The economics of preventive health

I'm at the Big Tent Ideas Festival today talking about public health spending. The blurb for the event is as follows:

There is an old adage, “an ounce of prevention is worth a pound of cure”, however, in the UK we do not seem to practice this. Of the Government’s total Department of Health expenditure, almost 75% goes on funding NHS England (which largely treats people) compared to the 0.2% which goes on funding Public Health England (which aims to prevent people from getting sick). This imbalance between treatment and prevention expenditure highlights a problem - how can we prevent people from getting sick in the first place? This session will explore the prevention/treatment imbalance and ask how we can make an ounce of prevention worth a pound of cure. 

I don't know where the 0.2% figure comes from. I think it might be PHE's operating budget. It's certainly not its full budget which exceeds £4 billion, of which £3.2 billion is sent to local authorities as their ring-fenced public health budgets.

£4 billion is a lot of money, but overall spending on preventive health is even higher. When you take NHS spending into account, the total was £9.6 billion in 2015 (I can't get a more recent figure). The proportion of the Department of Health budget spent on prevention is five per cent, not 0.2 per cent.

Should we spend even more? A lot of outlandish claims are made by people in the 'public health' industry about the efficacy of their interventions. It is often claimed, as the BMJ says, that...

Every £1 spent on public health in UK saves average of £14

Every £1.00 spent on public health returns an extra £14 on the original investment, on average—and in some cases, significantly more than that—concludes a systematic review of the available evidence, published online in the Journal of Epidemiology & Community Health.

The recent cuts made to public health budgets in the UK are therefore a “false economy” and are set to cost an already overstretched NHS and the wider economy “billions,” conclude the researchers, who warn other countries to think again before going down a similar route to claw back cash.

If it were possible to save £14 of NHS spending further down the line by spending £1 today, it would certainly make sense to make a greater 'investment' in public health. Alas, that is not what the evidence shows.

A study published in the Journal of Public Health in 2012 looked at a wide range of public health interventions in the UK and found that only 15 per cent of them save money in the long term. This echoed the findings of a systematic review published in the New England Journal of Medicine in 2008 which found that fewer than 20 per cent of interventions are cost-saving.

People need to abandon the delusion that preventive health interventions cuts costs in the long run. They do the opposite. If you can prevent an 18 year old from falling off a motorbike and spending the rest of his life in a wheelchair, you'll save the health service money. But if you prevent someone dying at the age of 80 so they live to 85, with all the hip replacements, cataract operations and pension payments that requires, it will cost money. The bulk of modern public health work involves the latter.

So what's all this about a 14 to 1 return on investment? That comes from this study published from 2017 (regular readers may be interested to know that one of its authors was Simon 'Caps Lock' Capewell). It is based on two assumptions from health economics:

1. NICE considers a health intervention to be cost effective if it produces one quality adjusted life year (QALY) for less than £20,000.

2. The Department of Health values a QALY at £60,000.

At the £20,000 threshold, Owen et al. (2012) estimated that 85 per cent of public health interventions were cost effective (not cost saving), ie. they cost less than twenty grand for every additional QALY. When the same researchers updated their work in 2018, this had fallen to 63 per cent. Still, not too bad by NHS standards.

The 14:1 ratio comes from comparing the cost of intervention against the £60,000 value of a QALY. So, if an intervention costs £10,000 and produces one QALY, it has a return on investment of 6:1 (£60,000 to £10,000). According to Capewell's study, the average public health intervention costs one fourteenth of £60,000 (14:1).

That's all well and good, but it's not a financial return, nor is it a future saving to the NHS. Unfortunately, the authors of the study portray it as both. The text of the study contains gross misrepresentations like this:

‘First, even with the most rudimentary economic evaluations, it was clear that most public health interventions are substantially cost saving…’

‘Overall, the results of our systematic review clearly demonstrate that public health interventions are cost-saving, both to health services as well as the wider economy… The cuts to public health budgets therefore represent a false economy. They are likely to generate billions of pounds of additional costs to the health services and wider economy.'

