Welch v Waterworth

Introduction

This is an appeal from the order of 8 October 2013 of His Honour Judge Armitage QC, sitting in the Manchester County Court, whereby judgment was entered for the claimant against the defendant for £155,000 as damages for negligence on the part of the defendant (“Mr Welch”) in a surgical procedure performed by him upon the claimant’s late wife, Mrs Marjorie Waterworth (“Mrs Waterworth”) on 29 February 2008.

On behalf of Mr Welch two grounds of appeal are advanced, with permission granted by Lord Justice Tomlinson, on 27 March 2014. There is also a Respondent’s Notice seeking to uphold the Judge’s order on alternative grounds.

Background Facts

Mrs Waterworth was born on 12 March 1952 and was, therefore, 55 years of age at the date of the procedure. She sadly died on 29 August 2011 from causes unrelated to it. The problem in issue in the proceedings is that, almost immediately after the operation, Mrs Waterworth suffered acute and irreversible failure of both kidneys, requiring her to have regular dialysis from then until her death 3 ½ years later.

The operation with which the court is concerned was a “re-do aorto bifemoral bypass graft” (a “re-do ABFG”). In ordinary terms, the object was to replace an arterial bypass, placed in a previous operation, which had become clogged (occluded) with degenerative plaques/hard calcified disease (atheromatomous material) in the interim period. The earlier operation had been performed by a different surgeon in 1999. This had effected a prosthetic bifurcation in place of the distal end of the aorta and its natural bifurcation into the left and right common iliac arteries serving the legs.

One side of that earlier graft had required remedial attention in 2000. Thereafter, Mrs Waterworth’s condition improved for some years. However, by late 2007 she was experiencing painful stiffness in the legs (claudication) severely impairing her ability to walk more than short distances. She was referred by her doctor to Mr Welch, who is a highly experienced vascular surgeon. The experts in the case (Professor Beard, instructed on behalf of the claimant, and Professor Thompson, instructed on behalf of Mr Welch) were in agreement that Dr Welch was likely to be one of the most experienced vascular surgeons in the United Kingdom.

Mr Welch saw Mrs Waterworth on two occasions before the operation. On the first occasion, on 28 December 2007 from the history alone (according to Mr Welch’s witness statement), he was very confident that the symptoms described were indicative of blockage in the previous graft. After preliminary tests, that view was reinforced and Mrs Waterworth was referred for more accurate and detailed imaging scans. Again, according to Mr Welch’s evidence, those scans confirmed blockage of the graft and of the abdominal aorta “commencing just below the renal arteries”, i.e. the arteries serving the kidneys. At a second consultation on 15 January 2008, Mr Welch advised Mrs Waterworth that surgery was required to replace the old graft in order to avoid further permanent damage by ulceration or gangrene (ischaemia). The operation was fixed for 29 February 2008.

At the outset of the proceedings, up to and including the trial before the learned judge, there was an issue between the parties as to whether or not appropriate consent procedures had been followed by Mr Welch in the consultations. However, that is no longer in issue on the appeal. The judge found no breach of duty in that respect and the point is not pursued on behalf of the claimant in this court.

It is clear that the occlusion of the old 1999 graft had proceeded into the aorta very close to its junction with the left and right renal arteries. In so far as there was any issue as to the precise level of the blockage, the judge concluded that this was a “flush occlusion”, i.e. the blockage had proceeded to the brim of the junction with the renal arteries. The judge recorded that there was in fact no real issue about the level of the occlusion and no point is raised as to his formal finding in this respect.

The Operation and the Judge’s Findings about it

The experts were agreed that the operation would have been technically difficult, being performed close to vessels that had been subjected to surgery on the earlier occasions. The previous surgery was capable of producing fibrosis and scarring which had a tendency to obscure the surgeon’s view of the anatomy. As this was a flush occlusion there was a risk in any event of disturbing the blockage material to some extent into the blood flow and into the mouths of the renal arteries.

In his witness statement (paragraph 30), Mr Welch stated that he was “satisfied” that the renal arteries “were not…occluded during or immediately after the operation”. However, the experts were agreed that the cause of renal failure “was most likely to have been precipitated by intra-operative atheroembolisation”: see their joint answer to question 7 of the questions for their joint meeting as posed by Mr Welch’s solicitors.

The question at trial, therefore, was how the damage to the renal arteries had occurred during the course of the operation.

