Mental Health Economics

Why this matters in health economics

Mental disorders account for a large and rising share of the world's non-fatal disease burden. They typically begin early in life, run for decades, and strike hardest at exactly the years when people are working, raising children, and building the skills a society depends on. That timing is the reason mental illness is an economic problem and not only a clinical one: its costs land as much on employers, families, and the wider economy as on the health budget, and a great deal of the loss never touches a clinic at all. A director who thinks about mental health only as a line in the treatment budget is looking at perhaps a third of the picture.

The discipline has a specific difficulty here. The standard machinery of health economics — the quality-adjusted life year, generic preference-based questionnaires, cost-per-life-year-saved reasoning — was built around conditions where you extend survival or restore physical function. Mental illness rarely kills quickly and rarely shows up as lost mobility, so the machinery systematically under-reads its severity and under-values its treatment. When your ruler cannot measure the thing, the thing loses every funding round it enters, quietly and repeatedly.

Three facts make this urgent for anyone allocating resources. First, mental health services are chronically under-funded relative to the burden they carry, in almost every system that has measured it — the gap between need and spending is wider here than for any comparable disease group. Second, most people with a mental disorder receive no treatment at all, and in low- and middle-income countries the untreated majority is overwhelming. Third, much of the return on treating mental illness shows up outside the health sector, as people staying in work and out of the justice and welfare systems, which means the body that pays is rarely the body that gains. Get the economics of mental health wrong and you do not merely misprice a service; you entrench a distribution of neglect.

Core concepts

Mental health is not simply the absence of illness but a state in which a person can cope with normal stresses, work productively, and contribute to their community — which already signals why its economics reach beyond the clinic. Mental disorders range from common conditions such as major depressive disorder and anxiety disorders, which affect very large numbers of people at moderate individual severity, to severe conditions such as schizophrenia and bipolar disorder, which are rarer but intensely disabling and costly. The economic shape of the two groups differs: common disorders dominate the aggregate productivity loss because so many people have them, while severe disorders dominate per-person service cost.

The first distinctive problem is outcome measurement. Generic preference-based measures such as the EQ-5D, which underpin the quality-adjusted life year (QALY) in many health technology assessment systems, describe health mainly through physical dimensions — mobility, self-care, usual activities, pain — with a single item for anxiety or depression. For a person whose disability is entirely psychological, that structure has little room to register change, and studies repeatedly find such measures less sensitive to mental-health improvement than condition-specific instruments. Analysts respond with mental-health-specific measures — for example the ReQoL and the Recovering Quality of Life family developed for this purpose, or clinical scales such as the PHQ-9 for depression — and sometimes with broader wellbeing or capability measures. Capability and wellbeing measurement is the province of Chapter 3.5 — Capabilities (which covers the ICECAP instruments and the welfarism/extra-welfarism debate); this chapter simply flags that mental health is one of the strongest cases for looking beyond the standard QALY, and the general mechanics of the QALY belong to Chapter 2.1 — Economic Evaluation.

The second concept is the weight of indirect cost. Mental illness damages human capital — a person's accumulated capacity to work and earn — and it does so through two channels that a clinical trial rarely measures. Absenteeism is time lost when people are too unwell to attend work. Presenteeism is the more insidious loss of being present but impaired — concentrating poorly, working slowly, making errors — and in mental illness the presenteeism loss is generally the larger of the two, precisely because these conditions degrade the cognitive and emotional functions that work depends on while leaving the body able to turn up. Because these costs fall on employers and the wider economy rather than on the health system, a health-system-perspective evaluation cannot see them, and a societal-perspective evaluation of mental health treatment often looks far more favourable than a narrow one.

The third concept is parity of esteem: the principle that mental health should be valued, resourced, and treated on equal terms with physical health. The reality is chronic under-resourcing relative to burden — expressed in law as mental health parity requirements in some jurisdictions, and as a policy aspiration in many others. Underpinning it is the burden of disease evidence, usually expressed in disability-adjusted life years (DALYs), which shows mental disorders carrying a share of total burden far larger than their share of most countries' health spending.

