Harris v Johnston

The Claimant, Mrs Sally Harris, is a 61-year old lady with a history of back pain which has attracted the attention of spinal orthopaedic surgeons, neurosurgeons, pain specialists and experts in pain neuro-modulation over many years. At the time of the events giving rise to this claim, she had already undergone extensive surgery to her cervical, thoracic and lumbar spine. Her claim concerns a revision foraminotomy carried out privately by the Defendant, Mr Johnston, a consultant neurosurgeon, at the C6/7 level of her cervical spine on 3 November 2011.

Each nerve root coming from the spinal cord passes through a bony canal on each side of the cord which is called the neuroforamen. If the canal narrows (the medical term for which is “stenosis”) as a result of bony growth, the compression applied on the nerve root from the bone and/or soft tissue can cause severe pain. A foraminotomy is an operation to enlarge the passageway where the spinal nerve root exits the spinal canal by removing enough bone from the edge of the canal to create more space and relieve the compression on the nerve root. The most severe narrowing is always at the proximal part of the canal. The surgery itself is very delicate, and is carried out under the microscope. It is important for the neurosurgeon not to remove too much bone, particularly when gaining access, as damage to the adjacent facet joint could lead to destabilisation of the spine.

There is no dispute that in the course of the operation an instrument used by Mr Johnston, a Cobb dissector, not only penetrated the dura, which is the sheath of membrane protecting the spinal cord, but came into contact with the spinal cord itself, on the extreme right hand side, causing an injury to it. The issue at the heart of this case is whether the injury was due to clinical negligence, for which Mr Johnston is legally liable, or whether it was a tragic accident for which no-one is to blame. The parties have agreed that if liability is established, damages should be awarded in the gross sum of £725,000.

The Claimant’s claim in respect of the spinal cord injury was originally put in three different ways. First, it was alleged that the initial foraminotomy carried out by Mr Johnston on Mrs Harris in the same location, in May 2010, was not competently performed, leading to the return of her symptoms and her decision to undergo a revision foraminotomy the following year. Secondly, it was alleged that Mrs Harris was not properly advised by Mr Johnston as to the risks involved in the second operation; and that if she had received such advice she would have postponed the surgery, thus avoiding the damage which occurred during that operation. Finally, it was alleged that the second operation was not competently performed. On the first day of the trial, the first two complaints were discontinued, in circumstances I shall go on to describe.

There was also a subsidiary claim for damages for clinical negligence relating to the way in which the damage to the dura occasioned during the second operation on 3 November 2011, and the ensuing leakage of cerebrospinal fluid (“CSF”), was treated. The Claimant’s pleaded case was that Mr Johnston should have repaired the dura in the course of the operation, or in any event stopped the leakage of CSF at some point prior to 18 November 2011, when Mrs Harris developed symptoms of a pneumohydrocephalus (pressure on the brain due to air entering the cranial cavity) and became confused and disorientated.

At the beginning of the trial Mr Davy, on behalf of the Defendant, applied for summary judgment in respect of this discrete aspect of the case. Mr Johnston’s evidence was that there was no segregation between the dura and the surrounding soft tissue at the point where the instrument penetrated. There is no dispute that, when a surgical repair was carried out on 28 November 2011, he repaired the soft tissues rather than the dura. Prior to that he had tried a variety of conservative measures recorded in the nursing notes, and a second opinion had been obtained from another consultant, Mr Stapleton, at the insistence of Mr Harris who was understandably very concerned at his wife’s condition.

The experts in neurosurgery, Mr Peter Kirkpatrick (for the Claimant) and Mr Robert Macfarlane (for the Defendant), agreed that if the lack of segregation was factually correct, it was not negligent of Mr Johnston to fail to repair the dura during the operation. Mr Kirkpatrick was critical of the speed with which the operation was performed (the agreed time for performing the surgery itself was around 30 minutes) and suggested that within that timescale, insufficient time was taken by Mr Johnston to ascertain whether or not the dura was susceptible of repair. However, there was no pleaded allegation in relation to that, and if the dura was indeed adhering to the soft tissue, those points did not improve the Claimant’s case, since however long he took Mr Johnston would have reached the same, correct, conclusion that he could not carry out a primary repair.

Mr Davy submitted that in the light of the agreed expert evidence and the absence of any factual evidence that the dura was separate from the soft tissue, the claim in negligence based on the failure to repair it during the operation itself could not succeed. Despite the failure by Mr Johnston to keep proper and detailed records, (of which I shall have to say more later in this judgment) and the obvious difficulties that this posed for the Claimant, I saw a great deal of force in that submission. Not only was there nothing in Mr Kirkpatrick’s report to indicate any evidential basis for inferring, let alone concluding that the position was other than described by Mr Johnston, there was a wealth of evidence from which the inference could be drawn that Mr Johnston’s description was more likely than not to be correct.

First, there was no dispute that the adhesion of the dura to the soft tissue was a likely consequence of the previous operations in the same area. Next, there had been a non-negligent penetration of the dura on the right hand side by a diathermy tool during the May 2010 foraminotomy carried out in the same place by Mr Johnston, which had been addressed by the suturing of the soft tissues. No complaint has ever been made about the cause of that dural leak or, more importantly, the steps taken to address it. Precisely the same steps were taken surgically on 28 November 2011 to deal with the leakage of CSF following the second operation (although the nursing notes establish that the earlier conservative remedial measures had actually stopped the leakage for several days until Mrs Harris unfortunately fell over on the ward and the wound was re-opened).

So far as the alleged failure to take steps to repair the leakage of the CSF before 18 November 2011 is concerned, the experts agreed in their joint report that if it were possible to have repaired the leak sooner, Mrs Harris may have avoided the pneumohydrocephalus, but they did not say that it was more likely than not that she would have done so. It would be impossible for the Court to reach a conclusion on the probabilities of development of a pneumohydrocephalus without expert assistance.

Mr Davy submitted that even if the Claimant could establish the failure to carry out the surgical repair prior to 18 November 2011 was negligent, (which was disputed) she could not prove a causative link between that negligence and the development of the pneumohydrocephalus and other unfortunate complications from which she suffered. Of course Mrs Harris only needed to establish on the balance of probabilities that the failure to repair earlier was a contributing factor, not that it was the sole contributing factor. However, in the light of the agreed position of the two experts, who are highly experienced in giving evidence in trials of this nature, he submitted that she could not prove that it was a probable cause rather than a possible one. I made it clear that I would not permit a re-opening of that agreed position in cross-examination, there being no evidence before me that Mr Kirkpatrick had changed his mind or was seeking to resile from the agreed position.

After I had heard argument on these points from both counsel, Mr Woolf requested time to consider the matter with his client, which I granted. This resulted in the discontinuance of the allegations relating to the management and repair of the dural leak which emanated during the second operation, without the necessity of a formal ruling on the summary judgment application.

I have already mentioned that two of the three ways in which the claim in respect of the spinal injury was originally pleaded were abandoned on the first day of trial. Mr Davy had made a further application for summary judgment in respect of the claim that Mrs Harris was given insufficient advice about the risks of the second operation. This was based on the agreed position of the experts that if the Court accepted Mr Johnston’s account of the counselling process, (as set out in his witness statement) that amounted to at least the minimum level of acceptable pre-operative information that Mrs Harris required. The Claimant’s pleaded case was that he should have said more to her than that. In the same way as the allegations about the treatment of the dural leak, this aspect of the claim was discontinued after I had heard the rival legal arguments on the summary judgment application.