They even have the nerve to write this...

‘An ROI [return on investment] of 14.3 implying a cash return of 1430% would sound too good to be true in the financial world.’

Indeed it would. The crucial difference is that the authors' ROI is not a cash return. It sounds too good to be true because it isn't true. It's a common misunderstanding which certain people in the 'public health' industry do nothing to correct.

The only valid use for an ROI in this context is to compare the cost effectiveness of different health interventions. In a world of limited resources, that is worth doing. So is there a case for spending more on prevention and less on healthcare? Possibly - it's difficult to tell from the data - but probably not.

NHS healthcare produces a QALY at an average cost of £13,000. That's a lot of money, as you might expect from something that uses highly trained professionals, expensive technology and a vast, centralised, bureaucracy. £13,000 implies an ROI of 4.6 to 1. If preventive health can produce an ROI of 14 to 1, it would make sense to put more money into it.

But it's not that simple. There are several important caveats.

First, the ROI for prevention provided by local authorities is only 4:1. This is significant because it is local authorities who spend the ring-fenced public health budget in England.

Second, there is no single preventive health intervention. The 14:1 estimate comes from averaging out a wide range of interventions which vary massively in their cost effectiveness. The most recent study found that the cost per QALY ranges from £300 to £82 million!

If 'public health' professionals are competent and rational actors, they will spend their first pound on the most cost effective intervention and spend their last pound on an intervention that is less cost effective but better value for money than any of the remaining alternatives. In other words, the more money they spend, the lower the marginal returns. At some point, it becomes less efficient to spend money on prevention than on healthcare. We cannot tell from the available data where that point is, nor whether it has already been reached.

If, however, 'public health' professionals are not competent and rational actors, the situation will be worse because they will waste money on projects which are relatively inefficient, if not absolutely useless. That, alas, seems to be the case in the UK where Public Health England spends money promoting policies which have no credible evidence of efficacy, such as food reformulation and minimum pricing, while local public health directors pursue policies such as fast food zoning laws that are demonstrably ineffective.

Third, cost-benefit calculations which depend on intangible benefits being bestowed on individuals fall apart when the intervention inflicts intangible costs on the individual. The intangible costs of traditional public health policies are trivial to nonexistent (eg. the slight pain of a TB jab), but they have become more significant since 'public health' turned towards coercive lifestyle regulation.

Assuming individuals understand the health risks of excessive drinking, smoking, sedentary behaviour and over-eating, the intangible costs of a lost QALY are priced in by the individual. They are making a trade off between the private benefits they enjoy from their risky lifestyles and the potential cost to their health. If, for example, I am prepared to take a 50 per cent chance of dying ten years early to enjoy my vice of choice, I am valuing my vice at more than £300,000 - (£60,000 x 10 = £600,000) divided by two - plus the out-of-pocket cost of buying the product.

If the state uses coercion to make me abandon my vice against my will, I will be worse off; I will have lost thousands of pounds of intangible benefits. People in 'public health' may not value those benefits, but you cannot make a cost-benefit calculation on the basis that life years are the only intangible benefits that matter.

In conclusion, it is impossible to tell whether preventive health requires more or less funding in general, but there are certainly areas of 'public health' that should receive zero funding.

Share

Tweet

A few facts about the UK economy

I'm on a panel at the Big Tent Ideas Festival today discussing this report by the IPPR which proposes 'fundamental reform' to the UK economy. In my view, the reforms are not particularly fundamental. Most of them involve setting targets, stating aspirations and creating more quangos. Most of it could have been written by Theresa May.

The report starts from several false premises and there are bound to be disagreements about the facts that won't be resolved in front of a live audience so I am posting some evidence here so I can point people towards it if they're interested.

'Most people are no better off than a decade ago' (IPPR, p. 6)

Not true. Provisional data from the ONS published last month shows that median household disposable income is well above the level of 2009 and has been for some time.