At its very basic, after obtaining access to the area of the aorta above the site where the old graft had been joined but below the renal arteries (itself difficult in the present because of the residue of adhesions from the primary ABFG) the operation entailed the aorta being transected (cut) to access the blockage and being sutured closed below the cut. The blockage would then be loosened around the circumference of the aorta, causing blood pressure to release the block which pops out of the open top end of the cut aorta (the “champagne cork” effect) which is then clamped closed quickly to prevent blood loss. The top end of the new graft is sutured to the transected aorta. The new graft is subsequently attached to the left and right femoral arteries in sequence.

It was contended by the claimant and found by the judge that the damage had been caused by a “sequencing error” in the procedure in that Mr Welch had clamped the aorta closed, across the material causing the blockage, before the aorta was cut and before the blockage was removed. The result would have been for waste material to be squeezed up the aorta and into the renal arteries – described as a type of “toothpaste tube” effect. All were agreed that if this had been done by Mr Welch it would have been negligent. Mr Welch denied that this was what he had done. His case was that he had merely “placed” a clamp over the aorta, in an open position, before cutting the aorta and removing the embolism. The clamp would then have been ready to close as soon as the blockage was removed. On the appeal, Mr Welch contests the judge’s finding on this issue.

It was also the case for the claimant that Mr Welch had negligently failed to protect the renal arteries, in a manner common in ABFG procedures, by one of two precautionary actions. The first measure would have been the temporary “slinging” of the renal arteries, by inserting the equivalent of a kink into a hose in those arteries for a limited time, to prevent waste material entering the renal system or by clamping higher up the aorta by means of a “supra-renal clamp” or a “supra-coeliac clamp”.

It was accepted that neither of these precautions was taken. However, the judge found that Mr Welch had not been negligent in failing to take either. That finding is contested by the Respondent in the Respondent’s Notice before us.

Mr Welch made two notes of the operation. One was a manuscript note written very shortly after the operation and another typewritten note later in the day at home.

The agreed transcript of the manuscript note (in its material parts) reads as follows:

“Re-do ABG

Transverse abdominal & bilateral vertical groin.

Adhesions +++

Aorta exposed. 5000 units heparin IV, clamps applied.

Left renal vein divided.

Aorta clamped below renal arteries, transected.

Distal end oversewn with 3/0 prolene.

14 x 7mm Hemashield graft anastomosed end to end to aorta with 4/0 prolene.

Limbs tunnelled to groins.

(L)

[left] – End to end anastomosis to femoral bifurcation with 5/0 prolene.

(R)

[right] – end to side anastomosis to hood of old graft extending into CFA with 5/0 prolene.

Excellent flow on clamp removal.”

The relevant part of the typewritten note is in these terms:

“Transverse supraumbilical and bilateral vertical groin incisions through old scars.

Very fibrotic scar tissue in groins; femoral arteries and limbs of previous graft carefully exposed.

Extensive intra-abdominal adhesions carefully divided; no 1 vicryl ties to divided omentum. Dense adhesions divided to expose old occluded aortic graft. Left renal vein ligated with no 1 vicryl and clips and divided to expose the pulsatile bit of the aorta which was then mobilised.

5000 units IV heparin, clamp applied to infrarenal aorta and aorta transected. Distal end occluded but oversewn with 3/0 prolene. Thrombo-endarterectomy of proximal end resulting in superb inflow. 14 x 7 mm Hemashield Dacron graft anastomosed end to end to aorta with 4/0 prolene. Limbs tunnelled retroperitoneally behind ureters to groins.

Clamps applied to femoral arteries; on left the old graft limb anastomosis was completely stenosed; femoral artery transected below this, spatulated into profunda. Excellent SFA and profunda backflow; hep saline flush; graft pulled to length, spatulated and anastomosed end to end with femoral bifurcation using 5/0 prolene. Excellent flow on clamp removal.”

The judge found, as the experts agreed, that the expectation is that the description of operations given in such notes is chronological and that surgical terms used are those commonly accepted. Mr Welch’s evidence was that, although he wrote both notes, neither described accurately what he did. The judge said, “That is most discouraging”. In another part of his judgment, the judge said that Mr Welch’s notes “did not serve him well”.