The final allegation that was discontinued (the criticism of the earlier foraminotomy) was not the subject of an application for summary judgment, because it was supported by the expert opinion of Mr Kirkpatrick: but it was obvious that his opinion was logically deficient. In the joint report, Mr Kirkpatrick had conceded that the 30 minutes taken by Mr Johnston to perform that particular operation on Mrs Harris was within the spectrum of times that a reasonable surgeon could have taken to perform a foraminotomy, and that it was not possible to say whether the procedure was performed negligently based solely on the time it took. He also accepted that one could not infer negligence from a comparison of the CT scans taken before and after the operation, and he accepted that a patient’s symptoms could return even after a non-negligent foraminotomy. Yet he was adamant that a combination of all three of these factors demonstrated that the operation had been incompetently performed.

There is a world of difference between a situation in which a number of factors which are indicative of negligence, when viewed collectively, demonstrate negligence on the balance of probabilities, even if taken individually they would be insufficient; and a situation in which it is sought to demonstrate negligence by adumbrating a collection of factors, none of which is even indicative of it. The situation here fell into the latter category, but when this was put to him in cross-examination Mr Kirkpatrick did not appear to appreciate, let alone accept the difference. He sought to draw an analogy between the present case and a road traffic accident in which the motorist who crashed was driving at 59 mph in a 60 mph zone in fog and in wet conditions, and said that the road accident investigator would look at all those factors together in reaching the conclusion that he was driving unsafely.

Whilst that is of course true, the analogy is flawed. The point demonstrated in the traffic accident scenario, by looking at the whole picture, is that the driver’s speed was excessive bearing in mind the factual conditions on the road – but in this case the experts agreed that the speed of the operation was not excessive given Mrs Harris’s presentation. That being so, the speed of the operation cannot be transformed into negligence by two obviously neutral factors, namely, the return of the symptoms within a year, and the inability to see on the CT scan that a significant amount of bone was removed.

It may well be that belated realisation that his position on the overall picture was untenable caused Mr Kirkpatrick to seek to resile from the concessions he had made in the joint report and to opine, during his cross-examination, (as he had in his original report) that no competent surgeon could have carried out a proper foraminotomy on Mrs Harris in 30 minutes. He said that the other two factors on which he relied “amplified” that position. This stubborn adherence to a position which was logically indefensible was one of a number of factors which substantially undermined Mr Kirkpatrick’s credibility.

Another unsatisfactory feature of Mr Kirkpatrick’s approach was that he appeared to equate professional negligence with the degree of competence that had to be demonstrated to pass a surgical examination. Thus when he was asked in cross-examination whether, at the time when he wrote his report, he considered the CT scans to be indicative of negligence, he replied “it was substandard in that the pathology had not been addressed”. When counsel asked whether by substandard he meant negligent, his answer was: “I think that operation was not carried out to a standard expected of, for example, an exiting exam individual.”That led me to question in my own mind whether Mr Kirkpatrick had ever addressed his mind properly to the principles set out in Bolam v Friern Hospital Management Committee [1957] 1 WLR 582. He should have asked himself whether what Mr Johnston did fell below the standards to be expected of the reasonably competent experienced neurosurgeon performing that operation on this patient, not whether an examiner would have failed a student who had done what Mr Johnston did.

If and to the extent that he did have the Bolam test in mind, it appeared that his reasoning depended substantially upon the alleged failure to achieve the “endpoint” aimed at (i.e. the alleviation of the patient’s symptoms) for more than a few months. That approach was founded on a different, but equally serious logical fallacy. Neither expert was critical of Mr Johnston’s advice to Mrs Harris that there was a 90% chance of success (which meant there was a 10% chance that the foraminotomy would not alleviate her symptoms). Therefore, one cannot infer professional negligence merely from the fact that the operation does not have the desired outcome. This was one of the points made by the Defendant’s expert, Mr Macfarlane. He said that a surgeon who makes a clinical judgment as to what he needs to do to deal with the pathology and who turns out to have removed insufficient bone is not necessarily negligent. It does not follow from the fact that in hindsight it transpires that he has not done enough, that his judgment was impaired or that he fell below the standards to be expected of a reasonable surgeon at the time when he made the decision as to how much bone to remove. He said that he had found himself in that position, and he doubted if there was any other spinal surgeon who had not had that experience at least once in his surgical career. It was plain to me that Mr Macfarlane had the Bolam test well in mind.

There is even less justification for drawing that inference if, as in this case, the procedure does have the desired outcome, but not for as long as one would have hoped. Nor can one infer it from that fact, plus the fact that the operation was at the fast end of a reasonable spectrum and the fact that the CT scans showed no substantial difference before and after the procedure. Zero plus zero plus zero does not equal three.

Mr Macfarlane’s evidence, which I preferred, was that the amount of bone to be removed in a foraminotomy is a matter of surgical judgment, based on what the surgeon finds when he opens the patient up, and the amount needed to relieve the pressure on the nerve may be so tiny that the difference would be almost imperceptible on a CT scan. He did not subscribe to the view that in order for a foramintomy to be competently performed, it must be possible to see clearly on a CT scan that a significant amount of bone has been removed. Nor did he criticize the speed of the operation, though he said that he would personally have taken longer. Some surgeons are faster than others (indeed, Mr Johnston’s unchallenged evidence was that he had carried out numerous successful foraminotomies in a similar timescale). Mr Macfarlane’s opinion was that given the operative difficulty and the fact that Mrs Harris’s arm pain resolved, the procedure did not fall below an acceptable standard of care.

Following some preliminary discussion of the logical fallacy underpinning Mr Kirkpatrick’s opinion on this aspect of the claim, in the course of counsel’s opening submissions, the decision was taken by Mrs Harris not to pursue that line of argument further and the claim in respect of the first operation was discontinued. Nevertheless, it remains relevant to my assessment of the reliability of the expert evidence.

Against that background I turn to consideration of the case concerning the spinal cord injury that occurred during the 3 November 2011 revision foraminotomy.

Before the first relevant foraminotomy was carried out by Mr Johnston, Mrs Harris had undergone 22 operations on her spine, 5 of which were on her cervical spine. These included a cervical canal decompression, with bilateral foraminal decompression, at C6/7, on 26 March 1994, followed by an open door laminoplasty at C5 to C7 on 15 September 1994. The latter is an operation designed to reduce compression of the spinal cord as a whole. The spinal cord is normally protected posteriorly by an arch of bone. A laminoplasty involves a full thickness bone cut to the protective lamina on one side of the spinal cord, and on the other side, a partial thickness bone cut to create a hinge. This allows the arch of the spinous process and remaining lamina to be slightly levered away from the spinal cord to relieve pressure on the cord. The operation was a particularly tricky one in Mrs Harris’s case, in the light of the surgery in that area six months previously. The laminoplasty carried with it a considerable risk of spinal cord injury, far greater than the remote risk of such injury occurring during a foraminotomy, because the neurosurgeon would be operating directly above the spinal cord.

In consequence of the laminoplasty, Mrs Harris had no bony lamina on the right hand side to shield the dura, and the usual spinous processes on that side were no longer present.

Mr Johnston is a highly experienced neurosurgeon specialising in spinal surgery. He has a combined NHS and private practice, working one day a week seeing private patients at the Parkside Hospital in Wimbledon. He operates on approximately 600 patients a year, most of whom are spinal surgery cases, though the number of patients with whom he has consultations would be far in excess of that. By May 2010 he had performed around 2,000 foraminotomies. However, Mrs Harris was the first patient upon whom he performed such an operation who had previously undergone a laminoplasty.

