As a young GP, Dr Heather Ryan misdiagnosed a patient’s abdominal pain and they later died. Here, she outlines why she would not like to see a physician associate assessing undifferentiated illnesses
All doctors have a handful of cases that haunt them, and Winnie* is one of mine.
I had met the affable 62-year-old a handful of times in my first year as a GP partner, primarily to manage her COPD and its occasional infectious exacerbations.
One day, she was booked into one of my same-day slots with a history of acute-onset abdominal pain and vomiting after eating a takeaway the day before. She looked well and her physical observations were normal. I examined her – her stomach was more tender than I expected, but I rationalised it away – and concluded that she had acute gastroenteritis. I trotted out the usual script about seeking review if her symptoms got worse, and then made a mental note to avoid the offending restaurant.
The following week, I received a discharge summary from the local hospital. I learned that the day after I had seen Winnie, she had taken a turn for the worse. She had started vomiting blood and had been driven to A&E by her son. She died of a perforated peptic ulcer later that night.
Last week, a Tweet from Norfolk & Waveney ICS aimed at members of the public suggested that patients with abdominal pain might see a physician associate at their GP surgery. As my experience shows, abdominal pain is one of the more challenging symptoms to safely assess in primary care, with no point-of-care tests and without the luxury of observing the patient for several hours.
If I struggled after 13 years of training – six years at medical school, six years of postgraduate training, and a year as a qualified GP – then I worry that a physician associate, whose qualification involves just two years of postgraduate study, has the potential to be downright dangerous in this situation.
I can see that physician associates may have some utility in a hospital setting – for example, undertaking administrative tasks like discharge letters, which have limited educational value for doctors. In some circumstances, they may help with service provision by undertaking certain procedures – though care must be taken to ensure that doctors in training do not lose out on learning opportunities. But it is unclear how to use physician associates in general practice efficiently and safely.
A core feature of general practice is that we see patients with undifferentiated presentations. This is one of the hardest jobs in medicine and one with real potential for harm if done badly: not just in the tragic, rare cases that make the news, but also in countless mundane little errors that have the potential to make patients’ lives worse – a delayed diagnosis here, an unnecessary antibiotic there. In its definition of a GP, the RCGP references the ‘complexity, uncertainty and risk’ inherent in what we do.
Those who defend the use of the physician associate role in general practice sometimes cite a 2015 study that suggested consultations with physician associates result in similar rates of re-consultation, prescriptions, referrals and patient satisfaction as GP consultations.
However, I would question how applicable this is now, several years down the line: in the study, almost all of the practices that employed physician associates had protocols about which patients were suitable to be booked in with a physician associate. In turn, this may limit the extent to which these findings can be generalised to physician associates seeing undifferentiated patients.
I believe that the expansion of the physician associate role in general practice is a calculated political strategy. The Additional Roles Reimbursement Scheme (ARRS), introduced in 2019, funds PCNs – and, in turn, practices – to employ various non-doctor primary care clinicians, including physician associates. Some of the roles funded by ARRS, such as first contact physiotherapists, are regulated professions with a clear scope of practice, which has a useful place in primary care.
But it is hard to see the inclusion of physician associates as anything other than an attempt to replace GPs. The NHS England role reference summary describes the physician associate role as being to manage ‘undifferentiated, undiagnosed cases’ and recommends a minimum supervision frequency of just once a month. Given the fact that even GPST3 trainees – who have completed at least four years of postgraduate training after medical school – have weekly tutorials and a named supervisor for debriefs in-between, this is staggering.
Due to a prolonged funding squeeze, GP surgeries are under financial pressure, and many partners report earning less than the salaried GPs they employ. As a result, locum work is drying up as practices simply cannot afford to engage them. Primary Care Facebook and WhatsApp groups are full of GP locums who are essentially now unemployed. One previously full-time locum told me that she had gone from working five days a week to one because work in her area had dried up. Meanwhile, one locum agency encouraged GPs to reduce their rates in order to secure work in the face of competition from ARRS roles.
In the wake of the death of Emily Chesterton after a physician associate failed to recognise that she had a deep vein thrombosis, the Nuffield Trust’s chief executive Nigel Edwards called for a ‘very clearly defined scope of practice’, close supervision by GPs, and for patients to have the right to choose to see a doctor instead. Her mother, Marion, said that Emily did not realise she had not seen a GP.
Earlier this month, the Daily Mail reported the case of a 69-year-old man who had presented to his GP surgery with abdominal pain. A physician associate diagnosed the patient with irritable bowel syndrome, but sadly, he actually had bowel cancer, and died less than a year later.