For the purposes of the principal factual issue in the case, the use of the terms “Aorta clamped” and “clamp applied” in the place in which they appear in the notes, if recording the sequence of actions chronologically, give rise to the inference that the aorta was indeed clamped prematurely. In contrast, Mr Welch was insisting that he did not close the clamp at the moment indicated by the references in the notes; he merely positioned it for speedy closure once blood flow was restored on clearance of the blockage. He identified the reference to “superb flow” as being consistent with the Champagne Cork effect caused by the loosened blockage popping out of the severed aorta.

The judge concluded that Mr Welch did not realise that he had made an error and still did not accept that he did so. The judge found that Mr Welch thought he had clamped a pulsatile (non-occluded) segment and had thus protected the renal arteries. Indeed, the note refers to exposure of “the pulsatile bit of the aorta” before the clamp was applied.

Professor Beard supported this view of events in view of the note and said that even experienced surgeons were vulnerable to such errors. The clamping sequence implied by the notes would be the correct one in operations for non-flush occlusions or for aortic aneurism, which were more common forms of procedure. The judge recorded Professor Thompson’s disagreement as to the reliability of the sequence recorded by the notes, particularly as clamping before removal of the blockage in this case would have been a major error. However, the judge concluded that the reference to a pulsatile section of the aorta in the second note indicated that Mr Welch had been led to conclude that it was safe to clamp across the aorta in that place before proceeding to remove the blockage. In a passage, on which the respondent placed great reliance before us, the judge found,

“A significant reason for accepting it [the manuscript note] as a reliable record of what Mr Welch had done would be his failure, when writing it, to recognise the significance of what he was recording and [sic: of] its reproduction, albeit in slightly different form in the later note.”

For the first time, in his evidence at trial, Mr Welch had said that it would not have been possible for the clamp to be closed over a blocked section of the artery. Professor Beard did not agree and sought to demonstrate the matter by use of a clamp upon a small toothpaste tube (toothpaste being of similar consistency, in his view, to the material in question). The judge did not fully accept the tenor of this demonstration, but his finding was:

“73…..It seems to me that two unknowns intrude. 1. The precise and possibly variable consistency of the atheromatous material and therefore the compressibility of a segment of aorta filled with such material, is not known; 2. Surgeons probably have limited experience of clamping across such material and thus of gauging the inhibiting effect of resistance. A further problem is that if a surgeon tries but fails to compress such material the damage may already have occurred.

74.

I am not convinced that Professor Beard’s demonstration proved more than that compression of an obstructed aorta was practicable. I do not accept it is proof that that could occur insensibly. Conversely, I am not convinced that the additional force necessary to compress atheromatous material, particularly the most recent deposits at the brim, would inhibit a surgeon who had decided to close a clamp. I am satisfied that Mr Welch thought he had elected to apply and compress a pulsatile segment and thus that he was not at risk of extruding material into the renal arteries. I am satisfied that the most likely rational reason for the outcome is that Mr Welch did sequence the clamping of the aorta as described in his notes and that whether that is described as a sequencing error or the compression of the aorta in a position where the radiology demonstrated that it was filled, it amounts to a negligent error which caused the relevant injury.”

It is to be noted that for a significant period after the operation, when Mrs Waterworth’s outcome was known, Mr Welch did not acknowledge that her renal condition had the operation as its source. In paragraph 30 of his witness statement in the action (dated 5/4/12), Mr Welch said this:

“30……..by clamping below the level at which these vessels intersect with the aorta. I cannot explain why the vessels became occluded so soon after the surgery, or why both vessels occluded completely simultaneously, but I am satisfied that they were not so occluded during or immediately after the operation.”

The Appeal and my own Conclusions

In his careful and cogent submissions on Mr Welch’s behalf, Mr Horne recognised that his appeal was against a finding of fact by an experienced judge who had presided over a 5 day trial and had seen the witnesses. He recognised the heavy burden that lay upon an appellant in such a case in seeking to overturn such a finding. However, he argued that the judge fell into error in his assessment of the evidence, in being over persuaded by the operation notes and by failing to address certain salient features of the case which pointed to a different result. In a second ground, it is submitted that the judge failed to give adequate reasons for his findings.

On the first ground, Mr Horne argued that the judge’s error was founded upon his failure to follow the consequences of his own finding that Mr Welch always honestly believed that he carried out the operative steps in the correct sequence. The core of this submission is recorded in paragraphs 18 and 19 of Mr Horne’s skeleton argument thus:

“18.