Mrs Harris was referred to Mr Johnston by a consultant neurologist, Dr Trend, who wrote in his own records at the time that “her cervical spine CT shows a significant degree of narrowing of the left C7 neural exit foramen… [Mrs Harris] remains in significant pain and I see little alternative to decompressing her left C7 nerve root.”

Mrs Harris was seen by Mr Johnston as a private patient in his clinic on 20 May 2010. His record of the consultation consists of a letter he wrote to Dr Trend the next day, the contents of which were dictated by him immediately after the consultation, as is his normal practice. The letter records that Mrs Harris told him that nine weeks previously she noticed the onset of very sudden left sided arm pain which radiated down into the index and middle fingers, but occasionally could affect all fingers. The pain was very severe and disrupting her everyday life.

As well as physically examining Mrs Harris, Mr Johnston examined the CT scans requested by Dr Trend, which had been sent to him on a CD. A CT scan was the best imaging that could be obtained in respect of Mrs Harris at that time, because she had a spinal stimulator implanted to try and address the pain in her lumbar spine, which precluded her from having any MRI scans. This point was initially overlooked by Mr Kirkpatrick, who criticized Mr Johnston in his expert report for not procuring an MRI scan. The CT scans showed that there was what Mr Johnston described in his letter as “a very, very severely narrowed left C6/7 nerve root canal” due to a build-up of bone. This was consistent with the symptoms Mrs Harris had described. According to his letter to Dr Trend, Mr Johnston told Mrs Harris that he had no alternative but to offer her a foraminotomy, which would give her at least a 90% chance of ridding her of her symptoms. The only specific risk that he mentioned was the risk of nerve root damage, which he quantified at 0.1%. Mrs Harris decided to take up the offer to have the foraminotomy and was booked in for surgery on 27 May 2010.

The risks and benefits of the operation should have been recorded, at the very least, on the consent form signed by Mrs Harris prior to the surgery, but they were not. This is just one example of Mr Johnston’s failure to keep proper or adequate records. The consent documentation is virtually illegible. Mr Johnston accepted in cross-examination that none of the three consent forms relating to the surgery he performed on Mrs Harris in 2010 and 2011 (including the form relating to the surgical repair carried out to the dural leak on 28 November 2011) was fit for its purpose. He forgot to sign one of the forms. His record-keeping, such as it was, with regard to consent, was contrary to the published guidance from the Royal College of Surgeons on “Good Surgical Practice” 2010 edition, which among other matters required surgeons to keep “legible, comprehensive and contemporaneous notes.” It was also contrary to the General Medical Council’s published guidance entitled “Consent – patients and doctors making decisions together” which had come into effect on 2 June 2008.

Three of the operations that Mrs Harris had previously undergone in this particular area had been from a posterior approach, which was also the direction from which each of the relevant foraminotomies would be performed. Thus the surgeon could expect to encounter scar tissue and adhesions, making the procedure more complicated than if it had been a first operation on a “virgin spine”.

In both the May 2010 and November 2011 operations (the latter being a revision of the procedure carried out the previous year), Mr Johnston performed a bony decompression. The bone was to be removed from the edge of the canal on the left hand side using a very small tool called a Kerrison punch. This required sufficient access space to be created to enable the instruments to be passed through at the correct angle to reach the nerve root canal. Mr Johnston began by making a midline incision using a scalpel or a knife. He then created the access space by dissection of sufficient muscle tissue, initially on the right hand side, using a diathermy tool to cut, and small strokes of a Cobb dissector to pull the fibrous tissue away from where the spinous process would have been but for the laminoplasty, towards the right hand side, using what Mr Johnston demonstrated as a scooping mechanism, with a degree of both downward and lateral pressure, in order to gain access for the diathermy. A Cobb dissector is a blunt, spoon shaped instrument, which in this context was being used to retract the fibrous tissue. The two instruments were used in tandem, with the Cobb dissector in Mr Johnston’s right hand, because he is right handed.

Mr Johnston was conscious that there was no lamina on the right hand side. He was therefore setting out to make a relatively superficial pocket on that side, with a controlled dissection down to the depth he thought was adequate to get the retractor blades safely in (he uses a particular type called a McCullough retractor) and then go over to the left hand side. As he proceeded with the dissection he was exposing the fibrous layer from the midline remaining soft tissues, which would enable him to see that there was still fibrous tissue in the way, so as to keep sufficiently far away from the dura. Neither expert criticized that methodology, or the decision to go in from the right, though Mr Kirkpatrick said he had never known a Cobb dissector to be used in this manner.

Mr Macfarlane’s experience was plainly different, as is evident from this explanation that he gave:

“when you make the skin incision with tough tissue, it doesn’t part, so the tissue is together. If you then use the cutting diathermy straight away, the diathermy will burn the skin. So you need the tissue – the skin out of the way a little. Sometimes you can put a retractor in – not the retractor you’ve seen here, but a superficial retractor – but as you’ve seen, it’s got curled blades underneath, so you need to have enough width in order to get those in.

So what we would often do, under those circumstances, is make a skin incision, but when we are going to use the diathermy we have to move the tissue out of the way, so we don't burn it with the cutting diathermy. And then as the cavity starts to deepen, as long as you've got it long enough, it will then start to move away. We don't worry about the skin anymore, and then you are just interested in the direction of travel that you are going”.

I asked Mr Macfarlane how far down the surgeon needed to go in order to get to the level where the retractor would fit. His response was as follows:

“He doesn’t want to be cutting through the muscle, so he wants to be in the plane between the muscle and the fibrous tissue, but he doesn’t want to be on the other side of the fibrous tissue where he’s going to find the dura, so he wants to be on the plane between them. And that’s the logic for using the Cobb. If you use cutting diathermy, it’ll cut through anything, but if you use the Cobb in a gentle way, it will tend to part in the plane that you want it to part. And that’s the reason for using that dissector in that way”.

Unfortunately, on 27 May 2010, as Mr Johnston was making the pocket in the tissue in the manner described, the diathermy breached the dura on the right hand side. Mr Kirkpatrick and Mr Macfarlane agreed that an inadvertent breach of the dura leading to a leak of CSF is, in and of itself, an acceptable complication of a foraminotomy. In Mrs Harris’s case, the pre-existing defect in the lamina caused by the earlier laminoplasty rendered the dura vulnerable on that side. Mr Johnston went just a little too deep in his dissection. No-one has suggested that this was negligent. The hole was small. Mr Johnston effected a repair at the end of the procedure by suturing the superficial tissues together, probably at fascia level because that is where Mr Macfarlane explained he would have been able to create a watertight seal. The dura did not exist as a separate layer and it was impossible for him to repair it as such.

It is Mr Johnston’s practice to dictate his operation note into a Dictaphone which is kept in the operating theatre, as soon as the operation has been completed, and then to get his secretary to type up the note. There is nothing wrong with that methodology, so long as the note is sufficiently detailed. Mr Johnston’s operation notes were too sparse to comply with the Royal College of Surgeons’ guidance on note keeping in “Good Surgical Practice” to which I have already referred. His note for the operation on 27 May 2010 reads as follows:

“With the patient prone, I localised the C7 interspace and then performed the usual muscle strip. This was quite difficult because of scar tissue and on the right side I did open the dura inadvertently. The approach, however, was made on the left to the C7 nerve root canal and this was decompressed thoroughly. I then closed the vacuum drain with vicryl and staples. The patient arose with no obvious neurological deficit.”