The ONS's last full (ie. not provisional) release showed that every quintile has a bigger income than it did in 2008/09, with the biggest gains going to the bottom (poorest) quintile.

I think the IPPR has confused wages with incomes here. At the last count, median wages were still £4 a week lower than their pre-crash peak, but this does not imply that 'most people are no better off'. The labour market has expanded greatly in the last decade. The median wage earner is not the same person as he was then, as I explain in this post and as the ONS explains here.

'The UK is the fifth most unequal country in Europe in terms of income...' (IPPR p. 6)

No, it isn't (original figures from here).

'...while inequality of wealth is even greater'. (IPPR p. 6)

This is true but misleading. Wealth inequality is greater than income inequality everywhere. But the UK is less unequal in this respect than most comparable countries.

'Many more people work in insecure jobs than in the past, with almost a million people on zero-hours contracts...' (IPPR p. 6)

Between 2012 and 2013, there was a huge rise in the number of people who said they were on a zero-hours contract. In fact, it was so huge that it could only be a methodological issue. Sure enough, it coincided with Ed Miliband and others getting the term 'zero-hours contract' into popular use. I have written about this before...

‘Zero-hours contract’ is a relatively new term for what used to be called casual labour. It was almost never used by the media before 2012. And so, although the number of people who said they were in these contracts more than doubled between 2012 and 2013, the ONS concludes that this ‘appeared to be due mainly to increased recognition and awareness of “zero-hours contracts”’. The fact that the trend flattens out after 2016 suggests that the term has now become universally understood.

.. Moreover, the majority of zero-hour contract workers seem to be satisfied with the hours they are given and there is evidence that they are more satisfied with their work-life balance than those on full-time contracts.

The shortcomings of the data make it difficult to establish whether there are more people in casual work than in the past, but, as the ONS explained on BBC More or Less earlier this year (from 9 minutes), the number of people working only a few hours per week has declined.

Share of GDP going to labour

The share of GDP going to labour has fallen since the 1970s, as the TUC often says. The IPPR says the share of wages in national income has dropped 'from almost 70 per cent in the 1970s to around 55 per cent now' (p. 11).

But if you look at the data, it is clear that it only rose to 'nearly 70 per cent' during an unusual period in the mid-70s when the economy was in crisis, inflation was at 20 per cent and GDP was low, so it is not the most representative era to compare it to.

Nevertheless, GDP share to labour has fallen. The IPPR blames this on the lack of bargaining power of trade unions, but the same decline occurred in more heavily unionised countries and to a greater extent. Indeed, the share of GDP going to labour has declined more in other countries than it has in the UK.

The IPPR don't claim that income inequality is rising or has been rising, but someone on the panel might. For the record, it isn't and hasn't - not for thirty years.

Nor do the IPPR claim that the UK's low rate of unemployment is due to people who hardly work being classified as employed, but it is a common misconception so it might come up. For the record, it isn't true.

Share

Tweet

Warnings on individual cigarettes - another scrape of the barrel

It looks like the tobacco control industry has settled on health warnings on individual cigarettes as their next make-work scheme. The idea goes back many years; it was discussed by the lunatic fringe of the anti-smoking movement in Canada in the 1980s. I mentioned it in my book Velvet Glove, Iron Fist in 2009 and again in 2012:

With plain packaging in place, the extremists have exhausted all of the options I listed in the final chapter of Velvet Glove, Iron Fist. What fresh lunacy will follow? Warnings on individual cigarettes? 

As I noted last year, this dozy idea made an appearance in the Scottish government's tobacco control plan:

This preposterous idea was first floated by a particularly deranged Canadian anti-smoking fanatic in the 1980s but was never taken seriously. The idea of changing the colour of cigarettes was mooted in New Zealand a few years ago but, again, was considered a joke. Only now, after every other idea has been tried, is it becoming policy in Scotland. The desperation is palpable.