The Appellant submits that if

(a)

the judge was correct in his finding that the Appellant had not actually followed the appropriate sequence but had instead closed the Satinsky clamp on an occluded segment of the aorta for the first time before the thrombo-endarterectomy; and

(b)

the Appellant had turned his mind to or reviewed what he had done at any time when he had an actual memory of this particular re-do ABFG

the Appellant he would inevitably have appreciated that he was making/had made a sequencing error in relation to the closure of the clamp.

19.

That is because there were a number of ‘alarm bells’ which would inevitably have caused him to appreciate his error.”

It is said that six identified “alarm bells” would have alerted Mr Welch to the alleged error if he had indeed made it. These were as follows:

In paragraphs 16 to 23 of their helpful skeleton argument for the claimant, Mr Melton QC and Mr Wells addressed each of these potential “alarm bells” individually. However, their principal point about them, which for my part I find persuasive, was that none of these formed part of Mr Welch’s own evidence nor were they raised as an issue for discussion between experts pre-trial. The simple fact was that Mr Welch never appreciated the impact of what had occurred in the operation and never suggested that such alarm bells (or any of them) would have operated upon his own mind if he had made the error suggested. The alarm bells first arose as a “tool” (and are none the worse as such) in the cross-examination of Professor Beard. More significant than these alarm bells to my mind, however, was Mr Welch’s total lack of appreciation of what had occurred here, not just intra-operatively but for many years thereafter. As previously mentioned, and as the judge noted, it seems apparent from the operation notes that Mr Welch was oblivious to the consequences of what he was recording, as having been the sequence of the operative actions, even when questions arose about them in the aftermath. In such a mental state, the potential for the sounding of alarm bells is, in my judgment, entirely speculative.

In paragraph 69 of the judgment the judge referred to the number of operations performed by Mr Welch in which the clamping procedure, erroneous in the present case, would have been the correct one. This, and the reference to a pulsatile segment of artery, was evidence upon which the judge could properly conclude that Mr Welch thought that it was safe to close a clamp in that segment. In paragraph 72 of the judgment, the judge found that Mr Welch assessed the relevant part of the artery as being “proximal to the occlusion and thus not giving rise to a risk of pushing atheromatous material into the free flowing renal arteries”.

Whether or not one goes the whole way to accepting the respondent’s submission that, “…the most likely explanation is that the Appellant proceeded on the basis that he was performing a re-do operation for a non-flush occlusion…” (and that might be right), I do accept the conclusion of that same submission, namely that Mr Welch thought he “…was clamping a soft pulsatile piece of aorta”. In contrast, Mr Welch thought it would be impossible to clamp an occluded section. The judge was not satisfied of this or of precisely what the sensation of clamping or trying to clamp such a section would be. However, he did find that Professor Beard’s demonstration proved that “…compression of an obstructed aorta was practicable”.

In my judgment, Mr Welch’s oblivion as to what went on in the operation and thereafter amply justified a finding that his operation notes provided very significant evidence of the clamping sequence that actually occurred. There was entirely adequate evidence for a finding that Mr Welch would not have been susceptible to the alleged alarm bells and the judge was entitled to fasten upon those aspects of the evidence which he found to be reliable pointers to what had actually occurred, without trawling seriatim though every issue, side-issue or speculation that arose on the evidence or in argument: see Biogen Inc. v Medeva Plc [1997] RPC 1 (per Lord Hoffmann at 31-45) and Henderson v Foxworth Investments Ltd. [2014] 1 WLR 2600, (per Lord Reed at [67]).

For these reasons, I would reject ground 1 of the appeal.

Mr Horne recognised, in his oral submissions, that the areas in which the judge did not, as Mr Horne argued, make specific findings of fact, did more to assist the appeal on ground 1 than to stand as an independent ground of appeal. I agree. I have given above my reasons for considering that the “alarm bell” scenario does not address the reality of this operation and the factors that were actually operating upon Mr Welch’s mind at the time, which were perceptively identified by the judge, in particular from Mr Welch’s own operation notes, the significance of which Mr Welch never noticed.

In my judgment, in the circumstances, the judge was not required to address individually each and every one of the “alarm bell” points raised in Mr Horne’s closing submissions at trial. With genuine respect to Mr Horne’s obviously very careful skeleton argument overall, I do not agree with the submission (made in paragraph 49) that the judge’s judgment does not inform Mr Welch why he lost.

I would, therefore, also reject ground 2 of the appeal.