Although Mr Johnston did not record what it was that opened up the dura, Mr Macfarlane said it was unnecessary to spell this out, because it would have been obvious in context (at least to a medical professional) that it was the diathermy tool. It would be a bit like requiring a surgeon to note that he used a scalpel to make the incision in the first place.

The operation appeared to have been successful, as evidenced by the fact that the pain in Mrs Harris’s arm was resolved. That was recorded when Mr Johnston reviewed Mrs Harris and discharged her on 22 July 2010. He described her as having “done very well” since the foraminotomy. His discharge note stated that the wound had healed very well, and that she had no neurological deficit in her upper limbs and symmetrical reflexes, including the triceps.

However, approximately ten months after the operation, in April 2011, Mrs Harris started suffering from really serious neck pain, and paraesthesiae which she described as feeling like she had scorpions under her skin, causing her arms to become so itchy that she developed scars from scratching them so vigorously. Some pain in her arm had returned at an earlier juncture, but Mrs Harris was unable to put a date on its recurrence other than to confirm that it was some time after she was discharged in July 2010. Although she saw a consultant neurologist for pain management in September 2010, there is no mention of these symptoms in his records. However, Mrs Harris, who I found to be a transparently honest and fair witness, explained that the earlier pain in her arm was not sufficiently bad for her to complain about it to her GP or to any other medical professional. She was used to coping with some degree of pain all the time.

Mr Johnston saw Mrs Harris in his clinic on 14 April 2011. His note of that consultation reads as follows:

“[Mrs Harris] has had 2 weeks of pain in the left clavicle with some radiation down into all fingers of the left hand and some into the left side of the face. The pain if anything, has fortuitously got a little better.

On examination she had no weakness in her upper limbs and her adduction test was negative implying negative acromioclavicular pathology. I have agreed with [Mrs Harris] that as her symptoms have only been present for a couple of weeks and are showing some signs of improvement we are best to leave things be. If her symptoms do not settle down she can telephone me within the next few weeks and I will order a CT scan”.

Therefore, symptoms of a similar nature and degree to that which had caused Mrs Harris to undergo the foraminotomy in May 2010 had returned at around the end of March or beginning of April 2011.

The symptoms did persist, so Mrs Harris rang Mr Johnston on 14 August 2011 and arrangements were then made for a further CT scan to be carried out on 18 August 2011. The results of the scan confirmed that Mrs Harris was still suffering from cervical stenosis on the left hand side, described as “moderate to severe”. Since the experts agree that there was insufficient time for this to have been due to bony regrowth, Mr Johnston must have removed insufficient bone in the first foraminotomy to give Mrs Harris lasting relief. Mr Johnston explained in his witness statement that the stenosis revealed by this CT scan was caused by a hypertrophied uncinated process, a hook shaped projection of bone on the lateral edges of the top surface of some of the cervical vertebrae, adjacent to the foramen, which had overgrown. He said that this change had been present at the time of the first operation, but was not then causing any symptoms (it follows that the pressure which gave rise to the symptoms on the earlier occasion was caused by other bony growth or a combination of bone and soft tissue). Mr Johnston also explained that it was possible that the onset of the symptoms in 2011 could have been triggered by a simple movement, such as twisting the neck in an unnatural manner. I accept that evidence, which was not challenged.

Mr Johnston spoke to Mrs Harris on the telephone on 5 September 2011 following receipt of the results of the CT scan. Although there is a complete absence of records as to what he advised her and when, Mr Johnston and Mrs Harris were in agreement that at around this time, he advised her that she was essentially in the same position as she had been prior to the surgery in May 2010 and that he could perform a revision C6/7 foraminotomy, but that she should wait until the symptoms became intolerable before doing so. He also advised her that the operation would involve the same issues and risks as the previous operation.

The note of the telephone call is brief in the extreme: “spoke to [Mrs Harris] today. She will come back to me if her arm persists with pain”. Mr Johnston’s explanation for this in cross-examination was that he was not advising Mrs Harris to undergo surgery but rather, advising against it at that juncture, and advocating conservative measures. There still should have been a proper record of what his advice was. It is extremely fortunate for Mr Johnston that Mrs Harris’ recollection of what he said to her broadly accords with his, and that the experts agree that what he said was acceptable, given that this was an intelligent lady who had undergone numerous procedures on her spine already, including the May 2010 foraminotomy in which she had suffered a dural leak. Although repeat operations do carry with them a degree of additional risk, in the context of Mrs Harris’ surgical history, the risks involved in the revision foraminotomy were not appreciably greater than the risks involved in the earlier foraminotomy in the same place.

Mrs Harris subsequently contacted Mr Johnston’s secretary and booked herself in for the revision foraminotomy. The first he knew about it was when his secretary informed him that Mrs Harris was on his list for 3 November 2011. Prior to the operation Mr Johnston saw Mrs Harris on the ward and she signed a consent form. Once again, the form does not record what it should have recorded. Nothing at all is recorded under the headings “explanation of the procedure” or “serious or frequently occurring risks”. Even if there was a valid reason for Mr Johnston’s advice regarding the risks and benefits of having a revision foraminotomy not being recorded in the note of the telephone conversation during which it was given, the advice should have been recorded at the very latest when Mrs Harris came to sign the consent form.

That was poor practice; but the fact that Mr Johnston, at the time, fell well below the standards of record keeping that could be expected of a professional in his position does not mean that he was incompetent in carrying out his surgery. The highest that it can be put is that sloppiness in record-keeping may be symptomatic of a more widespread cavalier attitude towards one’s professional obligations, or of a willingness to cut corners: but that was not the impression that I formed of Mr Johnston.

Mrs Harris’ operation was first on the list that afternoon. Typically, on Thursdays when he sees his private patients, Mr Johnston carries out 4 or 5 operations in the afternoon and evening, devoting the morning to his consultations. The 30 minutes or so taken for this procedure was the usual sort of time taken by Mr Johnston to carry out a foraminotomy. The operation was performed in essentially the same way as the previous surgery in May 2010. On this occasion, however, shortly after Mr Johnston commenced the dissection through the tough fibrous tissue under the skin, instead of encountering the degree of resistance he was expecting, the Cobb dissector entered what he described in the operation note as “a cavity”, and the momentum carried it downwards into the spinal canal, touching the side of spinal cord on the far right.

Mrs Harris no longer has the stimulator in place and so an MRI scan was able to be carried out in October this year. Mr Kirkpatrick told the court that the MRI scan showed that the injury was to the far right of the spinal cord and in keeping with Mr Johnston’s explanation of where the Cobb dissector went.

Mr Johnston’s operation note reads as follows:

“With the patient prone, I localised the C6/7 interspace. I performed the usual muscle strip. I entered a cavity with the immediate egress of a large amount of CSF. I was worried that there might have been a spinal cord injury because the patient had had a laminoplasty. However, I placed the retractors and performed a foraminotomy freeing up the C7 nerve root on the left hand side. I then closed over a vacuum drain with vicryl and nylon. The patient woke with no obvious neurological deficit”.

The final sentence was incorrect, and indeed Mr Johnston’s own summary discharge letter dated 2 February 2012 records that Mrs Harris awoke with neurological deficit. When Mrs Harris came round in recovery, Mr Johnston went to see her, accompanied by his regular consultant anaesthetist Dr Razis (who was present throughout the operation, but neither party called as a witness). They asked her whether she could move her right arm and leg; initially she was able to move the leg, but not the arm. Mr Johnston then told Mrs Harris that he had dropped a blunt retractor down onto the spinal cord during the procedure. He anticipated some temporary bruising to the cord which would resolve within 2-3 months and she would experience weakness in her right arm and right leg in the interim. Dr Razis was more pessimistic about the time it would take to recover. Mrs Harris described Mr Johnston as “devastated” and Dr Razis as “white and distraught”.

