> ## Content Index
> Fetch the complete content index at: https://www.healthdatacon.io/llms.txt
> Use this file to discover other available public pages before exploring further.

# NHS Data Reveal Fivefold Mortality Gap After Mental Health Referral
- URL: https://www.healthdatacon.io/nhs-data-fivefold-mortality-gap-mental-health-referral/
- Published: 2026-09-23T13:00:00.000Z
- Updated: 2026-09-23T13:00:00.000Z
- Description: New NHS data show adults referred to secondary mental health services were 4.8 times as likely to die before 75, exposing gaps between health checks, follow-up care and measurable physical-health outcomes.
- Author: Kenneth R. Deans Jr.
- Tags: EMEA

England’s latest linked health records put a precise number on a long-recognized failure: adults aged 18 to 74 who had been referred to secondary mental health services were 4.8 times as likely to die before 75 as other adults. Across 2022 through 2024, 129,216 people in that group died—an annual average of just over 43,000, according to the new [NHS data](https://digital.nhs.uk/data-and-information/publications/statistical/excess-under-75-mortality-rates-in-adults-with-serious-mental-illness/2022-to-2024?ref=healthdatacon.io).

The release is now driving a wider challenge to England’s health system. A new report from Rethink Mental Illness argues that the mortality gap persists partly because physical and mental health services remain poorly connected. Its central point is not that one program or one missed appointment explains the disparity. It is that routine information about risk, screening and follow-up is still too fragmented to show whether people receive effective preventive care after a mental health referral.

The distinction matters. The NHS figures are population statistics, not a clinical trial, and they do not prove why any individual died. They do, however, establish the scale and pattern of excess mortality with unusual clarity. For health leaders, the data turn an ethical problem into a measurable informatics problem: who is being identified, what care follows, and whether local systems can connect mental health records to physical health outcomes without losing patients between services.

## What the NHS measure captures

The official measure defines the mental health cohort as people with an open referral to secondary mental health services during the five years before death. That is a practical data proxy for a “severe mental health problem,” not a diagnosis-based registry. Some people had only a referral; many others had community, outpatient or inpatient contact. People supported only through primary care, talking therapies or voluntary services may not appear in the same cohort.

NHS England linked the Mental Health Services Dataset with mortality records and compared directly standardized death rates, adjusting for differences in age structure. The results combine three years of deaths and populations, smoothing some annual volatility. The companion [methods report](https://www.gov.uk/government/statistics/suicide-in-people-with-severe-mental-health-problems/suicide-in-people-with-severe-mental-health-problems-report?ref=healthdatacon.io) explains that denominators were estimated from mental health records and Office for National Statistics population estimates.

Those choices strengthen comparability but introduce important limits. The publication is labeled “official statistics in development.” Coverage of the Mental Health Services Dataset has expanded as more providers submit data, and referral patterns and service availability can change who enters the measured population. Local data-quality problems can also create apparent shifts. The 4.8-fold figure therefore describes a broad service-contact population and should not be read as a causal effect of referral, treatment or any particular diagnosis.

## Physical disease accounts for most deaths

The leading signal is not suicide, though suicide risk is exceptionally high. Cancer, cardiovascular disease, liver disease and respiratory disease together accounted for 56.5% of deaths in the referral cohort. Suicide accounted for 6.9%. Compared with adults outside the cohort, mortality was 2.3 times higher for cancer, 3.8 times higher for cardiovascular disease, 6.1 times higher for respiratory disease and 6.4 times higher for liver disease. The suicide rate ratio was 14.6.

That pattern supports a systems response broader than specialist psychiatric care. The government’s [inequalities briefing](https://www.gov.uk/government/publications/severe-mental-illness-smi-physical-health-inequalities/severe-mental-illness-and-physical-health-inequalities-briefing?ref=healthdatacon.io) identifies overlapping pathways: higher rates of smoking and some long-term conditions, medication-related metabolic effects, poverty and unstable housing, and barriers to screening or treatment. It also warns against assuming that deprivation alone explains the difference. Clinical symptoms can mask physical illness, while physical complaints may be misattributed to mental illness—a problem often described as diagnostic overshadowing.

