Challenging Findings of Primary Fact and the Burden of Proof: Davies v NMC [2026] EWHC 1139 (Admin)

Davies v NMC is well worth a read for at least two reasons. First, it is an example of a successful challenge to the facts, partly based in the Panel reversing the burden of proof, and partly based in a failure to do an overarching witness credibility assessment. Second, it is also a good illustration of how the High Court gets round the apparently very severe “virtually unassailable” criterion from Southall on overturning primary findings of fact.

Mr Davies was an intensive care nurse of 20 years’ standing. In all, he had enjoyed a forty-year nursing career free of any fitness to practise issues. Until, that is, a patient with a devastating and debilitating disease accused Mr Davies of ‘torturing” him on repeated occasions while administering personal care, and treating him in a dismissive manner.

The patient was wholly dependent on nursing care, as he could do little more than move his eyes and could not speak. He used a tracheostomy to breathe. The allegations included cleaning the patient’s penis with unnecessary force and (although the allegation was not framed in quite this form) putting some sort of inappropriate cleaning liquid into the patient’s penis, as well as ripping electrodes off his chest without warning, lifting the patient’s leg without consent (which was painful), and talking to him in insulting or dismissive ways. All were alleged to be done with the intention (or effect) of intimidating, humiliating, or scaring the patient, or causing him physical or psychological harm.

In the appeal, Mr Davies was challenging the primary factual findings that the events had happened, as well as the inferences as to his intentions when doing the alleged acts. Apart from a couple of minor exceptions, the only witnesses were the patient and Mr Davies. So the court was squarely faced with a request to overturn primary factual findings based on the Panel’s evaluation of the live and written evidence of two witnesses, largely unsupported by other eyewitnesses or contemporary documents.

In the section on the legal framework, Collins Rice J did not cite the Southall quote, although she did note that the court would be “extremely cautious” about upsetting findings of primary fact [at 10]. She stated that while the appellate court has to find that the decision is “wrong”, that does not mean simply that the appeal court would have found differently, but that “the Panel has gone wrong and made decisions it was not properly entitled to make at all”. [at 9, emphasis in original]

The route to deciding that the Panel had gone wrong on the facts was an examination of the reasons it gave. Her criticisms of the Panel included:

• Failing to start with an overarching assessment of reasons to doubt the patient [at 100], including: the intrinsic unlikelihood of the ‘torture’ described; the further unlikelihood, in a busy ITU, that no one else saw or heard anything amiss (lack of corroboration); the fact that, when it was suggested to the patient that other nurses would have been present for many of the events, the patient cited (for the first time) two “disciples” of Mr Davies, who he said were the only other ones present during the abuse. It had also failed to evaluate the poor physical and mental condition of the patient, which may have contributed to believing things that were not true; the patient’s evident animus towards Mr Davis; and the Panel had not dealt with inconsistencies in the patient’s evidence [at 109-120].
• Starting its analysis with a consideration and rejection of speculative reasons put forward by Mr Davies for why the patient might be lying [at 102-103]: Mr Davies, under pressure in cross-examination, had made some suggestions, but the Panel had been expressly warned by the legal advisor to bear in mind these might just be attempts by Mr Davies to find an explanation. This reasoning by the Panel amounted, the judge held, to reversing the burden of proof because the Panel analysed the suggested motives, rejected them, and then used that to support its belief of the patient. The Panel lost sight of the burden of proof – i.e. that the NMC had to prove these things had happened, and Mr Davies did not have to provide an explanation of why the allegations had arisen.
• Rejecting Mr Davies’ evidence despite the fact that it said he had given “consistent” evidence (something that the Panel had mentioned to support its belief of the patient), without explaining why his consistent evidence was less credible than that of the patient [at 106].
• Rejecting, on inadequate grounds, the evidence of another nurse which supported Mr Davies’ account of events on one charge [at 122ff].
• Not apparently applying the good character direction or saying why it had not [at 132].
• Not considering whether some of the incidents could be the result of misinterpretation by a distressed patient of legitimate treatment, although this plainly arose on the evidence [at 134].
• Treating credibility as all or nothing, rather than considering it as something that might be divisible [at 135].

In summing up why the findings could not stand Collins Rice J stated:

“Serious procedural irregularity can be constituted by a failure to maintain focus on the appropriate burden and standard of proof and to assess evidence coherently and consistently by reference to them. It can also be constituted by a failure of reasoning so that, looking at a statement of reasons (and, if the stated reasons are not sufficiently clear, at all the underlying material), a nurse is not fairly able to understand how or why a decision has been taken.”

However in deference perhaps to the warnings in authority about overturning findings of primary fact, she continued [at 143]:

“It is important …  to make clear what I am not deciding. My analysis and decision focus on how the Panel approached its decision-making, not on what it decided as such. [emphasis in original]

Consistent with this, no findings were substituted and submissions were invited from the parties as to what should happen next.

This case could potentially be very helpful in guiding defence practitioners as to the sort of challenges (procedural irregularity via the burden of proof, evidentiary analysis, and reasons) which might get over the high bar to reversing primary factual findings. Its emphasis on what was required for a fair process, rather than whether the factual determinations were right or wrong, opens up a simpler route to factual challenge.