The fourth concept is the treatment gap: the proportion of people with a disorder who receive no treatment, or no adequate treatment. It is large everywhere and enormous in low- and middle-income countries, where the field of global mental health has grown up to address it. The World Health Organization's Mental Health Gap Action Programme (mhGAP) responds by equipping non-specialist primary-care workers to deliver mental health care — a form of task-sharing that stretches scarce specialist capacity. Contrasting responses include England's Improving Access to Psychological Therapies programme, now NHS Talking Therapies, which built a large trained workforce to deliver evidence-based psychological treatment at scale, partly justified on the economic argument that treating depression and anxiety pays for itself through restored employment.

The final concept is the long shadow of deinstitutionalization — the twentieth-century shift from asylums to community mental health services. Where the promised community funding followed the patients, it worked; where it did not, costs were shifted onto families, homelessness, and prisons rather than saved, a cautionary tale in cost-shifting that still shapes today's debates.

Best practices

1. Take a societal perspective for mental health, and make the indirect costs explicit. More than in almost any other area, a health-system-only view of mental illness misses most of the economic story, because the largest costs are lost productivity and informal care that fall outside the health budget. Run the analysis from a societal perspective, or at minimum present health-system and societal results side by side, so decision-makers see that treatment which looks expensive to the payer may be strongly cost-saving to the economy. State clearly whose costs you are counting, because the perspective — not the treatment — often decides the answer.

2. Measure presenteeism, not just absence. If you count only days off work you will capture the smaller half of the productivity loss and miss the impairment of people who are present but unwell. Use a validated instrument that captures both absenteeism and presenteeism, and be transparent about the method used to value the time, since presenteeism estimates are sensitive to how you measure reduced output. Never assume a person at their desk is a person working at full capacity.

3. Choose an outcome measure that can actually detect mental-health change. Do not default to a generic preference-based measure and then conclude a service "shows no benefit" when the instrument was too blunt to register it. Pair or replace the generic measure with a condition-specific or mental-health-specific outcome, and where the value of a service is recovery, autonomy, or social participation rather than symptom reduction, consider a capability or wellbeing measure (see Chapter 3.5 — Capabilities). Decide the outcome space from what the service is trying to change, before you pick the tool.

4. Map the whole-of-government costs before you claim a saving. Untreated mental illness generates costs in welfare, housing, criminal justice, education, and children's services, and effective treatment can relieve them. Where you can, quantify these cross-sector effects, because they are frequently larger than the health-care costs and they are the strongest argument for investment. Be honest, though, about which budget captures the saving: a health payer investing to relieve a justice-system cost needs a mechanism to be repaid, or the business case will not survive contact with the finance director.

5. Treat prevention and early intervention as an investment with a lag, and model the lag honestly. Most mental disorders emerge in adolescence and early adulthood, and intervening early — in early intervention in psychosis services, school-based programmes, or perinatal mental health — can change a lifetime trajectory. But the returns arrive years after the spend, so standard short-horizon return-on-investment tools understate them badly. Use a long enough time horizon to capture the downstream benefit, apply discounting transparently, and flag that the case for prevention is real but back-loaded.

6. Design for the treatment gap, not only for the treated. A cost-effectiveness ratio per treated patient is irrelevant if the service reaches almost no one. In settings with severe specialist shortages, the decision-relevant question is coverage: how to extend adequate care to the untreated majority, usually through task-sharing that lets trained non-specialists deliver protocolized care under supervision, as in the WHO mhGAP approach. Evaluate the delivery model and its reach, not just the efficacy of the treatment in ideal conditions.

7. Use burden-of-disease evidence to expose funding disparity, and act on it. Compare your system's share of spending on mental health against mental illness's share of the DALY burden; the mismatch is usually stark and is the empirical core of the parity argument. Present the disparity plainly to those who set budgets, and pair it with a costed, deliverable plan to narrow it — a burden statistic without an implementation route persuades no one to move money.

8. Guard against cost-shifting dressed up as saving. Closing beds, shortening admissions, or moving care into the community saves money only if the displaced need is actually met elsewhere; otherwise the cost reappears as crisis presentations, family strain, homelessness, or imprisonment. This is the enduring lesson of deinstitutionalization. Before you book a saving from reduced institutional care, verify that the community capacity to absorb the need exists and is funded, and track where displaced patients actually go.

9. Cost informal and family care, because it is doing much of the work. A large share of care for people with severe mental illness is provided unpaid by families, and ignoring it makes services that shift care onto relatives look cheaper than they are. Value carers' time using a stated, defensible method, and count the effect of illness and caring on carers' own mental health, which is itself a significant and often invisible cost.