Mrs Harris quite understandably focused on the word “dropped” which she understood to mean that Mr Johnston had let go of the instrument. That was not what had happened, and it was accepted in the Particulars of Claim (and by both experts) that this was not likely. Mr Johnston accepted that he may well have said “dropped down” or “dropped into” which is consistent with the mechanism he described.

In his witness statement, dated 14 September 2015, as modified in his oral evidence (denoted in square brackets) Mr Johnston explained what happened as follows:

“I inserted a needle to localise the C7 interspace, checked this by means of an x-ray, and then opened the skin [with a scalpel]. Again, this was difficult because of the very tough and fibrous nature of the scar tissue.

I then used a blunt dissecting tool, known as a Cobb retractor, to exert gentle pressure to pull the soft tissue away from [where] the bone [would have been] so as to be able to get access for the cutting diathermy tool. As I applied the Cobb retractor on the right side, expecting to encounter resistance from the soft tissue as would ordinarily be the case, to my surprise and alarm, I encountered no resistance whatsoever and the retractor plunged straight into a cavity which I would not ordinarily have expected to be there.

The retractor went in by quite some distance, perhaps a couple of inches. I was absolutely devastated by this, and was immediately very concerned at the prospect of some injury having been caused to the spinal cord itself. Given the depth of penetration of the retractor, it was likely that this had come into contact with the cord.

I also noted that a large volume of CSF immediately emerged from the cavity. Retrospectively I wonder whether this may have been from a meningocele which had formed during the earlier surgery, which may explain the volume of fluid which emerged, and also the ease of penetration of the retractor.”

Technically what Mr Johnston is referring to as a meningocele is a pseudomeningocele, which is a sac-like cavity filled with fluid, caused by the leakage of CSF through a hole in the dura into the surrounding soft tissue but not through the skin closure. The shape of the pseudomeningocele is dictated by the configuration of the surrounding tissue, and if there is tough fibrous tissue the CSF will be unable to dissipate. As Mr Macfarlane explained, the fluid will follow the path of least resistance. If there is sufficient accumulation the pseudomeningocele can extend upwards towards the surface of the skin (or to any earlier barrier such as a watertight closure at fascia level). Large pseudomeningoceles may be visible on a high definition CT scan, (indeed, one can be seen on a CT scan carried out on Mrs Harris’s spine in 2012) but small or narrow ones may not show up, though they would on an MRI scan. If the Cobb dissector had pierced a pseudomeningocele, it would have been akin to penetrating a blister with a heavy long-handled spoon.

Mr Woolf was justifiably critical of the operation note, which not only failed to identify the surgical instrument that had caused the injury but failed to explain precisely what had happened, as it should have done, and contained at least one admitted factual inaccuracy (the final sentence). Mr Macfarlane fully supported that criticism. Mr Woolf submitted that in circumstances such as these, the Court should judge the Claimant’s evidence benevolently and the Defendant’s evidence critically, relying on the observations of Longmore LJ in Keefe v The Isle of Man Steam Packet Company Ltd [2010] EWCA Civ 683 at [19]. That was a case in which a breach of duty by the defendant, in that case the employer of the injured man, in failing to keep records of noise levels in the workplace, had caused the employee’s widow significant difficulty in adducing evidence to prove that the noise levels had been excessive and that the employer was therefore liable for causing his deafness.

I agree with the observations of Lloyd Jones J in the subsequent case of Shawe-Lincoln v Neelakandan [2012] EWHC 1150 (QB) at [80] - [82] that Keefe is not concerned with a reversal of the burden of proof, but rather with the weight to be attached to evidence, and the circumstances in which the Court may draw inferences, including adverse inferences. The Court may properly draw an adverse inference, in an appropriate case, from the failure by a defendant to create a record that he was obliged to keep, without which the claimant would have great difficulty in establishing a breach of duty on his part. Whether it is appropriate to draw an inference at all, and if so, the precise nature and extent of such an inference, will depend on the particular circumstances of the individual case.

In the present case, Mrs Harris was under anaesthetic throughout the surgery and would have been unable to give any evidence of what happened to her apart from the evidence that she did give about what was said to her in the recovery room immediately after the operation. She was always going to be reliant upon Mr Johnston’s version of events. The only other eyewitness, Dr Razis, could have been called by the Claimant; whilst one can understand that her legal team may have apprehended that the regular anaesthetist for Mr Johnston may have been reluctant to criticize him, the absence of a fuller operation note did not leave a complete evidential vacuum.

This is not a case of a failure to keep any contemporaneous record, but a case of the record being deficient in detail. Moreover, Mr Johnston’s failure to record the whole history in his operation note does not necessarily place the Claimant at an evidential disadvantage. It would only do so if he had omitted to record something which would point to his behaving negligently, or which would put a different interpretation on the information that was recorded. No-one criticized Mr Johnston for failing to record the precise depth of the cavity that he encountered. I do not accept that the Claimant or her legal team were materially inhibited or disadvantaged in putting her case or challenging his evidence by his failure to identify the Cobb dissector or to state when in the dissection process the pseudomeningocele was encountered.

One has to remember that the starting point in considering this claim is that during the course of an operation in which no surgical instrument should have gone anywhere near, let alone into, the spinal cord, that is what happened. Although this is insufficient in and of itself to prove surgical negligence, it does raise a prima facie case, and the evidential burden (though not the legal burden) has shifted to the defence.

Unlike Keefe, where the employer had a clear motive for its failure to keep any records, and it was difficult to prove liability without them, this is a case where the absence of a fuller explanation of what happened is potentially damaging to the Defendant. If anything, the absence of that further detail made it easier, not harder, to criticize Mr Johnston, and to suggest that his subsequent version of events is an attempt at self-exculpation long after the events would have been fresh in his mind. Such an attempt does not have to be deliberate; an honest witness can subconsciously place a favourable slant on his or her recollection, or indulge in wishful thinking. With that in mind, I have subjected Mr Johnston’s evidence which supplies the detail that was lacking in the note to rigorous scrutiny, but I am not prepared to draw an inference of suppression of unhelpful material, and it would be wrong in principle for me to conclude that Mr Johnston was negligent simply because he omitted certain important details from the operation note.

I found Mr Johnston, like Mrs Harris, to be an honest and straightforward witness. This was the first case in which he had given evidence in court about his surgical competence. Mr Johnston did not seek to defend the serious shortcomings in his record-keeping either generally or on this particular occasion, though he did observe that he was “not in a normal state of mind” when he dictated the 3 November operation note, this being the one and only time that he had good reason to believe that he may have caused an injury to a patient’s spinal cord in the course of a foraminotomy. He described his heart as either pounding or “in his boots”, and said that he was absolutely devastated. This description tallied with Mrs Harris’s own recollection of his demeanour in the recovery room, though as Mr Woolf pointed out, he did manage to compose himself sufficiently to complete his surgical list.

