The ADHD Annual Review and Physical Health Monitoring: What to Check and Who Checks It

Two audits, both of adults on stimulants, both measured against NICE guideline NG87, both published in BJPsych Open. One found that not a single patient in the sample had blood pressure, pulse or weight documented at six months. The other found blood pressure recorded in the last six months for every patient in its sample.

The difference was not the guideline. It was who was doing the measuring, and whether a system existed that made the measurement happen. That gap, between the monitoring requirement as written and as recorded - is what this page is about.

What NICE Requires, and How Often

The requirement itself takes four minutes to read. Recommendation 1.8.9 asks for heart rate and blood pressure before and after every dose change and every six months, compared with the normal range for age. Recommendation 1.8.5 adds six-monthly weight in adults.

Three words in 1.8.9 are routinely dropped. "Before and after", a post-increase reading with no comparator tells you the patient's blood pressure, not the effect of the change. "Compare with the normal range" - the recommendation asks for interpretation, not capture. And "every six months" is a floor, not a target: 1.7.28 requires more frequent monitoring where neurodevelopmental, mental health or physical health conditions coexist, which describes much of an adult caseload.

What the Baseline Has to Contain

Before the first prescription, recommendation 1.7.4 requires height and weight, pulse and blood pressure with an appropriately sized cuff, a medical history and current medication, a substance misuse and diversion risk assessment, and a cardiovascular assessment.

Two go missing with predictable regularity, and both require a clinician to do something beyond reading a number off a machine.

• The cardiovascular assessment. Not the same as recording pulse and blood pressure; the recommendation lists them separately. It is the history and examination that establishes whether any recommendation 1.7.5 feature is present
• The substance misuse and diversion risk assessment. A named component of 1.7.4, concerning a controlled drug, and the component most likely to be absent when a prescribing decision is later scrutinised

A 2023 audit of an adult community mental health team in Betsi Cadwaladr found baseline blood pressure in 57% of notes, baseline pulse in 39%, baseline weight in 51%, and a documented cardiovascular assessment in 6%. The pattern is consistent: the more clinical judgement an element needs, the less often it is recorded.

The Features That Trigger a Cardiology Opinion

Recommendation 1.7.5 lists what the cardiovascular assessment is looking for. Any one of these requires a cardiology opinion before starting.

• Congenital heart disease or previous cardiac surgery
• Sudden death in a first-degree relative under 40 suggesting cardiac disease
• Shortness of breath on exertion compared with peers, or fainting on exertion or in response to fright
• Rapid, regular palpitations of sudden onset and offset, or chest pain suggesting cardiac origin
• Signs of heart failure, a murmur on examination, or blood pressure classified as hypertensive

Put those on the proforma as fields rather than as free text. That single change is the highest-yield intervention available to most services, and it is the difference between a 6% recording rate and a defensible one.

Is an ECG Needed?

Not routinely. Recommendation 1.7.4 states plainly that an ECG is not needed before starting stimulants, atomoxetine or guanfacine unless a 1.7.5 feature is present, or a coexisting condition is being treated with a medicine that may pose increased cardiac risk. Recommendation 1.8.10 extends that into treatment and adds routine blood tests, including liver function tests, to the list.

So the question a protocol answers is not whether you do ECGs but what triggers one and who decides. The coexisting-medicine clause is the one most often overlooked: a patient on an antipsychotic with recognised QT effects meets it even with an unremarkable cardiac history.

Who Does What Once the GP Is Prescribing

Recommendation 1.7.29 moves prescribing and monitoring to shared care after stabilisation. That sentence allocates the task but not the detail, and the detail is where monitoring is lost. Working protocols allocate explicitly, and yours should answer three questions in writing.

• Who records the six-monthly observation, and what recalls the patient? An obligation with no recall mechanism is a hope. A 2022 closed-loop audit in NHS Greater Glasgow and Clyde improved blood pressure recording from 19% to 50% between cycles by intervening on process, not on knowledge
• Who performs the annual medication review, and do they have the training and expertise 1.10.1 requires? The six-monthly observation and the annual review are different events with different competence requirements
• What returns a patient to the specialist, and how fast? Name the parameters, the thresholds and the contact route in the document itself

One warning when you audit. Several shared care protocols in circulation predate the guideline they implement - the Oxford Health adult protocol was approved in 2017 and still cites CG72 from 2008. Check what document local GPs are working from before concluding the guidance was ignored.

What the Annual Review Has to Cover

The annual review is a medication review, not an observations check. Recommendation 1.10.1 requires a professional with training and expertise in managing ADHD to review medication at least once a year and discuss whether it should continue, covering:

• The person's own preference, and any wish to stop or change
• Benefits, including how well treatment works across the whole day rather than at peak
• Adverse effects, and whether the dose has actually been optimised
• Impact on education and employment, and the effect of missed doses and planned breaks
• The need for non-medication support where symptoms still impair despite optimisation

Recommendation 1.10.3 goes further and asks you to consider trial periods of stopping or reducing. A review that records observations, reissues the prescription and never seriously entertains "no" is documentation rather than review.

Running the Review at Scale

A 2025 service evaluation at Leeds and York Partnership NHS Foundation Trust replaced routine face-to-face annual reviews with a form-based first pass. Of 288 patients contacted, 262 responded (91%), and 60 (20%) were identified as needing a follow-up review.

That is a defensible way to run annual reviews at volume, provided the form covers every element of 1.10.1 and the fifth of patients who need a conversation actually get one. It is a way to triage the review, not to avoid it.

Frequently Asked Questions

How often should weight be measured in adults on ADHD medication?

Every six months, under recommendation 1.8.5. Recommendation 1.8.8 asks you to consider monitoring BMI where weight has changed as a result of treatment, and to consider changing medication if the change persists.

What should happen if a reading is out of range?

Recommendation 1.8.11 covers it: sustained resting tachycardia above 120 bpm, arrhythmia, or a clinically significant rise in systolic blood pressure on two occasions means reducing the dose and referring to an adult physician. NICE does not quantify "clinically significant"; local protocols do, and they disagree, so cite where your figure came from.

Conclusion

If you want to know where your service stands, pull twenty sets of notes and score them. The published audits suggest you will find blood pressure in better shape than pulse, pulse better than weight, and the cardiovascular assessment worse than all three.

Global ADHD Network's ADHD prescribing and management course works through review structure and the competence requirement in 1.10.1 in more depth. Our team is at info@globaladhdnetwork.com.

‍