10. Integrate mental and physical health economics rather than siloing them. Mental and physical illness are deeply comorbid: depression worsens the outcomes and raises the costs of diabetes and heart disease, and chronic physical illness raises the risk of depression. Evaluating collaborative-care models that treat both together often reveals savings on the physical side that a mental-health-only analysis would miss. Look for the interaction, and give credit for physical-health costs avoided when mental illness is treated.

11. Include the economics of suicide and crisis prudently and respectfully. Suicide is a leading cause of death in young people and carries measurable economic loss, and suicide prevention interventions can be cost-effective. Where you value a death averted, use the same conventions your system applies elsewhere and avoid sensational or speculative figures; describe the effect qualitatively rather than invent a number you cannot defend. Handle this evidence with the seriousness it demands.

Questions to discuss with your team

1. From whose perspective are we judging whether mental health care is worth funding — and are we honest that the answer changes with the standpoint? Mental health is the clearest case in this book where the perspective decides the verdict: a treatment that looks marginal to a health payer can be strongly cost-saving to society once lost work, informal care, and welfare are counted. The tension is institutional — your organization is usually accountable for one budget, and the gains from mental health treatment leak into budgets you do not control, so a "societal" case can feel like an argument to spend your money for someone else's benefit. Discuss which perspective your evaluations currently use by default, and whether that default quietly biases you against mental health. An honest conversation names the specific costs you are ignoring by taking a narrow view, and asks whether a societal analysis, presented alongside the payer view, would change any decision you have recently made. It should also confront the harder question of whether a cross-sector saving can be captured at all, or whether it remains a benefit no one will pay you for.

2. Do our outcome measures let mental health services show what they actually achieve, or are we scoring them with a broken ruler? Most evaluation toolkits were built for conditions that shorten life or impair the body, and they under-detect the improvements a mental health service produces — recovery, function, participation, hope. If your reference case mandates a generic preference-based measure, you may be systematically recording effective services as ineffective and then defunding them for "lack of evidence". Discuss where this has happened in your own portfolio, and whether the weakness lies in the services or in the instrument. The real tension is that the generic measure is comparable across all conditions and trusted by your funder, while a mental-health-specific or wellbeing measure fits better but sits outside the reference case. An honest answer identifies which of your mental health services are most likely being under-read, and decides whether to commission a fairer measurement before the next funding round rather than after.

3. Are we serious about closing the treatment gap, or only about improving care for the minority already in the system? Most people with a mental disorder receive no adequate treatment, and every marginal improvement you make to specialist services may reach only the small share who already have access. The uncomfortable question is whether your investment priorities are set by the visible, articulate, treated population or by the larger, quieter untreated one — and these often point to different spending. Discuss the trade-off between deepening care for current patients and widening coverage through primary-care task-sharing, digital tools, or community workers, which may deliver more health per unit of spend but feels less like "proper" specialist care. The tension is between quality and reach, and between the patients you can see and the ones you cannot. An honest answer sets an explicit coverage ambition, not just a quality one, and is candid about who is currently left out — by geography, income, language, and condition.

4. Are we measuring the productivity loss that mental illness actually causes, or only the part that is easy to count? Days off work are simple to record, but in mental illness the larger loss is presenteeism — people present but impaired, concentrating poorly and working slowly — and if you count only absence you capture the smaller half and understate the case for treatment. The tension is that presenteeism is genuinely harder to measure and its valuation is contested, so the convenient number and the honest number diverge, and the convenient one makes mental health look cheaper to neglect. Discuss which instruments your evaluations use and whether they capture impaired presence at all, or quietly assume a person at their desk is working at full capacity. Consider, too, that the productivity gain from treatment lands on employers and the tax base rather than on your own budget, which can make the whole exercise feel like measuring someone else's benefit. An honest answer commits to a validated instrument that captures both absenteeism and presenteeism, states the valuation method openly, and is clear about how sensitive the business case is to that choice. It also asks whether your organization is willing to act on a productivity argument it can measure but cannot bank.