Spotting a slew of research grants and the chance to extend their careers by a few more years, the dregs of the 'public health' research community have leapt into action, producing risible studies that are virtual carbon copies of the focus group-based garbage used to promote plain packaging. The latest was published this week. I don't advise wasting your time reading it but it's here if you want a reminder of what a Mickey Mouse field 'public health' academia is.

It was funded by Cancer Research UK so here's your regular reminder not to give them any money. There isn't much else to say about it, but I gave a comment to the Daily Mail when they asked for one:

'There isn't a single person in Britain who is unaware of the risks of smoking. People are bombarded with anti-smoking messages from the day they are born. 

'Cigarette packs are sold with graphic warnings covering two-thirds of the surface area. People smoke despite the risks, not because individual cigarettes haven't informed them about the risks.

'After years of being harassed by health zealots, smokers might relish the idea of setting fire to a health warning, but this is not a serious policy. 

'It is a desperate attempt by anti-smoking campaigners to keep themselves in business now that the campaign for plain packaging has ended.'

The lead author is serial trougher and chancer Crawford Moodie. If you want a laugh, check out his recent 'study' in which he throws leading questions at focus groups about adding audio warnings to cigarette packs:

This focus group study provides an understanding of smokers' immediate responses to cigarette packs which played a short health message when opened. Smokers generally viewed them as annoying or embarrassing, and some suggested the use of alternative storage.

.. Cigarette packs with audio messaging may have a role to play, now or in the future, as a novel way of communicating health and cessation information.

These people should be sectioned.

Share

Tweet

The nanny state - why do we put up with it?

I was interviewed by Douglas Carswell for his Room for Thought show recently. Check it out.

Share

Tweet

Public Health England's great chess board

Yesterday, the Daily Mail reported the findings of a study which found that packaged food in the UK is healthier than it is in eleven other countries. This isn't very meaningful because the authors averaged out tens of thousands of very different products.

Useless though it may be, it's the kind of aggregate data that Public Health England is relying on in its reformulation efforts, so I wrote about it for the Telegraph...

A study recently evaluated the nutritional quality of packaged food in twelve countries. It may surprise you to hear that Britain came top and the USA came second. The least healthy food was found in two of the world’s least obese nations: Hong Kong and India.

According to the research, average sugar content of packaged food in the UK is 3.5 grams per 100 grams; the lowest of the twelve countries studied and less than half of the 7.7 grams found in China. The overall energy content of Britain’s packaged food was also the lowest. With 252 calories per 100 grams, our processed grub is much less calorific than that of India, where the average is 380 calories per 100 grams.

These findings may strike you as counter-intuitive. After all, Britain is one of the fattest countries in Europe and the USA has the highest obesity rate in the western world. But crude averages don’t tell us much about the diet of individuals. In the UK alone, the researchers found 68,153 different packaged food and beverage products to study. In the USA, the number was 162,297. Taking an average from such a vast range of options is almost meaningless.

The average doesn't dictate what people eat. People have choice. Even under PHE's reformulated regime, people will still be able to switch from low-calorie food to high-calorie food.

After citing this sadly overlooked study, I end with a bit of Smith...

The great economist Adam Smith mocked central planners for believing that they can "arrange the different members of a great society with as much ease as the hand arranges the different pieces upon a chess-board". They forget that "in the great chess-board of human society, every single piece has a principle of motion of its own, altogether different from that which the legislature might choose to impress upon it."

It is a mistake to see the world through averages. The bureaucrats at Public Health England should lift their eyes from their spreadsheets, heed the warnings of Adam Smith and remember the power of human agency.

Do read it all. It's paywalled but you get two articles a week if you register for free.

Share

Tweet

Does alcohol really cause breast cancer?

The anti-alcohol lobby got very excited in April when a Mendelian Randomisation (MR) study suggested that there are no overall health benefits from moderate drinking. Put very simply, MR takes account of genetic differences to see if an epidemiological association has been confounded by genes or other factors.

