Your doctor
might not be
a doctor.
In the NHS today, you may be seen by a Physician Associate or Anaesthesia Associate — healthcare professionals with different training, scope, and accountability to a qualified doctor. Most patients don't know this.
- PAs planned by 2036
- 10,000
- of doctors raised safety concerns
- 87%
- patient deaths recorded by coroners
- 6
- of patients knew the difference
- 52%
What's happening
Rolled out at scale before the evidence existed
The NHS Long Term Workforce Plan, published in 2023, proposed expanding Physician Associate numbers to 10,000 by 2036 — roughly triple the current workforce — while committing to double medical school places. The plan described PAs as "senior clinical decision makers in general practice." That characterisation has since been directly contradicted by the government's own review.
England has 3.2 doctors per 1,000 population, against an EU average of 3.9 and Germany's 4.5. The shortage is real. But the response — deploying thousands of PAs and Anaesthesia Associates as doctor substitutes, at speed, without a nationally agreed scope of practice, without prescribing authority, and without adequate supervision infrastructure — was made before adequate safety evidence existed.
NHS spending on the Additional Roles Reimbursement Scheme, which until late 2024 funded PA salaries in full but did not fund GP posts, grew from £110 million in 2019/20 to over £1 billion by 2022/23. This created a structural incentive to hire PAs over GPs regardless of clinical need — in effect, making PAs free to practices that would otherwise have employed doctors.
No compelling evidence to conclude that PAs were either categorically safe or unsafe.
The evidence
Absence of evidence is not evidence of safety
The Greenhalgh and McKee rapid systematic review, published in the BMJ in March 2025, identified 52 UK studies — only 29 of which met their threshold for trustworthiness and relevance. Of those, no studies examined patient safety incidents as a primary outcome. No studies directly assessed the safety of Anaesthesia Associates. Only one study — involving four PAs — assessed clinical competence through direct observation.
The review's core finding was an epistemological one: the absence of recorded safety incidents cannot be interpreted as evidence that deployment is safe. PAs appeared to struggle most in primary care settings, where the case mix is more diverse, decisions are more uncertain, and supervision is harder to enforce. Preliminary evidence suggested PAs could support ward-based teams in lower-risk, closely supervised environments — but this is a different thing entirely from acting as first-contact clinicians for undifferentiated presentations.
Diagnosis-related allegations constitute 52.8% of PA malpractice claims in the United States, compared with 31.9% for physicians. US PA training requires a bachelor's degree, 3,000 hours of prior clinical experience, and 27 months of graduate study — significantly more than the UK's two-year postgraduate MSc. US PAs have held prescribing authority in all 50 states since 2007. The NHS deployed a less-prepared workforce into more autonomous roles, with less regulatory infrastructure.
The cases
Not isolated incidents — a pattern
Emily Chesterton, aged 30, attended her GP practice twice in November 2022 with calf pain and shortness of breath. She was seen both times by the same PA, who diagnosed muscle sprain and "long COVID." She died of a pulmonary embolism days later. Text messages confirmed she believed she was seeing a doctor. The coroner concluded she should have been immediately referred to hospital, and that "the likelihood is that she would have been treated and would have survived."
Pamela Marking, aged 77, presented to East Surrey Hospital's emergency department in February 2024 with vomiting blood-stained fluid and abdominal tenderness. The PA diagnosed a nosebleed and discharged her without medical review. She had a strangulated femoral hernia. She died two days later following emergency surgery. The coroner raised seven formal concerns, recording that the PA title was "misleading to the public" and that PAs were "working outside their capabilities."
Six patient deaths have been linked to PA contact in coroners' reports in England. The same themes recur across Prevention of Future Deaths reports: no national scope of practice, inadequate supervision, misleading job titles, gaps in competency assessment, failure of escalation. Only around 1.5% of inquests result in such reports — each published finding likely represents a substantially larger number of uninvestigated cases.
Supervision gap
Freedom of Information requests found PAs had acted beyond their legal scope in one in eight NHS trusts — attempting to prescribe medication and commission ionising radiation scans. The BMA's testimony portal collected over 600 accounts of doctors being asked to retrospectively sign off prescriptions for patients they had never assessed. A GP in the Chesterton case told the inquest she "had hundreds of prescriptions to sign off on a Monday afternoon and unfortunately this one slipped through the net."
July 2025
The Leng Review: what changed
Following sustained pressure from the BMA, the RCGP, royal colleges, patient groups and two judicial reviews, Health Secretary Wes Streeting commissioned an independent review in November 2024. Professor Gillian Leng's report — drawing on a survey of 8,558 NHS staff, patient focus groups, clinical interviews and site visits — was published in July 2025. The government accepted all 18 recommendations.
Rename to "physician assistant" — the title "physician associate" was found to be misleading to patients, who regularly confused PAs with doctors. Anaesthesia associates become "physician assistants in anaesthesia."
No undifferentiated patients — PAs must not see patients presenting with undiagnosed symptoms unless working within clearly defined national clinical protocols. This directly addresses the settings where coroners found the greatest risk.
Two years in secondary care first — newly qualified PAs must gain at least two years' experience in supervised hospital settings before working in general practice or mental health trusts, where autonomy is greater and clinical complexity harder to predict.
Named senior doctor supervisor — every PA must have a named senior doctor as formal line manager, with clear responsibility for oversight. The review found that in practice, supervision requirements were widely unmet and poorly understood.
The BMA has criticised the review for stopping short of a nationally mandated scope of practice, calling it insufficient to protect patients from local variation. Implementation faces legal challenge from the Union for Medical Associate Professionals. NHS England has been slow to update primary care specifications. Whether these recommendations are enforced remains an open question.
Your right
You can always ask: "Are you a doctor?"
Only 52% of people in a 2024 Healthwatch England poll understood the difference between a PA and a doctor. The Leng Review found that even after PAs introduced themselves, patients often remained unsure who they had seen. This campaign exists to close that gap.
You have the right to know who is treating you, to ask for a qualified doctor, and to understand what that means. Asking is not rude, unreasonable, or an insult to the person in front of you. It is a basic and legitimate expectation in a healthcare system that asks for your trust.
Spread the word
Help patients know their rights
Share this page with someone who uses the NHS.