5. Do our appraisal methods do justice to prevention and early intervention, or do they penalize the long wait for the payoff? Most mental disorders emerge in adolescence and early adulthood, and intervening early — through perinatal care, school-based programmes, or early intervention in psychosis — can bend a lifetime trajectory, but the returns arrive years or decades after the spend. Standard short-horizon return-on-investment tools and heavy discounting understate that back-loaded benefit badly, so the very interventions with the largest lifetime value can look the weakest on a three-year business case. The tension is real and not merely technical: money spent on prevention today competes with visible, urgent demand for treatment now, and the people who benefit from prevention are often not yet identifiable. Discuss whether your appraisal horizon is long enough to see the downstream gain, how your discount rate treats benefits that fall a generation away, and whether short budget cycles are quietly killing your strongest long-run investments. An honest answer extends the time horizon deliberately, applies and displays discounting transparently rather than hiding it, and states plainly that the prevention case is genuine but patient — and then decides whether the organization can hold its nerve for a return that outlasts the current leadership.

6. Does our spending on mental health bear any honest relation to the burden it carries — and if not, what will we actually do about it? In almost every system that has measured it, mental illness commands a share of the disability-adjusted life year burden far larger than its share of the health budget, and "parity of esteem" is easy to affirm in principle and quietly ignore in the allocation. The tension is that closing the gap means moving money from somewhere that already feels stretched, and that a bald parity statistic persuades no one to act without a costed, deliverable plan attached. Discuss where your own system's mental health spend sits against its share of burden, and whether the community and primary-care capacity exists to absorb demand rather than merely shift cost onto families, crisis services, and prisons — the enduring lesson of deinstitutionalization. Be candid about whether past "savings" from closing beds or shortening admissions were real or simply displaced to budgets you do not see. An honest answer pairs the burden-to-spend comparison with a concrete route to narrow it, verifies that any claimed saving is met need rather than moved need, and names the trade-off in plain terms: what you would spend less on to fund parity, and who would notice.

In practice: a health economics example

Selangor Utara — a fictional state health authority in Malaysia — is deciding how to spend a modest expansion in its mental health budget. Depression and anxiety are common across the state, but specialist psychiatrists are concentrated in the two largest cities, and rural districts have almost none; the authority's own estimate is that the great majority of people with common mental disorders receive no treatment. Two options are on the table. The first deepens the existing specialist service by funding more psychiatrists and clinical psychologists in the urban hospitals. The second trains primary-care nurses and medical officers across all districts to screen for and manage depression and anxiety using a structured protocol with specialist supervision — a task-sharing model in the spirit of the WHO Mental Health Gap Action Programme — and adds a stepped-care pathway with telephone follow-up.

The authority's health economist frames the choice around coverage as well as cost-effectiveness. The specialist-deepening option improves care for people who can already reach a city hospital, but her modelling shows it barely moves the treatment gap: the untreated rural majority stays untreated. The task-sharing option reaches far more people per unit of spend, though each episode of care is less intensive and outcomes per patient are somewhat lower than specialist treatment. Measured only as cost per patient treated to recovery in a clinic, the specialist option looks competitive; measured as depression-free days gained across the whole state population, the task-sharing option is well ahead because it reaches so many more people.

She resists two temptations. She does not evaluate the programme on a generic preference-based measure alone, because the state's own pilot showed it under-detected improvement; she uses a validated depression scale as the primary outcome and reports a mapped quality-of-life estimate alongside, clearly labelled. And she takes a societal perspective as well as a health-system one. A local employers' federation has shared anonymized data suggesting heavy productivity loss from untreated depression, dominated by presenteeism rather than absence. Incorporating restored productivity — valued by a stated, conservative method — the task-sharing option shifts from "affordable" to "largely self-funding to the economy over a few years", even though most of that return accrues to employers and the tax base, not to the health authority's own budget.

That distributional point is where the decision gets honest. The finance director notes, correctly, that a saving to employers does not refill the health budget, so the societal case cannot be spent. The economist presents both views without collapsing them: the health-system cost the authority must actually fund, and the wider societal return that justifies asking the state treasury for it. She also flags the cost-shifting risk — that a screening programme which identifies need it cannot then treat would simply move distress from invisible to visible without relieving it — and so ties the roll-out to a firm supervision and supply guarantee, not screening alone.

