People Without a Psychiatric Diagnosis Are Also at High Risk: What a Mortality Study Has Shown

Introduction

When excess mortality in mental healthcare is discussed, attention usually centers on people with named psychiatric disorders. That is understandable. Diagnoses organize services, research, and risk thinking. But a new BJPsych study shows that this map is incomplete. In South London mental healthcare records, nearly a third of service users had no recorded diagnosis at all, yet their mortality remained substantially worse than that of the surrounding population.

The paper’s importance lies in that simple but uncomfortable point: risk does not begin only where diagnostic certainty begins.

This is not only a clinical issue, but also a systems issue. If people can receive mental healthcare, remain outside clear diagnostic categories, and still have markedly reduced life expectancy, then diagnosis-centered care pathways may be missing a large vulnerable group.

Who Are Patients “Without a Diagnosis”?

The phrase “without a diagnosis” can easily be misunderstood. It does not mean these people had no contact with services, no distress, or no reason to be seen. In this study, all included individuals had past or current contact with mental health services. The distinction was in how that contact appeared in the record.

The authors compared three groups: those with ICD-10 F-code mental disorder diagnoses, those with Z-code diagnoses reflecting unspecified or contextual problems, and those with no diagnosis recorded.

That third group matters precisely because it is easy to ignore. In the combined annual cohorts, 29.3% of people accessing mental health services had no diagnosis recorded. That is not a marginal category. It is a substantial segment of the service-using population.

The Z-code group is also revealing. Most Z-codes in this study were coded as “person with feared complaint in whom no diagnosis can be made,” followed by general psychiatric examination not elsewhere classified. That suggests that at least some of the people outside standard F-code diagnoses were still presenting with concerns serious enough to bring them into mental healthcare.

What The Study Actually Examined

The study used routinely collected electronic health records from the South London and Maudsley NHS Foundation Trust, linked to mortality data, to examine annual cohorts of people with past or current mental health service receipt from 2015 to 2024. Mortality and life expectancy were then compared across the three diagnostic-status groups against the local catchment population.

Across the combined annual cohorts, there were 3,266,268 observations. Of these, 57.7% had an F-code diagnosis, 13.0% had a Z-code diagnosis, and 29.3% had no diagnosis.

This design matters because the study drew on routine service data, which makes it especially relevant for public mental health and system planning.

What The Mortality Findings Showed

The results show a gradient, but not a reassuring one. Mortality was worst among those with ICD-10 F-code diagnoses, intermediate among those with Z-codes, and still clearly elevated among those with no diagnosis. Annual standardized mortality ratios ranged from 2.25 to 2.56 in the F-code group, 1.88 to 2.18 in the Z-code group, and 1.59 to 1.87 in the no-diagnosis group.

In plain terms, even the people without a recorded diagnosis were dying at markedly higher rates than the general population.

Years of life lost were highest in the F-code group, but remained substantial in the others. For people with no diagnosis, estimated years of life lost were 9.4 years for females and 10.6 years for males.

The study also found raised mortality from both natural and external causes across all groups.

Why People Without A Diagnosis May Fall Outside Care Pathways

The paper itself offers the key phrase: these individuals may “fall outside care pathways.” That is more than a documentation problem. Many systems are built around recognized disorders. Referral criteria, treatment protocols, audit measures, and even research inclusion often assume that named diagnoses are the main way to define who needs follow-up.

People without one can therefore become less visible to structured support, even when they have already come into contact with mental healthcare.

Why This Matters Beyond Doctors

If almost one-third of mental health service users in these cohorts had no diagnosis recorded, and that group still showed markedly elevated mortality and reduced life expectancy, then service design that revolves too narrowly around diagnosis is likely missing an at-risk population.

That has implications for planning, commissioning, and public health strategy. It affects how outcomes are monitored, how thresholds for continued support are set, and how “need” is recognized in populations that do not fit cleanly into named disorder categories.

Conclusion

This study does not argue that diagnosis is unimportant. It shows something more specific: diagnosis alone is an incomplete map of risk. People receiving mental healthcare without a recorded psychiatric diagnosis still had substantially elevated mortality and significantly reduced life expectancy compared with the surrounding population.

That makes them important not only for doctors, but for the structure of mental health systems themselves. If these individuals remain outside standard categories, they may also remain outside standard care pathways. The paper’s real contribution is to make that invisibility harder to ignore.

References

Huang, H., Stewart, R., Dregan, A., Choi, K. W., Shetty, H., Singh, I., Schofield, P., & Hayes, J. F. (2026). Mortality and life expectancy in people receiving mental healthcare without a diagnosis: South London electronic health records linkage study. The British Journal of Psychiatry. Advance online publication. https://www.cambridge.org/core/journals/the-british-journal-of-psychiatry/article/mortality-and-life-expectancy-in-people-receiving-mental-healthcare-without-a-diagnosis-south-london-electronic-health-records-linkage-study/F2F98CDE26728342AF70681739BEB067