Johnson V Williams – Avoiding The “Holmesian Fallacy” And The Importance Of Responsible Expert Evidence

In June of this year judgement was handed down in the case of Johnson v Williams . The judgement contains useful pointers on applying the burden of proof (namely, do not be led astray by Sherlock Holmes!), and on the importance of experts being responsible and not adopting extreme positions when giving evidence.

The facts

The claimant, Mr Roger Johnson, was a professional footballer signed with Charlton Athletic Football Club. In January 2017 he suffered a meniscal tear to his left knee whilst in training. The defendant, Mr Williams, performed surgery to treat the same, which initially appeared to have been successful. Subsequently, however, the claimant’s knee showed signs of infection. The defendant performed a further surgery to remove infective material from the claimant’s left knee on 17 March 2017 (a synovectomy procedure), followed by a washout and arthroscopic debridement on 19 March 2017.

A few days after these procedures, a fluid filled egg-shaped lump appeared on the medial side of the claimant’s left knee, and an MRI scan on 11th April 2017 (some 25 days after the synovectomy) revealed a 3cm diameter tear to his medial retinaculum (the tendon that crosses the knee joint on the medial side of the patellar).

Unfortunately, despite treatment, the damage to the claimant’s medial retinaculum meant that he had to abandon playing football professionally. He therefore brought a claim for damages, including for the loss of the earnings that he would have received from his footballing career but for the alleged medical negligence.

At trial

The preliminary issues that Jeremy Hyam QC (sitting as a deputy judge of the High Court) had to decide were:

1. i) Did the defendant cause a large tear/defect in the claimant’s medial retinaculum on 17th March 2017 in the course of surgery (a synovectomy procedure); and
2. ii) If so, was it negligent for him to have done so.

The parties both instructed orthopaedic experts; Mr Ashok Paul for the claimant and Mr Sanjay Anand for the defendant. In their joint statement, the two experts agreed that a large rupture to the retinaculum would be an “extremely rare and unusual complication following synovectomy”. The judge unsurprisingly accepted this at trial.

There was some disagreement between the parties over when exactly the retinaculum damage was sustained but, upon review of the evidence, the judge was satisfied that the retinaculum damage either occurred at the time of surgery or later but before 23rd/24th March (i.e., within a window of 7/8 days from the date of the claimant’s synovectomy).

Despite the proximity of the onset of symptoms to the alleged negligence, the judge found that it was “highly improbable” that the medial retinaculum damage had been caused by the defendant. The judge focussed on how unlikely it was that the defendant, given his experience as a surgeon, would have caused the index injury in the synovectomy procedure without noticing his mistake on that day or during the further procedure two days later.

However, the judge also found it was unlikely that the two other possible explanations for the damage put forward by the defendant (namely alleged “non-compliance” by the claimant with the rehabilitation instructions, or an infection) had been the cause of the claimant’s medial retinaculum damage.

Burden of proof and the Holmesian fallacy

On the basis of the above, the claimant submitted that, even though it was very unlikely and perhaps “remarkable” that the defendant had caused his injuries, that was on the balance of probabilities what had happened in his particular case, because the other competing causal mechanisms put forward by the defendant were so remotely unlikely that they could be essentially ruled out, leaving the defendant’s surgical error as the only possible cause of his index injury.

In so doing, the claimant invoked the logic of Sherlock Holmes, who in The Sign of the Four said to Watson:

“How often have I said to you that when you have eliminated the impossible, whatever remains, however improbable, must be the truth?”

Unfortunately for the claimant and Sherlock, this logic has been criticised and in fact gives its name to a logical error known as the “Holmesian fallacy.” In outline, the fallacy is committed when one believes that a particular explanation is correct on the basis that alternative explanations have been ruled out, when in fact not all alternative explanations have been ruled out because (at least in most real life, complex situations) it is impossible to do so.

The judge was alive to this point and referenced it directly in his judgement, along with its application in the influential case of Rhesa Shipping Co. SA v. Edmunds [1985] 1 WLR 948. He therefore opted not to find that any of the explanations put forward, by either claimant or defendant, were on the balance of the probability correct. As the burden of proof was on the claimant to prove his version of events, the claimant therefore failed in his claim.[1]

This case is a salutary reminder to claimants that, unlike Sherlock Holmes, judges are under no obligation to solve every mystery. When faced with a put to proof defence, claimants should bear in mind that to succeed they must prove that it was probable that their losses arose from the defendant’s negligence, even after taking into account the fact that it may be impossible to rule out definitively all other possible explanations.

Expert evidence

As will already be apparent, the claimant’s case was a difficult one to prove. To make matters worse for him (and the court), the manner in which his instructed orthopaedic expert Mr Paul’s tendered evidence damaged his credibility as an expert and incurred criticism from the judge.

An exchange with the Defendant’s counsel (quoted below) typifies Mr Paul’s approach. In it, the Defendant attempts to highlight how unlikely it is that the Defendant could have caused the index injury and then not noticed it:

• “Q: So he would have to shear all the way across six times with this instrument 5.5mm multiplied by 3 gets you the 3 cm tear so he would have to cut the whole width of the shearer times 6; do you agree?

• A: Yes, I agree.

• Q: You would absolutely be aware of that?

• A: Yes, but it did happen. There is no other possible explanation, absolutely none”.

In saying that “There is no other possible explanation, absolutely none”, Mr Paul was not only committing the Holmesian fallacy, he was also adopting an extreme position and was failing to make any concessions to the Defendant’s case[2]. The judge was unimpressed with this, and found that it was “far too extreme and inflexible a position to take for such an unusual complication”.

This case is another reminder to experts (and those who instruct them) on the importance of tendering opinion evidence in a responsible way. When faced with a difficult case in which there are a range of possible explanations for what has occurred, experts should acknowledge the complexity of the situation and give a thought to alternative explanations before reaching their conclusions, rather than adopting a blinkered approach that will serve only to undermine their credibility.

[1] The judge did not go so far as to say that the defendant’s explanations for the claimant’s injuries were impossible, but he implies that, even if those explanations were shown to be incorrect, the claimant would not have succeeded in overcoming the judge’s “very considerable doubt” that the index injury was caused by the defendant in the synovectomy. The judge considered that the causation of the claimant’s injuries may have been multi-factorial, but the claimant did not advance a case based on material contribution.

[2] In direct contradiction to the guidance given by Green J to experts in the helpful case of C v North Cumbria University Hospitals NHS Trust [2014] EWHC