On the discontents of health systems: Part 2 Will Ozempic reduce healthcare costs?

This is the second of an occasional series on issues in health economics. It is not my intention to examine the details of the impact of Ozempic and other appetite-suppressant drugs. My intention is to use these as an example of the economics of preventive health measures. The UK and other countries are wedded to the idea that a focus on preventive healthcare will reduce the costs of healthcare in future, but most commentators are notably vague about the system-wide consequences of such interventions.

Let me start with an apparently separate question. Does the reduction of air pollution in cities save lives? That is the claim which underpins the case made by, for example, the Mayor of London that the introduction of controls on the use of older vehicles in cities - Ultra Low Emission Zones (ULEZs) – will reduce exposure to air pollution and thus mortality associated with various respiratory and cardiovascular diseases. In the case of London, it has been claimed that 3,600-4,100 lives are lost each year due to air pollution that would be eliminated by the ULEZ.

Anyone who thinks about such a claim for more than 5 minutes should realise that it is absurd. We all die – sooner or later. In aggregate, over an extended period of years the total sum of deaths in London, holding the population constant, must be the same whether or not the ULEZ is introduced. No lives will be saved because nobody lives for ever. It reflects very poorly on the competence and/or honesty of those who prepare or publicise such claims that they engage in what is little better than simple deception.

What is correct and relevant to the question of whether to implement a ULEZ is that some people may live longer if the level of air pollution is reduced. In addition, they or others may enjoy better health during the remainder of their life. Those are good reasons why controls on vehicles to reduce air pollution may be entirely justified, but none of that excuses the reliance on claims that give a bad name to snake oil salesmen.

Health economists use a variety of slightly different measures to assess the impact of health and other interventions on the average length and quality of life. Setting aside the technical details, the essential components are (a) average life expectancy, and (b) the impact of disability or ill-health on the quality of life. So, a policy that reduces air pollution in London may increase the average life expectancy from, say, 83 to 84 years. The average quality of life may improve because fewer people experience physical and other constraints on how they live because they no longer suffer (so much) from chronic respiratory or cardiovascular conditions. These numbers are purely illustrative. It would take a very large health intervention – and certainly not a ULEZ – to increase average life expectancy by 1 year.

I haven’t carried out a detailed study of the impact of air pollution on health in London, but I have done similar work for several cities in middle income countries. I would be extremely surprised if the impact of a ULEZ on air pollution comes anywhere close to the numbers claimed.

As a sanity check, Greater London has a population of roughly 9 million. The average all-cause mortality rate in the UK is roughly 1% per year, so the annual number of deaths in London is about 90,000. If reducing air pollution were to “save” 4,000 lives per year, the implication is that nearly 5% of total mortality is associated one way or another with air pollution.

There are two important lessons for preventive healthcare that can be drawn from this example. First, when discussing the impact of any healthcare policies we should focus on average life expectancy and quality of life, not spurious calculations of lives saved. Second, it is relatively easy for advocates of specific interventions to generate large numbers that purport to support huge claims for the benefits of their preferred policies.

Returning to more conventional preventive medicine, the widespread use of Ozempic is expected to reduce the incidence of poor health and premature mortality associated with obesity and excess weight. Rates of type 2 diabetes, which gives rise to a wide range of medical conditions, should decline. In addition, there are claims that Ozempic and similar products can lower rates of cardiovascular disease. There is clear evidence that early onset of diabetes 2 is associated with reduced life expectancy with reductions in life expectancy of 13-14 years for patients diagnosed at age 30 to 5-6 years for patients diagnosed at age 50.

The logic, therefore, of prescribing Ozempic for obese or overweight patients runs as follows. [Please note that the numbers used here are purely illustrative. In addition, I have made no attempt to calculate distributions that would be standard in a full analysis.] Suppose that 1,000 overweight or obese patient at age 30 are prescribed Ozempic. As a result, the number who develop diabetes type 2 is reduced from 400 to 300 and the average age at which they develop the condition increases from 40 to 50 years. For this group, the use of Ozempic would reduce the years of life lost (YLL) due to diabetes from 3,600 to 1,500. If we attach a value of £30,000 to each YLL saved, the total benefit for the group would be £63 million over, say, 50 years from age 30 to age 80.

The cost would be treating 1,000 people with Ozempic but for how long? A once-off intervention might be relatively cheap, but the evidence suggests that any weight loss is wholly or partly reversed once treatment ceases. If the average cost of the drug, patient monitoring, and other costs amounts to £1,500 per year, the annual cost of treatment would be £1.5 million per year for the group. The total cost of treatment would exceed the estimated benefit if the full benefits of treatment can only be realised by taking Ozempic continuously from age 30 to age 80. However, if the average cost of maintenance treatment is only £1,000 per year after the initial intervention, then the expected benefit would exceed the total cost for the group of £50.5 million.

The point of this example is that even the most basic calculations of costs and benefits are rarely simple. Adding distributional aspects such as differences in ages or responses to treatment means that conclusions are usually more nuanced than is convenient for political and media discussion.

This is only the starting point. Extending the life expectancy of overweight or obese patients by treating is a desirable goal, especially as living with diabetes and excess weight usually means that people can do less and get less out of life than they would without the effects of such conditions. Still, this gain does not come without additional costs on top of the simple costs of treatment. In 2025 total public spending on healthcare per person will be about £4,000 in nominal terms.[1] Thus, reducing the years of life lost by treating a group of patients with Ozempic will add about £8.4 million to the total costs of healthcare spending on these patients over their extended lifespans.

While not part of healthcare spending, the treatment will add about £25 billion to the cost of state pension payments for the group as most of the saving in years of life lost will be at the end of life during retirement. We should also take account of public spending on disability and other benefits. The net effect should be an overall reduction in expenditures on these programs. There is (old) evidence which suggests that complications of diabetes and obesity contribute to a substantial share of claims for various disability payments. However, estimating the magnitude of the impact would require a proper longitudinal study to collect reliable evidence rather than relying upon generic assertions.

The factors that I have outlined are only the beginning of a full assessment of the net benefits or costs of adopting preventive healthcare measures such as the widespread reliance on drugs such as Ozempic to mitigate the increasing rates of obesity and diabetes in the population. There are both financial and non-financial benefits that flow from their potential impact in reducing years of life lost as well as improving overall health and well-being for many people.

If we value these gains sufficiently highly – or can reduce the costs of treatment – the net economic and social benefits may be large. Even so, there is a large element of either naivety or deliberate blindness among both politicians and lobbyists who claim that such preventive interventions will significantly reduce the growth in healthcare and related costs. Sadly, it is all too likely that they will increase such costs and worsen the difficult choices that must be made if modern European states are to become solvent at some point in the future.

Finally, I should emphasise that I am not trying to make a case for or against the widespread use of semaglutides or similar molecules to manage obesity and diabetes. My point is that there are difficult trade-offs which must be made. Collectively and personally, we have benefitted greatly from the substantial increase in life expectancy over the last eight decades. However, all too often there is an easy adoption of mantras about preventive medicine. The questions should be asked in future require (a) more scepticism about whether its benefits are as large or unambiguous as usually claimed, and (b) some open discussion about how the costs should be distributed.

[1] For such a large sector, data on public spending on healthcare in the UK is poorly documented. Part of this is accounts for devolved administrations are rarely consistent with those for England. The average figure per person in 2025 is based on extrapolating ONS figures for 2023 to allow for inflation and real growth and using a population figure of 69 million.