Every guideline says people with severe mental illness need annual physical health monitoring, and the audits keep showing it does not happen. The reason is rarely disagreement. It is that the task sits in the gap between psychiatry, which sees the patient regularly but regards physical health as someone else's remit, and primary care, which owns the remit but often does not see the patient at all.
The fix is ownership. Decide, for each patient on your caseload, who is doing the physical check and write that name in the notes. If the answer is 'their GP', confirm the patient actually has one and attends. If they do not, the answer is you. Then attach the check to a contact that already happens — the depot clinic, the medication review, the follow-up appointment — rather than creating a separate appointment that becomes another thing to not attend.
The minimum set is short and mostly free: weight and waist circumference, blood pressure, pulse, smoking status, a fasting glucose or HbA1c, and a lipid profile. Add an ECG where the drug or the patient warrants it. Do it at initiation, at three months after any change, and annually. The reason to fix the ownership problem specifically, rather than the knowledge problem, is that today's cohort data show the divergence in physical health begins in the early twenties — during the years when the patient is most engaged with psychiatric services and least engaged with anyone else.
- Name the clinician responsible for physical monitoring in the notes for every patient on the caseload.
- If the patient has no functioning primary care relationship, the psychiatric service owns it by default.
- Attach the check to an existing contact — depot clinic or medication review — not a new appointment.
- Minimum set: weight, waist, blood pressure, pulse, smoking status, HbA1c or fasting glucose, lipids.
- Do it at initiation, three months after any medication change, and annually thereafter.
The statistics, in plain English
The evidence that monitoring itself improves outcomes is thinner than the evidence that the excess physical morbidity exists — screening studies show detection improves with structured programmes, while the downstream mortality benefit is inferred rather than demonstrated. That inference is reasonable: the conditions being detected are treatable and treating them works in the general population. But it means the case for monitoring rests on a chain of reasoning, and the weakest link is not detection, it is whether anything is done about what is found.
Read the rest in the app
You have read your two free briefings this month. The app carries all 27 specialties, every morning, free — and this finding is waiting in it.
Scan to keep reading on your phone. No account needed to start.
Tomorrow morning, before your first patient
One edition a day for psychiatry, written by the desk, every claim tied to its paper. Six minutes.
Get the app — free