Conservative party speech. 

I’m not going to check every sentence. I’m going to examine the bigger claims that I am concerned about.

“But the truth is that for many years we have had the lowest cancer survival rates in Western Europe.”

The most recent EUROCARE study published in 2015 examines the cohort of patients 1999-2007, i.e. is now 9 years past, followed up till 2008. Additionally, “Survival appears to correlate with macro-economic determinants, particularly with investments in the health care system“.

Of the 15 original EU nations, we are ranked 13th for funding. We are underspending on the NHS. The speech doesn’t say that the NHS will get any more money. Instead it focusses on

“We know with cancer the key is to catch it early. So every day, compared to 2010, we are doing 16,000 more diagnostic tests. As a result we’re starting cancer treatment for 130 more people every day and now have our highest ever cancer survival rates.”

This is not always correct. ‘Cancer’ in many instances is easier to treat if treated early. But it can also be treated needlessly, for example, many cases of DCIS diagnosed at breast screening, and many prostate cancers diagnosed at prostate cancer screening, will never go on to harm the person. Diagnosing more cancers which behave in benign ways will automatically improve survival rates (if you include lots of  people who have a cancer which will never harm them, it will look as though your service is very effective, when in fact it is giving no benefits.) Cancer survival has been steadily improving over several decades. It would be brave to attribute this to a particular recent set of political policies. (Mike Baum has pointed out that cancer survival metrics are pretty meaningless outside of RCTs; he rightly suggests that mortality corrected for age would be a better measure. And that’s even before we go into the differences caused by coding between different countries….)

“So our new cancer plan will introduce a maximum 4 week wait from GP referral to diagnosis; bring in Ofsted-style cancer ratings for CCGs; do more molecular diagnostics and immunotherapy and save an estimated 30,000 lives a year.”

Diagnosis can often be very difficult. There is an additional problem in that most people with ‘red flag’ symptoms of cancer actually don’t have cancer. The lowering of thresholds of referring people urgently has meant that fewer people being seen in urgent clinics have cancer. Locally, referring more people has meant that many services can’t manage the usual system of seeing the patient and then organising tests, are changing their way of interacting with patients – for example, not seeing them but sending them for a scan –  to exclude ‘their’ speciality of cancer and then discharge them back to their GP. This might serve some people well. It is likely to serve other people far less well. Diagnosis isn’t a simple ‘black box’. It can be complex and require inputs from several different specialists and generalists. I would like diagnostic tests to be faster, but I would like the right ones to be done, interpreted by the right people who are able to act on them appropriately and take responsibility for the patients’ care when the patient does not fit the clinical pathway they have been assigned (often). The target culture paired with lack of resources inevitably produces quality problems in the NHS.

Ofsted style ratings: this government is very, very keen on ratings. I have yet to see data which tells us how accurate or useful they are to patients and families, or qualitative data examining them for potential to cause misconceptions, harms or poor quality decisions on the back of them. Big data dumps tend to be unintelligent and can lead us to unsafe conclusions. Here’s the kind of work that has to be done in order to help people understand them usefully. As an example, targets to diagnose dementia: surgeries were given targets, financially incentivised to reach those targets, and then it was realised that these targets were inaccurate – the prevalence of dementia had been over estimated. What a mess.

Molecular diagnostics and immunotherapy is already happening. The 30,000 figure seems to come from this report which is rather light on calculations, and sets targets for e.g. bowel cancer screening take up (when the only ethical target is to reach 100% of fully informed consent.)

And it’s not ‘lives saved’ – it’s ‘deaths delayed’.

“So many will be shocked to know that, according to independent research, every week in our hospitals we have around 150 avoidable deaths.”

This seems to be citing this paper and I don’t think it’s quite true. The researchers made a judgement on likelihood of being an avoidable death on that occasions of more than 50%. They reviewed 100 sets of notes. Overall, they reckoned this at 3.6%. However, there is a big difference between auditing notes to see what went wrong and what could be improved and knowing that different care would definitely have made a difference ( they reckoned on a definite avoid-ability far less often, 0.3% in 2009, see table 1, and 1 case only in 2012/3.)

This isn’t to say that we shouldn’t reflect, learn improve: it’s more that we shouldn’t be ‘shocked’ (or more commonly, scared to go into hospital for relatively minor and effective procedures) by numbers that are presented without an explanation of the uncertainties. (Incidentally, the purpose of that paper was really to examine how good HSMR ratios are; not very, which hasn’t stopped Mr Hunt from using them frequently.)

“Other countries have the same issues, often worse than us actually, but why anywhere in the world do we accept these kind of statistics as somehow inevitable in healthcare when we would never do so in the airline industry or the nuclear industry?”

Well: I don’t think we do, but human error is human error. Humans are more likely to make errors when stressed, understaffed, tired, etc. As Phil Hammond says, if the NHS was a plane, it’d be like trying to take off with half the crew, one wing, and the engine on fire. The airline industry stops flying in bad weather – the NHS doesn’t have much of a choice.