Rethink’s [new report](https://rethink.org/media/s0uht1y5/closing-the-mortality-gap.pdf?ref=healthdatacon.io), “Closing the Mortality Gap,” estimates that people affected by severe mental illness die 15 to 20 years earlier on average. It cites smoking as a major modifiable risk and calls for prevention services designed around the realities of medication, symptoms and access. The report is an advocacy synthesis, not an experimental evaluation. Its value is in combining national statistics with service experience; its recommendations should still be tested against outcomes.

## A health check is not an outcome

England already asks general practices to provide annual checks covering alcohol use, blood glucose, blood pressure, body mass index, lipids and smoking status for people on severe mental illness registers. The [national collection](https://digital.nhs.uk/data-and-information/publications/statistical/physical-health-checks-for-people-with-severe-mental-illness?ref=healthdatacon.io) tracks delivery quarterly. Rethink reports that 58.6% of eligible people received all elements of the check in the first quarter of 2026/27.

That percentage is useful operationally, but it answers only whether measurements were recorded. It does not show whether a high blood pressure reading led to treatment, whether a smoker received effective support, whether medication was reviewed, or whether a cancer referral was completed. Rethink says the national system does not routinely monitor the interventions that follow the checks or their effect. This is the data gap at the center of the report: process is visible, continuity and outcomes are not.

A 2024 [peer-reviewed study](https://pmc.ncbi.nlm.nih.gov/articles/PMC11350427/?ref=healthdatacon.io) of physical health checks in deprived communities reached a similar practical concern. Records can show that checks occurred while providing a less complete view of follow-up care and patient experience. The study’s setting and observational design limit generalization, but it reinforces why commissioners should not treat check completion as proof that cardiovascular, metabolic or respiratory risk has been reduced.

## The gap is widest in working-age adults

The relative disparity was greatest among people aged 30 to 49\. Within those age bands, adults in the referral cohort were 6.5 to 6.6 times as likely to die as peers outside it. Rate ratios can be especially large where baseline mortality is lower, so they do not by themselves show the largest number of deaths. Still, the result signals preventable loss during years when chronic disease identification and sustained primary care could have decades of benefit.

The new suicide analysis adds another layer. It found an annual average of 2,967 registered suicides in the cohort during 2022 through 2024, close to 60% of suicides among all 18- to 74-year-olds in England. About 80% of those who died by suicide had an open referral in the prior year, but roughly 14% of that subgroup had no recorded contact or hospital stay. The figures cannot explain individual circumstances, yet they give local systems a baseline for examining referral-to-contact delays and service transitions.

Trend interpretation requires care. The measured mental health population grew by 45.6% from 2015–17 to 2022–24, reflecting some combination of need, demand, referral practice, service availability and improved dataset coverage. Suicide registration was also affected by a 2018 change in the legal standard of proof and by longer post-pandemic registration delays. An observed increase in recorded deaths is therefore not automatically evidence that clinical care deteriorated.

## From linked records to accountable care

Rethink recommends integrated physical and mental health care, better workforce training, stronger prevention, peer support and co-designed services. The World Health Organization’s updated [global guidance](https://www.who.int/news-room/fact-sheets/detail/mental-health-strengthening-our-response?ref=healthdatacon.io) points in the same direction, noting that people with severe mental health conditions often die 10 to 20 years early and urging community-based services that connect mental and general health care.

For England, the immediate technical task is to move beyond counting completed checks. Integrated care systems need measures that follow the sequence from risk identification to intervention, referral completion and clinical outcome. Those measures should be stratified by age, sex, ethnicity, deprivation and geography, with safeguards against using data to penalize services caring for higher-risk populations. They should also preserve the distinction between service contact and diagnosis that the current NHS indicator necessarily blurs.

The evidence does not establish that any single integration model will close the mortality gap, and it does not justify blaming patients or clinicians for outcomes produced across many institutions. It does show that physical disease dominates premature mortality after mental health referral and that current activity metrics reveal too little about what happens next. The next phase is not merely another dashboard. It is an accountable chain of care in which each abnormal finding, missed contact and completed referral can be followed to a meaningful health result.