The authority funds the task-sharing model, phased by district, with the depression scale and coverage rate as its headline metrics rather than cost per QALY, and takes the societal analysis to the state treasury as the basis for a cross-government contribution. The specialist service receives a smaller uplift focused on supervising the new primary-care workforce and handling severe cases the protocol refers up. The economics did not make the decision by itself, but it reframed the question from "which service is more cost-effective per patient" to "how do we buy the most mental health, and the widest coverage, for a population most of whom currently get nothing" — and that reframing changed the answer.

Four sector lenses

Startup

A digital-mental-health start-up — a therapy app, a symptom-tracking tool, a peer-support platform — usually pitches on reach and cost: the promise of narrowing the treatment gap cheaply by delivering support without a clinician for every user. That is a genuine strength, but the evidence bar is where these ventures live or die, because engagement often collapses outside a trial and generic measures may not capture real benefit. A credible start-up measures outcomes with a validated mental-health-specific instrument, reports real-world retention honestly rather than trial-condition efficacy, and is candid about which severities it is and is not safe for. Its strongest economic argument is productivity: an employer or insurer buyer will respond to a defensible presenteeism-reduction case far more than to a soft wellbeing claim (see Chapter 5.2 — Digital Health Economics for the wider evidence-standards picture).

Small business

An established small provider — a group psychology practice, a single community counselling clinic, a modest employee-assistance supplier — competes not on novelty but on a steady reputation and a referral base it cannot afford to damage. Its economic problem is neither the start-up's survival gamble nor the enterprise's siloed accounting, but cash-flow under payer terms it does not set: session caps, reimbursement rates, and reporting demands that fall heavily on a business with no analytics team. Such a practice should measure outcomes with a validated mental-health-specific instrument anyway, because a defensible recovery record is its best protection in contract renewals and its answer when a commissioner questions value. It knows its own treatment gap intimately — the local people it turns away for lack of capacity or ability to pay — and its honest move is to be clear about which severities it can safely hold and which it must refer on, rather than stretching beyond safe limits to fill the diary. Its most credible pitch to an employer or insurer buyer is reliable access and continuity for a defined population, priced sustainably, not a promise of scale it cannot deliver.

Enterprise

A large employer, insurer, or integrated provider sees mental health from both sides: as a payer for treatment and as an organization whose own workforce loses productivity to it. This gives the enterprise an unusual incentive to take the societal view, because the "indirect" costs that a health payer externalizes are, for an employer, its own wage bill and its own absenteeism. The enterprise move is to run mental health as a linked investment — workplace prevention, fast access to treatment, and collaborative care for employees with comorbid physical illness — and to measure the return in retained productivity, not just claims cost. The obstacle is siloed accounting, where the health-benefits budget and the productivity loss sit on different ledgers and no one owns the whole; the mature enterprise brings them together so the business case can be seen.

Government

A ministry or national payer holds the parity question directly, because it sets the budgets in which mental health is under-funded relative to burden and defines the reference case that decides how mental health outcomes are measured. Government's task is to align spending closer to the DALY burden, to fund the community and primary-care capacity that makes deinstitutionalization and task-sharing work rather than merely shift cost, and to build the cross-sector mechanisms that let a health investment be repaid from the justice or welfare savings it creates. It also carries the treatment-gap responsibility that no single provider can, deciding coverage ambitions for whole populations, including the rural, poor, and marginalized who are furthest from care. The honest governmental position treats mental health as an investment in human capital and social participation, funds it accordingly, and holds itself to a coverage target, not only a quality one.

Common failure modes

• Judging mental health on a health-system perspective alone. Counting only treatment costs and ignoring the larger productivity and informal-care costs, so effective treatment looks unaffordable. Fix: use a societal perspective, or present it alongside the payer view, and make the indirect costs explicit.
• Measuring with a blunt instrument. Using a generic preference-based measure that cannot detect psychological improvement, then recording an effective service as ineffective. Fix: use a condition-specific or mental-health-specific outcome, and consider capability or wellbeing measures (Chapter 3.5 — Capabilities).
• Counting absence, missing presenteeism. Capturing only days off work and so under-stating the productivity loss, which in mental illness is dominated by impaired presence. Fix: use an instrument that measures both, and value the time transparently.
• Booking cost-shifting as saving. Closing beds or shortening admissions and recording a saving while the need reappears in crisis care, families, or prisons. Fix: verify and fund the community capacity that absorbs the need before claiming the saving; track where patients go.
• Short-horizon evaluation of prevention. Applying a standard return-on-investment tool to early-intervention or prevention programmes whose benefits arrive years later, and so understating them. Fix: extend the time horizon, discount transparently, and flag the lag.
• Optimizing for the treated minority. Improving specialist care for the few already in the system while the untreated majority stays untreated. Fix: set an explicit coverage ambition and evaluate reach, not just per-patient cost-effectiveness.
• Ignoring comorbidity. Evaluating mental health in isolation and missing the physical-health costs that treating it avoids. Fix: evaluate collaborative-care models and credit avoided physical-health costs.

