could respond, as poor care is not resolved by attacking those who might expose it. Or is it not? [sic]
Dear Mr XXXXX,
XXXXXX forwarded me your email. She’ll answer for her own part, but I will answer for myself. To your questions (with a few additional facts thrown in):
- I receive no funding from anyone for my blog, nor from any political party for any purpose whatever.
- My ‘claims’ are made based on evidence gathered from various sources, including the transcripts of witness evidence to the Francis inquiry. All my sources are fully stated in my articles, with links where applicable – as you’ll know if you’ve actually read them – so anyone is free to check whether what I’ve pointed out, and the conclusions I’ve drawn, are correct and valid.
- I am not linked in any way to the BMA.
- I have no link to Stafford or Cannock hospitals, no relatives work there, and I live a long way from Staffordshire.
It’s not a matter of ‘attacking those who might expose poor care’. Poor care is not at issue, at least in anything I’ve written. My articles acknowledge poor care very frankly and explicitly.
What is at issue is the mishandling and misrepresentation of statistics, and the damaging headlines that have been spun out of the misrepresentation. Media from the BBC to the Telegraph to – yes – the Eye have repeated as fact the idea that ‘hundreds of needless deaths’ occurred at Stafford hospital, but the statistics say no such thing.
Robert Francis took care to say that HSMRs cannot be used to extrapolate numbers of avoidable deaths, and Prof Jarman and Roger Taylor admitted the same in their testimony to the inquiry – yes Prof Jarman is now giving interviews to TV and press saying there were 20,000 avoidable NHS deaths on exactly the same basis that he acknowledged could not be used for Stafford.
Many of the media have a clear and negative agenda in proclaiming these headlines. I trust that is not the case with the Eye – but recent articles have still been extremely misleading.
‘Avoidable’ deaths occur in NHS hospitals every day – and in every other hospital and healthcare system in the world. Healthcare is intrinsically risky, and because it is delivered by fallible human beings things will be missed, or done incorrectly.
Did people die avoidably at Stafford? Without question. Was the poor care at Stafford resulting in an elevated death rate that could justify headlines of ‘hundreds of needless deaths’? Absolutely not.
The corrected HSMRS (reflecting proper depth of coding and fully-audited palliative care codes) were substantially below the national average.
Since the existence of poor care in parts of Stafford because of severe short-staffing (as recognised by the Francis report) is uncontested, the fact that the death rate was below average means the headlines should be telling a completely different story – one in which overstretched staff managed to hold the essentials together well enough to prevent the understaffing from increasing the death rate.
The headlines say something else, and the end result is that the people of Stafford look likely to lose their hospital altogether.
I’m trying to present the facts in a proper light to redress the balance a little, and I have attacked unfounded/ill-founded statements, shoddy interpretations and lazy/malicious journalism, not individuals.
Within a minute or two of sending my response, I received another (and answered it):
Thanks for that Steve. Perhaps you could set out your qualifications – any mathematical or statisical or medical – to make the assertions you do about lazy malicious and shoddy journalism.
I have asked the Labour Party today who have distanced themselves from the views of yourself and XXXXX XXXXXX.
Labour will have to take a wider political view than I (happily) do. I have no qualifications other than intelligence, common sense and a certain dogged analytical bent – nor is it remotely relevant whether I do or don’t. If you can show my facts and conclusions to be wrong, do so. My qualifications have no bearing on the correctness or otherwise of what I’ve written.
Thanks for that Steve. Are you happy for me to forward this to the letters page of the Eye?
As you should be aware, HSMR is based on the coding of hospital episode statistics. That process can be “gamed” and your claim that Stafford was below average is a bizarre one which implies to me you simply haven’t looked at the HSMR figures (before attacking them).
Your point about a ratio of actua/expect as opposed to data on actual deaths is a fair one. And the media have sometimes taken a ratio to denote something else. However, HSMR do show up problems confirmed by real intelligence in the hospitals – surgeons, doctors, nurses. Even in failing hospitals there will be good doctors who are unfairly maligned. Nevertheless, apart from your general intelligence and common sense it might also be worth speaking to people who work in hospitals.
Feel free – I’m going to post the exchange to my blog, so you’re welcome to print them in the magazine (which I buy, btw).
My articles address ‘gaming’, so your inference is incorrect. Gamingat Stafford is out of the question – the hospital’s statistics were audited by the Audit Commission and by Capita and graded 97-98% accurate. Gaming at other hospitals is possible – though more likely for income purposes rather than specifically to affect HSMRs – but the fact that they can be ‘gamed’ shows the system as it stands now and stood in the critical period at Stafford to be useless in terms of actually identifying genuine problem areas.
Hospitals might conceivably have poor HSMRs simply because they’re more honest than some others and are not gaming their figures. Or they might have genuine issues – or might just not be very good at coding, since the vast majority of coders are unqualified (and usually overstretched since theirs is a ‘back office’ function likely to be an easy target for cuts by people who don’t understand the potential consequences).
This is a tragedy, since HSMRs could be a vital tool – but only if money is invested in training coders and auditing coding strictly and regularly, to ensure consistency of input. Otherwise, ‘rubbish in, rubbish out’.
I do suspect you haven’t read most of what I’ve written, though – or else you’d know I’ve spoken to a lot of people in hospitals, both on the clinical side and an in-depth interview with Stafford’s unfairly-maligned (and proven by audit to be capable and rigorously honest) coding manager, who joined halfway through the debacle and ‘righted the [coding] ship’.
There have been hospitals with poor HSMRs who had clinical problems, it’s quite true. It’s also true that there have been hospitals with poor HSMRs and no substantial clinical problems, and hospitals with great HSMRs who, on inspection, were found to be riddled with problems. That’s the problem – HSMRs are too random to be useful, because of the problems with poor coding, lack of training and auditing, etc.
The rest of the emails so far then get into details of evidence etc that won’t make very interesting reading here.
By the point in the correspondence shown above, the tone seemed to have shifted considerably and were turning into an interesting and much less confrontational correspondence.
It’s been a fairly intense few weeks, one way or another. But if a more balanced, less damaging view percolates out into the public consciousness, I’ll consider it all more than worthwhile.