How many more patients have to die before the powers that be recognise that physician associates are not adequately trained to assess undifferentiated illness with minimal supervision?
Dr Heather Ryan is managing director of Formby GP, a private GP clinic in Formby, Merseyside
*Please note that some details have been changed to maintain patient confidentiality
READERS' COMMENTS [19]
Please note, only GPs are permitted to add comments to articles
The clue is in the name: Physician assistant. A physician focuses on non-surgical treatment of patients in their chosen Medical Speciality. In America, the role is to assist in the initial history taking and observations of patients in these same medical specialities and to perform specific niche tasks. They are not specialised generalists and absolutely do not have the skills to work independantly sifting through undifferentiated illness, both surgical and medical. But the reality is that General Practice is complicit in allowing this state of affairs to arise having embraced the NHSE PCN ARRS model as cheaper alternatives to GPs and it is beginning to bite back, with catastrophic consequences a real and present concern
Thank you, Dr Ryan, for your honest and comprehensive appraisal of the situation! Much needed and appreciated.
I’m pleased that you mentioned all non-GP roles rather than just PAs as replacing GPs.
ACPs (Advanced Care Practitioners) are not necessarily as highly trained as one might assume from their title. They are also seeing undifferentiated cases with minimal supervision.
It is possible for a podiatrist, nurse, paramedic, physiotherapist, pharmacist etc to apply for a ACP course after just 2 years experience in their own field! The course is usually 3 years part-time (so 1.5 year’s additional training?). Is that really enough to start doing the job of a GP with minimal supervision?
Excellent post Dr Ryan with some very good points well made.
‘Primary Care Facebook and WhatsApp groups are full of GP locums who are essentially now unemployed.’
We estimate this figure comes to several thousand [NASGP believes there are 17,200 GP locums across the UK/28% of the workforce] yet the public have repeatedly been told there is a mass ‘shortage of GPs’. Some people are clearly not telling the truth.
As a retired GP, former trainer and recipient of poorly trained young doctors under my wing, I can confidently say that the diagnostic method is poorly taught at medical school, on the wards and in play school.
The favoured ‘spot diagnosis’ of our hospital colleagues rarely, if ever, applies to GP. Maybe I missed all the hoards of myxoedematous old ladies.
The modern fashion for diagnosis by algorithm which are dispensed to nurses and pharmacists are like teaching directions using satnav. The learner never gets to understand the route or find another way if there’s a road block.
Even the hypothetical-deductive method is taught without including the possibility that I might not know what’s going on, safety-netting or (dare I say) to ask someone else. The management of uncertainty is given lip service. The appropriate taking of risks is frowned upon and if it goes wrong, vilified.
What chance do PAs have. They are not, nor should be put in a position to make these complex decisions.
The PAs that foul up should not be responsible for the errors. Their supervisors, GP partners and others who treat them as cheap labour should look at themselves to where the responsibility lies.
Very good article, well done Heather.
I dont agree at all with this assessemnt of the capabilities of Physicians Associates to work in General Practice, assessing and working with undifferentiated patients. We have many PAs working at our surgery and PCN and they are invaluable. They help GPs. They dont make dicisions. The GP supervises every patient contact. GPs make mistakes every day. I notion that PAs are inherently unsafe is just wrong and we need to wake up to skillmix and MDT working.
Dr Steven Cleasby
HALIFAX
UK
Physician associates have a role in hospital medicine where they work effectively in specific roles and become knowledgeable at seeing similar presentations. I really worry about the role in primary care. As a GP working in a practice with PA’s it is clear there is a difference in their level of confidence and ability to manage risk. This worries me as ultimately we the GP is responsible and it is our registration in he line if they choose not to discuss a patient. We don’t have time to discuss every patient and therefore there is an element or risk taking when supervising. This is not safe for patients or the individual doctors supervising.
I concur with steven cleasby – we have had many excellent Physicians Associates and reserve time of our duty doctor so they are able to supervise patient contacts, be asked for advice or jointly see patients.
Just as hospital consultants don’t see every patient to take the full history & total examination, they rely on less experienced junior doctors to take the basic history and examination and then the consultant focuses on the key parts of history or where there may be inconsistences or uncertainties and then may undertake limited focused examination. Nominally hospital care is provided by a consultant and the juniors are just in training posts, but juniors rightly or wrongly feel they often carry the clinical workload, but we delegate and supervise. We’ve always delegated – a practice nurse assessing a patient for annual asthma review who has got worse will have the training and experience to suggest changes to their inhalers, they don’t always need the GP to come see the patient for fear of missing a lung cancer.