“So we need to help them too – by dismantling a culture that prevents improvement and replacing it with one that supports it. A blame culture and not a learning culture.”

I question the validity of this argument from a man who has suggested that doctors work 9-5 ; who has promoted dementia screening despite the evidence of ineffectiveness and harm, who has alienated a generation of junior doctors….a culture of improvement and learning needs trust and mutual support. I don’t see it here. Instead we are proceeding with wasting millions on commissioning, tendering, PFI – these are huge, politically cultural problems that prevent improvement because they waste money.

“It’s simply not acceptable that according to 8 recent studies we have a ‘weekend effect’ which means mortality rates up to 15% higher for those admitted on or around weekends.”

The 8 ‘recent studies’ are presumably these : some are studies, some are reports from various people/agencies, some are borrowed from each other. He hasn’t updated this list with more recent studies which have contradicted the claims Hunt has made. Nor has the lack of relationship with medical staffing (the major reason, apparently, for the fall-out with juniors) been explained. And 15% is a relative risk. We really need absolute risk to give us better perspective.

“Another standard says that whatever day of the week, patients should be checked by a senior doctor within 14 hours of being admitted. Again pretty vital for patients. But again when we checked, happening in just one in ten hospitals.”

These standards are  here. In fact, a similar claim (referencing weekends) was made by Hunt in relation to weekend review earlier this year and firmly rebutted; in fact, on average, about 79% of  patients being seen by consultants within 14 hours of admission, see this excellent blog

“The background to the story is Hunt’s comment in the Commons on October 28th, that “currently, across all key specialties, in only 10% of our hospitals are patients seen by a consultant within 14 hours of being admitted at the weekend”. This simple statement managed to achieve two errors and one misleading claim. Close examination of the data source from NHS England (table 2.1) reveals –

1. Error: The available data refers only to emergency admissions, not all admissions.
2. Error: The data refer to all seven days of the week, and so nothing specific can be said about the weekend
3. Misleading claim: the statement gives the clear impression that few patients are seen within 14 hours by a consultant. In fact, over all the reports in the NHS England dataset, the average is 79%. By using a very stringent criteria of hospitals in which all specialties report at least 90% of patients seeing a consultant within 14 hours, Hunt manages to cherry-pick a statistic that gives a very negative view of the situation rather than reasonably describing the patient experience.” (from Prof David Spiegelhalter’s blog) . Also, this. The ‘we checked’ comment rankles – there were very few case notes examined. It’s not robust research.

“I say to the BMA and all junior doctors let’s not argue about statistics or whether we can do more to raise standards for patients.”

Hmm. Well: arguing about statistics is pretty important. It would be possible to spend the entire GDP of the country on the NHS and waste every penny if we didn’t pay attention to statistics and use them safely and wisely. It is not a good idea to repeat statistics which can be misleading or inherently badly uncertain and misleading without acknowledging either.

A better culture means trust, trust means honesty, it means owning up to errors and trying to do better. I screw up, we all do, but we do not learn or improve if we continue to repeat our errors and create avoidable harm – whether a doctor or a politician.

after note: The more I think about this, the more upset I feel. Not arguing about statistics would infer that we should simply just believe what Mr Hunt tells us. In fact, the statistical messes that have been created needlessly, and gone uncorrected, have been the hallmark of this particular health secretary.

Now: I voted no for Scottish independence, I have no dog in this fight: but Shona Robinson as Scottish Health secretary is worth comparing in terms of attitude and engagement with staff. See Catherine Calderwoods’ realistic medicine report. There really could be better ways to work with policymakers which doesn’t result in repeated basic errors.

“In fact safer care doesn’t cost more, it costs less”

This is true and also untrue. Safer care saves avoidable harm and litigation. But you have to spend to get it. The need for safer staffing despite working in the face of budget cuts was a crucial finding of Francis. We are now hearing from hospitals told to curb spending on staffing to make efficiency savings. The debacle over the evidence on safer staffing, let alone the lack of statutory implementation, has been dreadful despite the evidence that it improves quality of care (though there are important gaps in the evidence that do need to be addressed).

“When you look at our safest hospitals, our best schools and our top police forces and you see it’s not about the level of funding, but the quality of leadership.”

Hmmm. If widely regarded as good CEOs are leaving and the average tenure is 2.5 years and with a third of trusts having a vacancy/interim in a director role, is the only thing standing in a Trusts’s way of making the austerity cuts and reaching the targets (despite the freefall in social care) a good leader? Or are they leaving because it is often impossible? Evidence, please.

“Training a doctor costs over £200,000. So in return we will ask all new doctors to work for the NHS for four years, just as army recruits are asked to after their training.”

Evidence based policy making should ask: what are the harms of imposing further working practices on doctors after imposing – sorry, not imposing – a deeply disliked contract on juniors? Will this improve loyalty to the NHS? Or might there, just perhaps, be better ways to do it (given that most doctors not only have a vocational career in the NHS but routinely work way longer than their contracted hours to do so?)

Bottom line:

Careless attention to statistics detail: a belief that leadership can compensate for austerity; and some new junior contract impositions. That would be a fail.