Although Mr Johnston’s detailed account of what happened during the November 2011 operation was not given until his witness statement was served, some 3½ years later, it was consistent with such contemporaneous evidence as does exist, and his evidence in the witness box that he could remember it very clearly, as if it had happened yesterday, had the ring of truth about it. A neurosurgeon would have good reason to remember an event that was so extraordinary and so shocking. There was no hint of any self-exculpatory revision, whether deliberate or subconscious. Indeed, when I asked Mr Johnston about why he did not go back over his operation note when he had calmed down, to see if it was sufficiently detailed and accurate, he gave this very frank answer:

“I thought that doing something in retrospect may, you know, may -- for the purposes of the court I could have made it up. But I wouldn't do that….

I don't do addenda to an operation note, that's not something I do.”

The consistencies between Mr Johnston’s account and the contemporaneous evidence are (a) the express reference in the operation note to a “cavity” and to the immediate egress of a large amount of CSF (b) Mrs Harris’s evidence of the explanation given to her in the recovery room and (c) the symptoms that Mrs Harris experienced, indicative that the injury had been sustained to the far right hand side of the spinal cord, which has been confirmed by the recent MRI scan.

Mr Woolf suggested to Mr Johnston in cross-examination that the “cavity” referred to was the spinal canal, which he denied. A surgeon of Mr Johnston’s experience would know the difference between a cavity and the spinal canal, and if he entered the spinal canal he would have known that was what it was, and said so in his records. In the note he dictated at the time, which on the Claimant’s case is the best and most reliable evidence, he used the word “cavity”. If a cavity was encountered early in the dissection, or indeed at any stage outside the spinal canal, Mr Kirkpatrick had no suggestion as to what else that cavity might have been, if it was not a pseudomeningocele.

Mr Macfarlane told the court that on reading the operation note he understood that what happened was that whilst performing the muscle strip (not after completing it) Mr Johnston unexpectedly penetrated a pseudomeningocele. He said that was the only explanation that made sense. The contemporaneous record, when interpreted by another neurosurgeon, therefore supports Mr Johnston’s account, even though it failed to indicate at what point in the course of the muscle strip this occurred, or which instrument had penetrated the pseudomeningocele.

It is true that Mr Johnston’s evidence was not that he realised at the time that he had penetrated a pseudomeningocele, but that retrospectively he wondered if that is what had happened. In his oral evidence he said that the only other possibility was that the CSF came from the spinal canal and that the Cobb dissector had plunged through soft tissue. That is not the evidence of a person who is trying to cover something up. I find it plausible that at the time, as he said, Mr Johnston’s primary concern was with his patient and with the prospect that she had suffered a spinal injury, and therefore he would not have stopped to think clearly about what it was that caused the unexpected lack of resistance he encountered. He was not prepared to be categoric in his witness statement or in his oral evidence, because he was speculating with hindsight and could not be sure. The contemporaneous operation note is only consistent with one of the two possibilities that occurred to him afterwards.

Mr Johnston’s explanation of what happened is also consistent with Mrs Harris’s surgical history. The experts were agreed that when he first penetrated the skin, Mr Johnston would have encountered, and expected to encounter, tough fibrous tissue (as he did in May 2010). Accordingly, in my judgment, nothing turns on the fact that the operation note in November 2011 makes no specific reference to the difficulty of dissecting through it, as the previous operation note had done. The conditions for formation of a pseudomeningocele were also present: CSF could have leaked through the hole in the dura on the right hand side which was caused by the diathermy tool in May 2010. It would have persisted if either the dura did not heal, or if the cavity was surrounded by hard tissue that was unable to absorb the fluid that leaked out – which was highly likely given the extensive prior surgery in that area. Mr Macfarlane’s evidence was that pseudomeningoceles are fairly rare, and he agreed with Mr Kirkpatrick that they are usually spherical in shape, but he had encountered long narrow ones on more than one previous occasion.

I accept Mr Johnston’s evidence as to the mechanism of injury. There was a long narrow pseudomeningocele which extended from the area of the May 2010 dural tear either up to or very close to the fascia at an angle to the right of the midline. That was the “cavity” recorded in the operation note, and there was no reason for Mr Johnston to have expected to encounter it; he had reviewed the CT scan and the pseudomeningocele was not apparent. He could not order an MRI scan, which would have shown it up. He encountered the cavity at a point in the dissection when he was using the Cobb dissector to retract tough, fibrous tissue and expecting resistance. The width of the cavity was self-evidently sufficient for the Cobb dissector to fit into it.

If, as I find as a fact it did, the Cobb dissector unexpectedly entered the top of a narrow pseudomeningocele with tough fibrous tissue on either side, emanating from a dural leak on the right hand side, it would follow the path of least resistance, leading inevitably to the dura and the spinal cord unless the surgeon had sufficiently fast reflexes to reverse the momentum. That would explain why the instrument travelled down at an angle towards the right and through the dura before connecting with the spinal cord, the consistency of which was described in the course of evidence as being akin to “cold porridge.” Therefore, it would not take significant force to transect it, though fortunately in the present case it was bruised, not severed.

The Defendant’s pleaded case as to the mechanism of injury is clearly set out in paragraph 11(d) of the Defence, in which it is admitted that the Cobb dissector perforated the dural sac via a meningocele, leading to a CSF leak and causing an injury to the spinal cord. That case is nowhere addressed in Mr Kirkpatrick’s expert report, and he did not engage with it adequately in the joint statement either. An expert witness is, of course, entitled to take issue with the factual premise on which a defence rests, if he regards it as implausible or impossible, and to point out to the court any evidence or scientific reasons why it should not make those fact-findings. However, an independent expert fulfilling his duties to the court should also give an opinion based upon the hypothesis that the court does make those fact-findings. In this case Mr Kirkpatrick did neither of those things, but instead based his opinion on a mistaken factual premise.

I accept that a degree of confusion was caused by Mr Johnston referring to the Cobb dissector as a “Cobb retractor” in his witness statement of 14 September 2015, and as a “blunt retractor” when he spoke to Mrs Harris in the recovery room, although he also described it as “a blunt dissecting tool” and the Defence clearly and unequivocally referred to it as a Cobb dissector, which is its proper name. There is apparently also an instrument known as a Cobb retractor which is similar to the McCullough retractor and serves the same purpose as that instrument. Mr Kirkpatrick was well aware that there were two different Cobb instruments, performing different functions, one of which was sharp (the retractor) and one of which was blunt (the dissector).

Whilst Mr Macfarlane appeared to be under no misapprehension as to how the injury is said to have occurred, or what the instrument was that caused it (at least after reading Mr Johnston’s evidence), Mr Kirkpatrick produced an expert report in which he assumed, wrongly, that the instrument that entered the spinal canal was a large retractor with a sharp claw mechanism. He therefore proceeded on an assumption as to how the injury occurred, without any reference to what was actually pleaded in the Defence or the evidence as to how it actually did, ignoring all references to the instrument being “blunt”. This mistake permeated his evidence. In commenting on the Defence in his report, he said that it admitted that a “Cobb retractor” was inserted and entered the spinal canal: but the expression used throughout the Defence is “Cobb dissector” and Mr Kirkpatrick knew that was a different instrument. Mr Kirkpatrick observed: “how one can defend the use of a Cobb retractor as a dissector defies the purpose of the instrument” and he goes on to refer to the purpose of that instrument and its use after the anatomy has been dissected clearly, and to its “sharp claw mechanism”. At the very least this indicates that he did not read the material before him with the appropriate degree of care or ask the questions one would have expected him to ask to obtain clarification.

Question 29 in the joint expert report asked whether, if the Court accepted Mr Johnston’s account as set out in his witness statement, that was an acceptable complication of the procedure. Mr Kirkpatrick’s response was that “to penetrate the spinal canal to the degree described by Mr Johnston when using an instrument designed to retract the musculature outside the spinal canal represents an unacceptable standard of care…” [my emphasis].