What is required are PAs well trained who know their personal limitations and are confident to ask for supervision and advice (my experience so far has been of pleasant surprise at just how good they are and how good at identifying red flags or prudently request sensible investigations). Likewise as a GP need to be willing to delegate (not as if there is a tide of doctors trying to fill empty salaried posts) and be approachable & supportive to provide the supervision, and that we are given the time to undertake this. Clearly unsupervised unsupported PA is not acceptable, and nor is expecting a duty doctor to have a full clinic of patients of their own and other work during the day and then resentfully shoe-horn in a few PA’s queries.
Problem is they don’t know what they don’t know and the lack of undergrad and in the job training together with the somewhat ropey entry requirements leads to a huge variation in ability. Completely different to junior doctors.
If GPs have to review each one what on earth is the point. Easier to see them yourself than have a 70k obs taker?
It is cheap labour that has lower skill levels being thrust into an every expanding complex and litigious system.
The leeches at grabbit&run solicitors are going to move this in a few years.
Thank you, Dr Ryan. Brave statement in several ways.
They don’t know what they don’t know. As more preventable disasters happen, the GP bears the final responsibility through employing them . They will never be employed in private practice like where the author works, but for some reason they are a very expensive saviour of general practice.
Being of a similar vintage to Dr Bennett, and as a former GP trainer and Undergraduate tutor, I am inclined to agree with him that medical graduates and many GP trainees increasingly seemed to lack practical experience and clinical “nous”. Despite being well versed in theory and knowledge of guidelines
Medical educators (and the GMC) have turned medical education and post grad training into a tick box exercise driven by algorithm and protocol that ill prepares students and trainees for clinical practice. But the two year course that PAs follow still pales in comparison.
Financially motivated GP practices and partners will continue to employ Physician Associates into inappropriate roles despite the lowering of quality and increased risks as there is is money to be made and increased drawings for the partners.
They are not interested in the damage caused to the careers of locums as a result and we in our practice would not employ PAs into unwarranted positions.
While there is a a financial gain for certain partners and practices the patients and the public will continue will continue to receive a lower quality service.
There is of course a role for ARRs if used appropriately!
@ Past Caring
“medical graduates and many GP trainees increasingly seemed to lack practical experience and clinical nous”
While that’s clearly true, you imply that it has worsened over your practice. But surely it’s just as likely that you’re the one who’s changed? That as your practical experience and clinical nous has grown, you’ve become more aware when these qualities are lacking in others?
I think GP training is rather good at encouraging clinical “nous”. Though for some it is slower to blossom than for others!
(typo: “over your years of practice”
The narrative in these Comments seems to suggest that greedy Partners are spurning expensive Locums for cheaper PAs in a cynical exercise of recklessly endangering patient care to boost their own personal profits.
This is a gross misrepresentation of recent history. For years Practices would hose away endless thousands on unanswered adverts for salaried/partner GPs, forcing them to employ ruinously expensive Locums.
In the absence of actual GPs, many practices reluctantly employed ARRS, including PAs, in a desperate attempt to avoid the nightmare of closure with potential unlimited liability bankruptcy.
But once we found superb Pharmacists, PAs, ANPs etc, we accidentally reinvented ourselves as “Consultant GPs” supervising a tight knit MDT, with decreasing necessity for Locums.
The advent of PCNs with their ARRS-only staff budgets clearly shows that government have twigged that the future of Primary Care is replacing a GP-heavy service with plentiful PAs supervised by the odd doctor.
Whilst it’s fair enough to denounce this devious government plan as second-rate and reckless, it is ridiculous to blame Partners. We loyally stayed on deck whilst others abandoned ship, and found a way to survive.
Unemployed Locums didn’t want to be tied down with Partnerships or salaried posts. They spurned the “job for life” for the freedom of the road less travelled, with all the uncertainty that comes of the feast or famine of available work.
And fair play to Locums for being there when we need them. But we don’t owe you a living. Back in the day we would do a long Locum post to prove we were Partner-worthy. Trust me, if you shine at your Locum practice, show willingness to do the dirty admin jobs, Practices will be keen to sign you up…..if you are prepared to sacrifice your freedom. There’s no cake and eat it any more.
Would most partners chose a locum to see their patient or a PA if the government gave them funding to choose?
Ive previously been a partner and care about my patients so I would choose quality of care over P.a’s and lower skill mix clinicians.
Problem is NHSE decide, not GP partners, because of imposed business and funding arrangements.
We are all GP’s we dont need to battle each other….