The authors of the study chose to look at people living in a specific area of China where two genes associated with teetotalism are very prevalent. They didn't explain how these genetic differences explain the J-Curve and - more importantly - they didn't explain what relevance their findings had to countries like the UK where these genes are rare. As I said at the time...

In short, the study looks at a Chinese population and argues that the J-Curve showing lower levels of stroke risk for moderate drinkers is an artifact of genetic differences. Two genotypes - ALDH2-rs671 and ADH1B-rs1229984 - are identified as suspects. The problem is that the former 'is mainly absent among Europeans but is prevalent in populations in East Asia' and the latter 'is found in 19 to 91% of East-Asians and 10 to 70% of West-Asians, but at rates ranging from zero to 10% in other populations'. Since most of the evidence for the J-Curve comes from western countries, it is not at all obvious that this explanation would hold up outside of Asia.

Nevertheless, it was good enough for The Lancet which published an extraordinary editorial pronouncing the death of the J-Curve, declaring that there is no safe level of drinking, and calling for a Framework Convention on Alcohol Control. Hello confirmation bias, my old friend.

The 'no safe level' claim is largely based on epidemiological studies of drinking and breast cancer, which purportedly show an increase in risk from very low levels of alcohol consumption. The evidence for this, in fact, very weak and a new MR study - currently in pre-print and based on a large sample of people in the UK - has found that the relationship between drinking and breast cancer doesn't exist at all. Moreover, it found that there is a threshold (ie. a safe level) for the handful of rare cancers that are genuinely associated with alcohol consumption.

Alcohol was observationally associated with cancers of the lower digestive system, head and neck and breast cancer. No associations were observed when we considered those keeping alcohol consumption below the recommended threshold of 14 units/week. When Mendelian randomisation was used to assess the causal effect of alcohol on cancer, we found that increasing alcohol consumption, especially above the recommended level, was causal to head and neck cancers but not breast cancer.

Our results where replicated using a two sample MR method and data from the much larger COGS genome wide analysis of breast cancer. We conclude that alcohol is causally related to head and neck cancers, especially cancer of larynx, but the observed association with breast cancer are likely due to confounding. The suggested threshold of 14 units/week appears suitable to manage the risk of cancer due to alcohol.

Strangely, the temperance lobby alcohol research community hasn't said much about this, even though it was reported in New Scientist.

Share

Tweet

The madness of food reformulation

I've written about the government's food reformulation ruse for Spiked...

At the heart of the reformulation delusion is an ignorance of market forces, a deep suspicion of industry and a naive faith in the power of bureaucracy to remedy supposed market failures. One of David Cameron’s greatest mistakes as prime minister was creating Public Health England in 2013. This quango, which relieves the taxpayer of over £4 billion a year, was always going to attract ideologues and activists from the clown show that is ‘public health’ academia. These people are relatively harmless when confined to their echo-chamber conferences and rinky-dink journals, but are a menace when allowed off the leash. At Public Health England, they have real power and influence. It is telling that the only ‘stakeholders’ from civil society involved in the reformulation work are Action on Sugar and the Obesity Health Alliance, two mouthpieces of the fanatical Graham MacGregor, who flood the media with hysterical claims about the ‘shocking’ levels of various ingredients in normal, everyday food.

As Josie Appleton showed in her superb report for the IEA last week, these activist groups are the outriders of reformulation, working hand in glove with PHE to soften the public up for further interventions in the food supply. The bone-headed approach of these extremist pressure groups has been bought wholesale by the apparatchiks at PHE. They allow no room for personal autonomy. As they see it, the public will buy whatever products the food industry throws at them. For some mysterious reason, the industry has traditionally chosen to put lots of unnecessary fat, sugar, salt and, er, calories in these products. Therefore, all the government needs to do is to tell them to use saccharine and brown rice instead and the British public will lose weight without even noticing.

It is the kind of idea you might hear from someone who owns a collection of bongs, but thanks to Public Health England it is official government policy.

Do read it all.

Share

Tweet