Four documents, one promise, three unanswered questions

Healthy Lives, Healthy People, published in November 2010, set out the government's long term vision for public health in England. Its stated aim was to create a "wellness" service, Public Health England, and to strengthen "both national and local leadership" (Healthy Lives, Healthy People, 2010). That placed the local authority at the centre of prevention and made public health a local duty rather than a hospital one. The Five Year Forward View of 2014 supplied the phrase the shift still uses. The future health of millions of children, the sustainability of the NHS and the economic prosperity of Britain, it argued, all "depend on a radical upgrade in prevention and public health", and it noted that local authorities now had "a statutory responsibility for improving the health of their people" (Five Year Forward View, 2014). It also fixed the problem in numbers: one in five adults still smoked, a third of people drank too much alcohol, a third of men and half of women did not get enough exercise, and almost two thirds of adults were overweight or obese. Smoking in pregnancy ranged from 2% in west London to 28% in Blackpool.

The NHS Long Term Plan of 2019 gave prevention a chapter of its own and a renewed programme. It ranked the leading modifiable risks as smoking, poor diet, high blood pressure, obesity, and alcohol and drug use, and set out action to cut smoking, to double enrolment in the diabetes prevention programme, to limit alcohol related hospital admissions and to lower air pollution (NHS Long Term Plan, 2019). The 10 Year Health Plan of 2025 made prevention the third of three shifts, from sickness to prevention, and promised to "tilt the NHS towards prevention" and to make "the healthy choice the easy choice" (10 Year Health Plan, 2025). Its stated ambition is to halve the gap in healthy life expectancy between the richest and poorest regions while raising it for everyone.

Read the four together and something is missing. Each is a promise about direction. None answers three questions: what is being prevented, for whom, and who pays for the years of benefit. That gap is the subject of this week.

What the evidence supports, and where it stops

Immunisation and screening are the offers the plan treats as immediate. Vaccines prevent an estimated 3.5 to 5 million deaths globally each year, and each year over 15 million people in England are invited to screening, with around 10,000 lives saved a year (10 Year Health Plan, 2025). The evidence on uptake is strong, and it is evidence about gradients. Flatt and colleagues used general practice data for 2,386,317 children and found that the absolute inequality in uptake widened for all five childhood vaccinations between 2019 and 2023. The six-in-one vaccine moved from -5.1% to -7.7%, the first MMR dose at age five from -3.1% to -5.6%, and the second MMR dose at age five from -9.6% to -13.4% (Flatt et al., 2024, BMJ). von Wagner and colleagues analysed the first 2.6 million invitations to the English bowel cancer screening programme and found uptake of 35% in the most deprived areas against 61% in the least deprived (von Wagner et al., 2011, International Journal of Epidemiology). The offer is evidence based. The take up is not neutral.

Smoking is where the design lesson is clearest. The Long Term Plan recorded 6.1 million people in England still smoking, with nearly half a million hospital admissions a year linked to it (NHS Long Term Plan, 2019). Smith and colleagues reviewed the evidence on stop smoking services and ran a national equity analysis. Quit rates were lower in more deprived areas, 52.6% against 57.9%, but reach was higher, 16.7% against 13.4%, so the overall effect on smoking was equity positive (Smith et al., 2020, Addiction). A service that reaches further into the groups with most to gain can beat the gradient.

Alcohol offers a smaller, cheaper and more honest effect. Kaner and colleagues pooled trials of brief interventions in primary care and found that at twelve months the intervention group drank about 20 g per week less than the control group, with substantial variation between the included studies (Kaner et al., 2018, Cochrane Database of Systematic Reviews). Against that modest effect, the 10 Year Health Plan puts the societal cost of alcohol harm at 27.4 billion pounds a year, over 1.2% of GDP (10 Year Health Plan, 2025).

Weight and physical activity show the pattern the rest of the week turns on. The NHS Diabetes Prevention Programme, evaluated by Valabhji and colleagues, referred 324,699 people and brought 152,294 to an initial assessment. Mean weight loss was 2.3 kg in the intention to treat analysis and 3.3 kg among those who completed the course (Valabhji et al., 2020, Diabetes Care). The programme worked, and it worked least well for the people it most needed to reach: younger age, female sex, Asian and black ethnicity and lower socioeconomic status were each associated with less weight loss. The physical activity figures in the 10 Year Health Plan are a promise rather than a trial: it estimates that physical inactivity costs about 10.5 billion pounds a year in NHS pressures, and reports that where investment in physical activity is designed with local people, inactivity rates are 2.5 percentage points lower in the most deprived places (10 Year Health Plan, 2025).