Even if I were to be really generous to Mr Kirkpatrick and absolve him from any blame for failing to appreciate prior to the joint experts’ meeting that there was at least some confusion about what the instrument was, he was well aware of the possibility that he had made a mistake when he discussed the joint report with Mr Macfarlane at the experts’ meeting. Mr Macfarlane also drew it to Mr Kirkpatrick’s attention that the “probable mechanism” of injury that he postulated in answer to question 30 (which asked the experts, among other things, whether they considered that another mechanism was the probable cause of the injury to the spinal cord) was neither supported by the evidence of the parties, nor recorded in the contemporaneous medical notes. If nothing else, that should have set alarm bells ringing. Yet Mr Kirkpatrick did nothing to check if his assumption about the instrument that was used was correct. He accepted in cross-examination that he was under a duty to have done so, and that he was in breach of that duty.

His explanation was that he did not regard it as relevant when giving his opinion what the instrument was, because whatever the identity of the instrument the location where it ended up was the same and that was indicative of an unacceptable loss of control. That was tantamount to arguing that there were no circumstances whatever in which this injury could have occurred without negligence. That intransigent mindset coloured his evidence throughout, and it did so in a way which was not helpful to the Court. An expert is of course entitled to hold a firm opinion, but in fairness he is obliged to give proper consideration to any material that might test it or potentially cause him to take a modified or different view. He cannot possibly assist the Court if he does not take appropriate steps to ensure that he is giving his opinion based on the facts that are alleged, rather than on assumptions about the facts which may turn out to be completely ill-founded.

What is all the more surprising about Mr Kirkpatrick’s failure to carry out even the most basic of checks in the face of the defence expert pointing out to him that he could be labouring under a serious misapprehension, is that he had recently been criticised by Her Honour Judge Melissa Clarke in the Oxford County Court for making factual assumptions about key matters in the case (including the nature of the operation that was performed by the defendant surgeon) without taking any steps to check that his assumptions were correct. In consequence, his evidence was given on the basis of a complete misunderstanding, and he had to be recalled after the misunderstanding was demonstrated. The judgment in Kneuss v BMI The Chiltern Hospital and Stuart Blagg [Case No 2YN28735] was handed down on 15 July 2016 and Mr Kirkpatrick confirmed that he had read it before he came to give his evidence in this case.

That judicial criticism was put to Mr Kirkpatrick, very properly, by Mr Davy in cross-examination, but I deliberately refrained from reading the judgment in Kneuss before forming my own views about him. Mr Kirkpatrick denied that he was under any misapprehension about the nature of the operation when he wrote his report in that case, and suggested that the defendant changed his evidence in the witness box. Now that I have read the judgment, I cannot see how that can be an accurate explanation for the mistakes he made, not least because Mr Macfarlane (who happened, coincidentally, to be the opposing expert) had accused him of “poor attention to detail” and said in his own expert report that “all the allegations of negligence made in relation to these matters are based on inaccurate information”. The Judge found those comments to be well-founded. Mr Kirkpatrick also sought at one point in his evidence before me to suggest that the points of difference between the operation he assumed had taken place and the one which actually did take place in Kneuss were immaterial to his opinion, which is a strikingly similar stance to the one he took about the identity of the instruments in this case.

I am less concerned with whether that earlier criticism of Mr Kirkpatrick was or was not justified in this specific respect, than with the fact that, having been subjected to that criticism, he did nothing to avoid being criticised on that basis ever again – criticism which I should say, on this occasion, has ample justification. I would have expected someone of Mr Kirkpatrick’s standing, who regularly appears as an expert before the court in cases of this nature, to have been mortified by that criticism (all the more so, if he felt the criticism was unfair) and therefore to have taken great pains to ensure that he could not possibly be said to have made a fundamental factual error in future, by checking and double-checking. It beggars belief that he would allow this to happen a second time.

The effect of Mr Kirkpatrick’s failings goes far beyond matters of credibility. His fundamental misapprehension as to how the injury was sustained meant that the first time that he ever addressed the Defendant’s case was when he was in the witness box being cross-examined.

One could be forgiven for thinking on reading the joint report that the presence of a (pseudo)meningocele from the previous surgery was not disputed, because in answer to question 20, “was there any feature of the Claimant’s presentation which might have enabled the procedure to be completed more quickly than might typically be the case?” the experts have answered “the presence of the meningocele from the previous surgery may have been of some assistance.” They refer to “the meningocele” not “a mengingocele, if there was one.” However, doctors do not always use language with the same degree of precision as lawyers do, and the Court has to make allowances for that. Mr Kirkpatrick’s position, which he clarified in his supplementary report, served on the second day of the trial (at my direction) was that “I agreed that theoretically if such a meningocele was present it may facilitate muscle dissection and reduce the time for this part of the operation. This was discussed purely in the context of the speed of the operation, and not the mechanism of injury”.

There is no suggestion in Mr Kirkpatrick’s expert report that there was no cavity, as recorded in the operation note. If Mr Kirkpatrick was going to dispute the existence of the “cavity” referred to in the operation note, or of the meningocele referred to in the Defence, I would have expected him to have said so, setting out his reasons, and he did not do so. That was a serious breach of his duty to the Court as an independent expert, and it placed both his own client and the Defendant at a forensic disadvantage. Even in his supplementary report, he said no more than that “I believe this point [about the existence of the pseudomeningocele] to be somewhat academic as the preoperative 2011 imaging shows no such collection of significance”. Anyone reading that might infer that he could be going to say that there was no pseudomeningocele, because if there had been, it would have shown up on the CT scan. Logically, therefore, the “cavity” referred to must have been the spinal canal, and therefore it appeared to be the Claimant’s case that Mr Johnston simply dissected down too far.

That, indeed, was the case put to Mr Johnston by Mr Woolf – but it suffered from the problem that it was highly implausible that a surgeon dissecting tissue in the manner and for the purpose described by Mr Johnston would (a) go beyond the point he needed to with the Cobb dissector in order to insert the blade of the McCullough retractor, (b) breach the dura with the Cobb, rather than the diathermy (c) if he dissected down too far, not realise he had breached the dura before he got anywhere near the spinal cord, and (d) cause an injury in the place where the injury was sustained, rather than directly in line with where he was dissecting. As Mr Macfarlane said, the closer to the midline he was when he entered the spinal canal, the less likely he was to have caused an injury to the lateral aspect of the cord.

In fact, that was not the theory that Mr Kirkpatrick espoused when he finally opined on the matter in the witness box. He came up with two completely new possible mechanisms of injury, both of which Mr Macfarlane considered (with ample justification) to be implausible. The first was that Mr Johnston, immediately after the skin incision, applied the Cobb dissector with so much force that it went through several inches of muscle and/or tissue and into the spinal canal. That hypothesis is self-evidently far-fetched, as it would require huge force to travel that distance through the tissue. As Mr Macfarlane pointed out, the surgeon would then be using a blunt instrument to do something that it was never designed to do, namely, to cut through the thick fibrous tissue, and to do so would undoubtedly be negligent.

The second was that, having initially dissected appropriately through the initial scar tissue for some distance, Mr Johnston then applied too much force to the Cobb dissector, causing it to go through around 2 inches of muscle and/or tissue and into the spinal canal, veering off to the right as it did so. Mr Macfarlane readily accepted that a surgeon who did that would undoubtedly be negligent, as he would still be applying far too much force and using the instrument for a purpose for which it was not intended. However, he pointed out that the further down in the dissection Mr Johnston had progressed, the less fibrous tissue there would be, so there would be nothing to steer the instrument away from the midline towards the right, and on that scenario, the injury was likely to have been closer to the centre of the spinal cord.

