21 September 2026 · England · Evidence and interpretation
Imagine someone arriving at an appointment with a new physical symptom and an established psychiatric diagnosis. The diagnosis may help a clinician understand their circumstances. It may also become an explanation too quickly. This is an illustrative situation, but it poses a concrete question: what ensures that the new symptom receives the attention it needs?
Reporting on 20 September renewed attention to early deaths among people using specialist mental-health services. The news report is a starting point; the underlying statistics deserve a careful reading.
What the figures describe
An NHS England release dated 2 September 2026 covers deaths in 2022–2024 among adults aged 18–74. It reports an excess mortality measure of 379%: about 4.8 times the mortality of the comparison population. Cancer, cardiovascular disease, liver disease and respiratory disease together accounted for 56.5% of deaths in the group. These are population comparisons, not a forecast of what will happen to an individual.
The category needs explaining. For this indicator, contact with secondary mental-health services means an open referral in the five years before death. It is an administrative definition rather than a single diagnosis or a measure of everyone experiencing distress. The release is labelled official statistics in development and warns about data-quality issues in provider submissions, particularly when interpreting local changes. A mortality comparison cannot, on its own, establish which factors caused the difference. Those qualifications affect how the figures should be used; they do not remove the need to investigate the inequality.
Following the symptom beyond the appointment
Rethink Mental Illness’s Closing the Mortality Gap summary identifies interacting problems: poverty, stigma, social exclusion and fragmented care alongside physical-health risks. It describes “diagnostic overshadowing”, in which physical symptoms are wrongly attributed to mental illness. Its proposals include better integration of care, workforce training, funded physical-health interventions and involvement of people with lived experience. These are the charity’s policy recommendations; the mortality figures do not independently demonstrate the effectiveness of every proposed intervention.
A useful question for services is what happens after a concern is voiced. Does it become a recorded symptom? Does the record produce an investigation? Is somebody responsible for following the result? Can the person question an explanation that does not fit their experience? These questions concern the route through care, including the moments when responsibility passes between teams.
This is where an autopoietic ecology reading can help organise an inquiry. AE asks how an activity becomes possible and how it changes the conditions for what follows. Here, an appointment depends on time, clinical judgement, trust and usable records. Its value also depends on whether those conditions support the next necessary action. AE supplies questions for investigation; it supplies neither a diagnosis nor a tested treatment.
Consider two hypothetical services reporting the same number of completed health checks. In one, an abnormal finding leads to accessible follow-up and someone checks that it happened. In the other, a referral is issued but nobody owns the next step. Counting checks alone would miss that difference. A more informative evaluation would follow the chain, including the work required of patients and carers.
What would count as progress?
The alternative explanation must remain visible: differences in underlying illness and circumstances can contribute to mortality differences even where clinicians provide good care. It would be unfair to infer individual negligence from an aggregate indicator. Equally, that uncertainty should prompt better investigation of avoidable barriers, rather than treating the whole gap as inevitable.
A practical research programme would examine follow-up completion, delays, patients’ ability to correct records and outcomes over time. It would ask whether improvements reach people facing the greatest obstacles. Taking physical symptoms seriously becomes meaningful when a person can see a concern carried through to an appropriate response.
Subscribe free for published EdgeLab posts.
Leave a Reply