In the circumstances, it is not necessary in my view to address the issues arising on the respondent’s notice as to the alleged error by the judge in failing to find negligent the absence of precautions taken by Mr Welch to protect the renal arteries, independently of any issue as to the sequence of the other steps in the operation. However, I will shortly state my reasons for concluding that the judge was essentially correct in his assessment of this part of the case.

It will be recalled that the areas of criticism of Mr Welch in this respect were his failure to clamp or sling the renal arteries themselves and/or his failure to employ either a supra-renal clamp or a supra-coeliac clamp.

The judge accepted Mr Welch’s evidence (not significantly contested by either expert) that the extent of the scar tissue/fibrosis present after the earlier surgery would have impeded access to the relevant arterial system. There were thus inherent perils in adopting techniques directly to protect the renal arteries.

So far as clamping or slinging the renal arteries was concerned, the judge found that there was respectable contemporary medical thinking, endorsed by both Mr Welch and Professor Thompson, to support the view that there was sufficient difficulty or risk in such an exercise as to point against interference with the renal arteries themselves. Equally, given the narrow spaces for manoeuvre, clamping immediately above the renal arteries also gave rise to significant risks. As the judge found, provided the access was obscure or significantly obstructed there was support in the evidence of both experts for Mr Welch’s view that the risks of supra-renal clamping outweighed the benefits. In the light of the factual findings as to the extent of obstruction to this area, the judge cannot in my view be criticised for his finding in this respect.

With regard to supra-coeliac clamping, in the absence of renal clamping, there was a stark contrast between the experts and the judge’s finding was very short as follows:

“59.

It seems to me that the additional measure, supra-coeliac clamping, raises the same issues as supra-renal clamping. The more distant the proposed clamping from earlier surgery, the less likely is the presence of adhesions. Mr Welch did not consider it, because he was confident that his general approach to this type of operation provided sufficient safeguards.”

The challenge to the finding is summarised in paragraph 64 of the claimant’s skeleton argument in these terms:

“64.

Given the unchallenged evidence regarding the virgin tissues where a supracoeliac clamp would be placed, the Judge’s finding at paragraph 59 (B/B1/a/17/59) (the Appellant did not consider supracoeliac clamping, because he was confident that his general approach to this type of operation provided sufficient safeguards) is not consistent with his correct statements at paragraph 55 (B/B1/a/15/55) (there are no responsible body of surgeons who would not attempt to protect the renal arteries) and paragraph 58 (B/B1/a/17/58) (provided there is sufficiently safe access to the structures, slinging or clamping of the renal arteries or clamping of the aorta above the renal arteries, that is to say that the benefits outweigh the risk of any necessary dissection, one of those precautions ought to have been taken.)”

I can understand why the claimant argues that the judge’s conclusion did not deal fully with the criticism of Mr Welch which was raised on this part of the case. However, Mr Melton QC stated in argument before us, in his submissions on the Respondent’s Notice as a whole, that the height of the claimant’s case on the issue of supra-renal/supra-coeliac clamping was Mr Welch’s stated reliance upon an outmoded view that the technique that he had adopted was reliable because it was one that he had been taught by an eminent teacher of vascular surgery: see the evidence of Mr Welch 18.4.13 at p.141C et seq. It was this to which the judge was referring in paragraph 59 of the judgment. Nonetheless, at p. 144D of the transcript one sees that Mr Welch had accepted that some supra-renal clamping was “do-able”. It was argued for the claimant that, in such circumstances, the risks of such clamping outweighed the risks and to fail to take these steps was negligent.

I have already dealt with the question of clamping in the area immediately above the renal arteries. On the issue of supra-coeliac clamping, in my judgment, Mr Horne’s answer to this issue, in supplement to the judge’s brief conclusion on it, is persuasive. It was simply that Professor Thompson did not accept that it was negligent practice not to use a clamp of this nature. This was because of the risk to blood supply not just to the kidneys but to all the abdominal organs. Professor Thompson was clear that he would not have applied such a clamp in this operation and his view could not be shown to be unrepresentative of reasonable medical opinion. If Mr Welch was negligent in this respect then so would Professor Thompson have been and the latter was not challenged on the balance of risk, which he apprehended, in cross-examination. His view represented one strand of reasonable medical opinion on the issue and, therefore, not to employ supra-coeliac clamping could not have been found to be negligent. I agree.

Accordingly, had it been necessary to do so, I would not have upheld the grounds raised in the Respondent’s Notice.

Proposed Result

For the reasons given above, I would dismiss this appeal.

I agree.

I also agree.