This hypothesis was contrary to Mr Johnston’s explanation of his methodology, which was to go even more gently with the Cobb the further down he reached, which is what Mr Macfarlane said he would expect a surgeon to do; and it was also contrary to his evidence that he penetrated the “cavity” early in the dissection process.

The defence (and presumably the Claimant’s own legal team) had no prior warning of what Mr Kirkpatrick was going to say, and therefore the new alternate hypotheses that he advanced were never put to Mr Johnston (there was no application to recall him for that purpose). On that basis alone, I would probably not have allowed reliance to be placed on that evidence to make a positive case as to how the injury occurred; but since I have accepted Mr Johnston’s evidence as to the mechanism of injury and found that there was a pseudomeningocele, there is no need to dwell on this further. Tellingly, Mr Kirkpatrick gave no evidence in support of the proposition that if the Cobb dissector did penetrate a pseudomeningocele, a more controlled use of it would have avoided the injury.

The duties of an expert are set out not only in Part 35 of the CPR but in the Experts’ Declaration in Mr Kirkpatrick’s expert report and in the joint report. He declared that he had done his best in preparing the report to be accurate and complete, but he plainly had done nothing of the kind. I do not consider that he gave proper consideration to the pleadings or the witness statements or to the evidence as a whole. In this case, as in Kneuss, there was poor attention to detail. He really had no choice but to accept that he should have sought clarification as to the nature of the instrument that caused the spinal injury.

Mr Davy made five criticisms of Mr Kirkpatrick against the background that Mr Kirkpatrick had accepted that he was in breach of his duty as an expert witness, namely:

It gives me no pleasure to make such findings about a professional of Mr Kirkpatrick’s standing, but all these points were well-founded. Moreover, his reasoning was unreliable. As I have already observed, Mr Kirkpatrick took a position in respect of his criticism of the first operation that was logically indefensible, by combining three neutral factors to conclude that the operation had been performed incompetently. When each of the planks in his reasoning was removed, he still tried to cling on to the opinion he had originally expressed, just as it appears he did (and was also criticised for) in Kneuss. His general intransigence, his sloppy attention to detail, and his failure to abide by his duties as an independent expert did not just lead me to question his reliability, it left me with no confidence in him.

It is bad enough that Mr Kirkpatrick fell so far of the standards to be expected of an expert witness in this case, but what makes it particularly serious is that he did so against a background where another judge had said, for very similar reasons, that her “confidence in the quality, reliability and impartiality of Mr Kirkpatrick’s evidence has been so severely undermined that I treat it with very great caution”. This is therefore not an isolated aberration.

I am not prepared to rely on Mr Kirkpatrick’s evidence unless it is supported by other evidence, including that of Mr Macfarlane who, by contrast, appeared to me to be the model of an independent and impartial expert, balanced, fair and objective.

Faced with the absence of any reliable evidence from his own expert, Mr Woolf had little choice but to resort to reliance on Mr Macfarlane’s evidence in answer to question 29 in the joint report, which sets out the sole, and limited, basis on which Mr Macfarlane was prepared to accept that this was just a very unfortunate surgical accident rather than a case of clinical negligence. There he said: “if the Court accepts Mr Johnston’s account, namely that he had found the tissues to be “very tough and fibrous” and that he was exerting only gentle pressure to pull the soft tissue away and unexpectedly “encountered no resistance whatsoever” then this did not fall below an acceptable standard of care.”

When Mr Macfarlane was asked by counsel about the degree of force that would be appropriate in using the Cobb dissector for the purpose described by Mr Johnston, he said this:

“These things are all relative, but perhaps “minimal” would be a better word than absolutely nothing at all. I think that in order to use this instrument, particularly if you're within fibrous tissue, you do need to apply a little bit of force. Normally, in someone who's not had surgery before, the tissue is very elastic. We can stretch skin, we can push in muscle, and so you expect what we call "plasticity" in the tissues. But fibrous tissues in people who have had multiple operations, it becomes stiffer and stiffer, in order to expose it you need to exert a little more force than you would otherwise, because otherwise you can't part the tissues.

MRS JUSTICE ANDREWS: Force in which direction?

What he was trying to achieve on the right-hand side, to go down and across, sufficiently to put his retractor in. He didn't need to be all the way down on the dura on that side. But you'd need to be far enough in that the retractor will sit properly, because otherwise, it sort of wants to cant over at an angle, even if you use a slightly shorter blade on that side, so you have to have it in far enough that it's going to support the tissues. So he would be going down and outwards -- if you imagine the lamina is in that sort of shape, and that's the sort of direction that he'd be wanting to go.”

Mr Woolf submitted that I could infer from the fact that the Cobb dissector travelled all the way to the spinal cord that Mr Johnston was using more than gentle pressure and that the force he used was excessive. He pointed out that Mr Johnston knew that he had gone too far in his previous dissection and punctured the dura with the diathermy because there was no protective lamina on the right hand side. Therefore, it was incumbent on him to be even more cautious this time round. I agree, but he encountered the pseudomeningocele before he reached the depth where that lamina would have been.

Mr Woolf next submitted that because the Cobb dissector is a heavy instrument, with a head of approximately 4 x 2 cm, the notion that the cavity encountered should be able to accommodate the entirety of the head without encountering any resistance seemed rather far-fetched. However, that depends on the momentum carrying the instrument downwards, and the depth of the pseudomeningocele. Mr Woolf submitted that it was more likely than not to have been a small pseudomeningocele which was closer to the dura than Mr Johnston suggested, given the size of the dural perforation in May 2010 and the unlikelihood that much CSF would have leaked out before the hole in the dura itself healed naturally. However, a long narrow shape surrounded by tough fibrous tissue is also consistent with those factors and with Mr Johnston’s evidence of what happened. I was entirely persuaded by Mr Macfarlane’s explanation as to how such a pseudomeningocele could have formed in this case without anyone becoming aware of its existence.

I can legitimately infer that Mr Johnston was using a degree of force to pull the tough fibrous tissue to the right to gain access for the cutting tool, but there is no, or no sufficient evidence to justify a conclusion that the level of force was excessive or inappropriate in all the circumstances, bearing in mind that he was not expecting to encounter a pseudomeningocele, certainly not in the place where he entered it, and that he cannot be criticised for not anticipating it. I consider it would be a counsel of perfection to have expected Mr Johnston to have had the speed of reaction in that split second when the instrument met no resistance, to be able to stop the head of the Cobb dissector from plunging downwards in the manner that it did, or to apply sufficient opposing force to prevent it from going as far as it did. It is small wonder both he and his anaesthetist were so badly shaken. It was an accident of a type that no-one could have anticipated in those circumstances and for which Mr Johnston is not to blame.

Mr Woolf also relied on the fact that the operation was being performed at the fast end of the acceptable spectrum, but I cannot see how speed can have had much to do with this incident, especially as it occurred so soon after the dissection commenced. The pseudomeningocele was an unexpected development which Mr Johnston would have encountered however slowly he was progressing down through the layers, and the lack of resistance would have been the same.

I am satisfied that the Defendant has discharged the evidential burden upon him; Mr Johnston’s level of care did not fall below the standards to be expected of a competent neurosurgeon faced with that unexpected scenario.