Maturity model

Checklist

• State the perspective explicitly and present a societal view, or payer and societal side by side.
• Measure both absenteeism and presenteeism, and value the time by a stated method.
• Choose an outcome measure that can detect mental-health change; do not rely on a blunt generic measure alone.
• Consider a capability or wellbeing measure where recovery and participation are the goal (Chapter 3.5 — Capabilities).
• Map cross-sector costs (welfare, justice, housing, education) and identify who captures any saving.
• Use a long enough time horizon for prevention and early-intervention cases; discount transparently.
• Frame the decision around coverage and the treatment gap, not only per-patient cost-effectiveness.
• Verify and fund community capacity before booking any saving from reduced institutional care.
• Cost informal and family care, including the effect on carers' own mental health.
• Compare mental health's share of spend against its share of DALY burden, and pair the gap with a costed plan.
• Look for comorbidity and credit physical-health costs avoided by treating mental illness.
• Handle suicide and crisis evidence seriously; describe effects qualitatively rather than invent figures.

Key sources

• WHO Mental Health Gap Action Programme (mhGAP) — World Health Organization guidance on scaling up care for mental, neurological, and substance-use disorders through non-specialist providers.
• WHO — mental health financing, atlas, and burden data — WHO Mental Health Atlas and burden-of-disease evidence on the funding-to-burden mismatch and the treatment gap.
• NHS Talking Therapies / Improving Access to Psychological Therapies (England) — a large-scale programme partly justified on the economics of restored employment; associated evaluations and the economic case set out by the London School of Economics (Layard and colleagues).
• Institute for Health Metrics and Evaluation — Global Burden of Disease — DALY estimates for mental disorders used in parity arguments.
• OECD — mental health and work; Health at a Glance — international evidence on the productivity and employment costs of mental ill-health.
• Economics Network — Health Economics for Teachers — https://economicsnetwork.ac.uk/health/teachers

References

1. Mental disorder — Wikipedia — https://en.wikipedia.org/wiki/Mental_disorder
2. Major depressive disorder — Wikipedia — https://en.wikipedia.org/wiki/Major_depressive_disorder
3. Anxiety disorder — Wikipedia — https://en.wikipedia.org/wiki/Anxiety_disorder
4. Quality-adjusted life year — Wikipedia — https://en.wikipedia.org/wiki/Quality-adjusted_life_year
5. Presenteeism — Wikipedia — https://en.wikipedia.org/wiki/Presenteeism
6. Mental health parity — Wikipedia — https://en.wikipedia.org/wiki/Mental_health_parity
7. Disability-adjusted life year — Wikipedia — https://en.wikipedia.org/wiki/Disability-adjusted_life_year
8. Global mental health — Wikipedia — https://en.wikipedia.org/wiki/Global_mental_health
9. Improving Access to Psychological Therapies — Wikipedia — https://en.wikipedia.org/wiki/Improving_Access_to_Psychological_Therapies
10. Deinstitutionalisation — Wikipedia — https://en.wikipedia.org/wiki/Deinstitutionalisation
11. Early intervention in psychosis — Wikipedia — https://en.wikipedia.org/wiki/Early_intervention_in_psychosis
12. Suicide prevention — Wikipedia — https://en.wikipedia.org/wiki/Suicide_prevention
13. mhGAP: Mental Health Gap Action Programme — World Health Organization — https://www.who.int/teams/mental-health-and-substance-use/treatment-care/mental-health-gap-action-programme
14. Mental Health Atlas — World Health Organization — https://www.who.int/publications/i/item/9789240036703
15. Health at a Glance — OECD — https://www.oecd.org/en/publications/health-at-a-glance-19991312.html
16. Economics Network — Health Economics for Teachers — https://economicsnetwork.ac.uk/health/teachers