Social prescribing is where the week has to be candid. Social prescribing and link workers are named as core in both the plan and the neighbourhood framework. The evidence is weak. Bickerdike and colleagues reviewed 15 United Kingdom evaluations and rated every one at high risk of bias, with no controls, short follow up and no validated measures, concluding that the evidence "fails to provide sufficient detail to judge either success or value for money" (Bickerdike et al., 2017, BMJ Open). An exploratory randomised trial of link workers for people with multimorbidity in deprived urban communities, by Kiely and colleagues, recruited 251 patients and found no significant difference in health related quality of life or in anxiety and depression, and no cost effectiveness advantage (Kiely et al., 2024, BMC Primary Care). Husk and colleagues' realist review found the evidence base too underdeveloped to draw general conclusions about which models work (Husk et al., 2020, Health and Social Care in the Community). A plausible mechanism and a thin evidence base should mean funding as an intervention under evaluation, not as a proven one.

Why a neighbourhood prevention offer can widen the gap

The mechanism is that take up follows advantage. An offer that asks people to opt in is taken up first by those with the time, confidence, transport and trust to opt in, and they are rarely the people at greatest risk. A universal offer, well delivered, can therefore improve everyone's health while widening the absolute gap between groups. The immunisation and screening gradients above are the proof that this is happening for services to which everyone is already entitled. The 10 Year Health Plan names the same risk for weight loss medication, noting that as many as 1.5 million people are already taking GLP-1 drugs through private prescriptions and conceding the risk that these "become the preserve of those who can afford them despite those without the financial means having higher need" (10 Year Health Plan, 2025). The prevention paradox sharpens the problem: a measure that produces a small benefit for everyone produces its largest population gain from the many at low risk, so a programme judged on total benefit will look best when it is aimed at the already healthy. The counter example is Smith's stop smoking finding, where reaching further into disadvantage, and accepting a lower quit rate, still produced an equity positive result. The question is not whether an offer is universal or targeted, but whether it is delivered so as to reach the people least likely to opt in.

Who pays for the years of benefit

Prevention's benefits arrive over years, and they land on a budget that may not be the one that paid. The Long Term Plan was unusually explicit. Local government had become "responsible for funding and commissioning preventive health services, including smoking cessation, drug and alcohol services, sexual health, and early years support for children such as school nursing and health visitors", funded by central government from the public health grant, and "funding and availability of these services over the next five years ... directly affects demand for NHS services" (NHS Long Term Plan, 2019). It also priced the cost of inaction, calculating the extra costs to the NHS of socioeconomic inequality at 4.8 billion pounds a year in greater hospitalisations alone. The 10 Year Health Plan shows how far the time horizon stretches. It projects that its tobacco legislation will deliver cumulative savings to the NHS of 6.6 billion pounds and 64.2 billion pounds in economic productivity gains by 2100 (10 Year Health Plan, 2025). A benefit that arrives by 2100 appears in no three year plan; the service that pays for smoking cessation today sits in the local authority grant, and the budget that saves is the NHS.

Systems work around this in three ways. They pool money: the neighbourhood framework describes place based budget pilots in five mayoral strategic authority areas that "explore how public services can refocus onto prevention and early intervention through pooled budgets" (Neighbourhood Health Framework, 2026). They write risk into contracts: the framework says new risk sharing approaches "incentivise neighbourhood providers to deliver effective preventative care that reduces avoidable non-elective admissions". And they localise the decision. The honest limit is that the framework does not resolve the misalignment. It states that neighbourhood health "will be funded by rebalancing existing resources rather than relying on new funding", and it treats prevention as an area where systems are "expected to go beyond" the minimum if they choose. The money question is therefore handed to local leaders who pay now and bank a benefit for a successor.

Communication and swallowing as prevention

The sharpest example of a prevention outcome that the policy never names is aspiration pneumonia. Smithard and colleagues report that 940,341 people were admitted to hospital in England in 2018 with pneumonia, 44.45% of them aged 80 or over, and that a further 61,248 were admitted with aspiration pneumonitis, 93% of them over 80. They also found that 62% of the health services that responded do not, or only occasionally, screen older people for swallowing difficulties (Smithard et al., 2019, OBM Geriatrics). Swallowing difficulty is the mechanism, and speech and language therapy is the profession that assesses and manages it, in the home and the care home as well as in hospital. This is prevention in the plainest sense: it averts an admission, it belongs to the neighbourhood rather than the ward, and it is invisible to a prevention programme built around smoking, alcohol and weight. A neighbourhood offer that does not name dysphagia will not commission it, and the admission it could have prevented will be counted instead as an emergency.

Universal with intensification, or the most deprived first

That is the week's contested question, and the evidence supports a narrower answer than either camp states. Concentrating everything on the most deprived cohorts protects the gradient but forfeits the population effect, because the prevention paradox means most of the gain comes from the many at moderate risk. A universal offer without deliberate intensification widens the gap, and the immunisation and screening data show it happening even where entitlement is universal. The defensible position is progressive universalism: keep the universal offer, and design every element of it for the people least likely to take it up, as the stop smoking services were designed. The test is not the offer's reach but its distribution: did the subgroup at highest risk improve at least as much as the subgroup at lowest risk?

That test is the thesis made practical. A prevention shift means something at neighbourhood level only when it can name what is being prevented, for whom, and who pays for the years of benefit. The four policy documents name the direction and leave the three questions open. The evidence fills in the first two. The third, who pays, is the one the neighbourhood service has to answer for itself.