AC-2024-LON-003354 - [2025] EWHC 2270 (Admin)

Ground 1(a) and Ground 1(c): process irrationality

I address the process and outcome irrationality arguments separately but, as I have already observed, neither the claimants nor the BMA differentiated between the two aspects of the rationality challenge either in their written or oral submissions. There is considerable overlap and what follows on process applies to the outcome rationality challenge also.

I start by recording that I do not accept that the contents of Professor Melville’s statements amount to an exercise in which he seeks, retrospectively, to rationalise the approach taken by the defendant. Although in the letter of response the defendant took the point that the 2024 Order did not confer a power on the defendant to impose limits, this letter was written on 16 August 2024. As Mr Dunlop has observed, the process of evolution of GMP and the other advice and guidance started four years’ earlier in 2019 at around the time that the defendant was selected to be the regulator. Whatever the position may have been in 2024 after the defendant had had sight of the draft Order could have had no bearing on its approach some years earlier.

Nor am I persuaded by the claimants’ further point that there can have been, in reality, no active consideration of the issue of the imposition of limits because there is no contemporaneous documentation from the defendant confirming or recording the debate. Professor Melville explains that the decision-making was iterative in the sense that there was no particular date upon which a decision was made not to impose limits on associates’ practice. In his second statement he says that it is not unusual for the defendant not to record formally a decision not to do something and that everyone in the organisation thought that the imposition of scope of practice was inappropriate. He added that the defendant had a very large number of matters to consider when deciding how to regulate associates and that formal records and consultation tend to relate to actions which the defendant was minded to take and not to actions which no decision maker thought appropriate. I accept his evidence.

In respect of the process challenge, as I understand the claimants’ case, they make two main points. First that the defendant adopted a flawed starting point. Second that the defendant failed to take into account the extensive evidence of patient safety risk due to associates working beyond their competence.

I reject the submission that the defendant’s starting point was irrational. The decision to apply the medical model to associate regulation:

The defendant took into account the findings from its own COI survey. It accepts that those findings demonstrated significant concerns over patient safety due to associates working beyond their scope of practice, concerns over the supervision of associates and delegation of tasks to members of the healthcare team. The defendant considered that those concerns would be addressed by regulation. This response was not irrational. The registration process would include quality assurance of all courses leading to qualification and for those who are already associates the registration process would include evidence that the associate possessed the specified competencies. The defendant was satisfied that by 2026 (although it anticipated the majority of currently working associates to have sought registration within 6 months) all associate registrants would have demonstrated their ability to comply with professional standards and recognise the limits of their competence. The regulation process will include a process of revalidation and appraisal. I agree that other surveys (BMA, DAUK) revealed the same concerns, albeit that those surveys may have been of lesser quality (due for example to bias and anonymity of respondents) and importantly Mr Dunlop does not dispute that much of what is recorded by the claimants in annex 2 of the skeleton argument “evidence of problems with AAs and PAs” from practitioners, Royal Colleges and patients is germane. However, I accept his further submission that, having identified those serious concerns through its own COI survey, the defendant was not obliged either to seek out further evidence of the same points of concern and review or to reconsider its position on limits of practice in the light of other material to the same effect.

Mr de la Mare appeared to suggest (in his response) that the evidence of risk was qualitatively different by 2024 compared with earlier in the process leading up to regulation. I do not accept this. By 2024 there had been more surveys but they revealed the same concerns which had been raised earlier. There was evidence of patient safety issues from a different source, in the coronial reports which were generated between 2023 and 2025. Such reports contain material of great importance. But following Ms Chesterton’s death, there was no PFD report (to any person or agency) and no doctor was referred to the fitness to practise processes of the defendant. There was a scope of practice document which if followed ought to have prevented the associate from seeing Emily Chesterton on (at least) the second occasion. The report concerning the death of Mr Peters was not addressed to the defendant, but the information provided by the Trust concerning that death did not appear to raise patient safety concerns about PAs (albeit that document has been reviewed by Professor Melville in the context of this claim).

There were two PFD reports (Susan Pollitt 31 July 2024) and Pamela Marking (24 February 2025) addressed to the defendant, amongst others. Professor Melville’s response to each was to emphasise that the concerns would be addressed by the introduction of regulation. This response was not irrational. At the time of the deaths, associates worked in regulated healthcare settings and had to be supervised, but there were no profession-specific mandatory standards for their pre-qualification training or education or conduct, nor any professional accountability to any statutory body.

In addition to his complaint about the starting point (which I reject), Mr de la Mare submits that the defendant took another wrong course by applying a “compelling evidence” test. In my judgment, however, Professor Melville was not intending to set out a legal test which the defendant was then applying to the question of limits. All that he was saying was that the medical model had been a successful model of regulation and there would therefore need to be good reasons to do something different. That was a rational conclusion. The defendant was not entrenched in its view, as Mr de la Mare suggests. Rather, it raised questions relevant to the limits issue in its own COI survey and considered the concerns raised. The question of limits appears to have been raised on at least two occasions in meetings of the EAG without objection.

As Professor Melville states, the defendant’s approach to the imposition of limits on associates’ practice was informed by its understanding of the role of employers and local clinical governance arrangements in ensuring that associates practised safely. Scopes of practice or limits on practice were best identified and implemented locally, taking into account the individual competence and experience of the associate. The level of supervision was best determined locally on an individual basis. This is how supervision generally within the NHS workforce is undertaken. Guidance had been issued by other bodies on the role of associates and the need to have clinical governance arrangements in place. In March 2024, NHSE published guidance for medical directors which emphasised the need for Trusts to have policies and systems in place to ensure that associates were supported, supervised and integrated into the multidisciplinary team. Clinical governance is the responsibility of NHS employers and the various oversight bodies.

I find no failure by the defendant to take into account material evidence or mandatory relevant consideration. I find no logical error or critical gap in the defendant’s reasoning. The defendant’s approach was I find coherent and not irrational. I therefore reject the claimants’ submission based on process irrationality.