May (A Child) v Lancashire Teaching Hospitals NHS Trust [2009] EWHC 3175 (QB)

[2009] EWHC 3175 (QB)

Case No TLQ/09/0656

IN THE HIGH COURT OF JUSTICE

QUEEN'S BENCH DIVISION

Venue Royal Courts of Justice, Strand, London, WC2A 2LL

Date 04/12/2009

Before

THE HONOURABLE MRS JUSTICE SLADE DBE

Between

Laura MayClaimant

Lancashire Teaching Hospitals NHS TrustDefendant

William Featherby QC and David Kenny (instructed by Rostrons) for Claimant

Margaret Bowron QC (instructed by Hempsons) for Defendant

Hearing Hearing dates: 26-30 th October 2009

Judgment

[1]The Claimant, Laura May, is 16 years of age having been born on 26th November 1993. In June 2004 her mother noticed that the upper right side of the Claimant’s back was misshapen. The Claimant was referred by her General Practitioner to an orthopaedic surgeon who correctly diagnosed that she was suffering from a scoliosis. After further investigations she was referred to Mr Roger Smith, consultant orthopaedic surgeon at the Defendant’s Royal Preston Hospital. He concluded that she had a severe convex idiopathic thoracic scoliosis of her spine. The curvature had progressed from about 50º in June 2004 to 72º in February 2005. On 25th February 2005 Mr Roger Smith performed an operation to try to achieve some correction of the curvature and to prevent its progression and a significant deformity. He used the Mehdian method, and operating from the posterior position with the Claimant lying on her front. The operation involved improving the angle of the spine by the use of pedicle screws, hooks and rods. The spine was manipulated to achieve a correction.

[2]Sadly, when the operation was over the Claimant was paraplegic. She had lost movement in her body and limbs below the level of the sixth thoracic vertebra, T6, in her spine. In this judgment the vertebrae are referred to as thoracic ‘T’ or lumbar ‘L’ followed by a number indicating their position in the spine.

[3]The parties are agreed that the Claimant went into the operation with full use of her limbs, that after the operation she was paraplegic from T6 down and that her paraplegia occurred during the operation. It was further agreed that the Claimant’s dura was pierced during the operation and that this caused a leak of cerebral spinal fluid. It was further not in dispute that an epidural haematoma was seen on the MRI scans performed on 25th and 28th February and 16th March 2005. A large infarction was seen on the 16th March scan. A large syrinx was seen on a scan carried out on 27th August 2009.

[4]The decision as to the event which led to the Claimant’s paraplegia is both difficult and important. By the time of the trial the parties contended that on the balance of probabilities one of two events led to this outcome. Mr Featherby QC for the Claimant submitted that on a balance of probabilities the cause of the damage was misplacement by Mr Smith of a pedicle screw probably at the thoracic pedicle at T6. He considered it likely that the pedicle screw pierced the dura and compressed the arteries in the spinal cord. Miss Bowron QC for the Defendant contended that the likely cause of the damage was manipulation of the spine once all the metalwork of the Mehdian pedicle screw system was in place.

[5]The issues for determination are: a) Whether Mr Smith carried out the operation negligently; b) Whether such negligence, if found, caused the Claimant’s paraplegia.

[6]On the facts of this case these two issues are closely related. As is submitted by Ms Bowron QC rightly on behalf of the Defendant, if the Claimant does not establish that the misplacement of a pedicle screw by Mr Smith led to the damage of the spinal cord and her paraplegia her claim must fail. The Claimant does not assert that Mr Smith was negligent in any material respect if the paraplegia was the result of manipulation of her spinal cord. All the allegations of negligence against the Defendant relate to the failure to prevent or detect damage to the cord by insertion of a pedicle screw or associated instrumentation at T6 or, possibly, T8, T10 or T12. Accordingly I will consider negligence and causation by determining whether the Claimant has established on a balance of probabilities that:(i) her paraplegia was a result of the misplaced insertion by Mr Smith of a pedicle screw at T6 or possibly T8, T10 or T12; if so

(ii) Mr Smith was negligent in any material respect; if so

(iii) Such negligence caused her paraplegia.

[7]The Claimant relied on the evidence of Dr Evans, a consultant radiologist, and of Mr Thompson, a consultant orthopaedic surgeon. Mr and Mrs May also gave evidence for the Claimant. For the Defendant, in addition to Mr Roger Smith who performed the operation, Mr Robert Smith who was then in training and Dr Azavedo the anaesthetist gave evidence as did Tina Armstrong the radiographer. The Defendant relied on the expert evidence of Dr Connolly, consultant paediatric radiologist and Mr Grevitt, consultant orthopaedic surgeon. The evidence developed and the contentions of both parties were refined during the course of the hearing. Some factual issues which appeared relevant at the outset are now to be seen as of little or no importance. Did the Claimant’s paraplegia result from the misplaced insertion of a pedicle screw? The operation

Did the Claimant’s paraplegia result from the misplaced insertion of a pedicle screw?

[8]Mr Roger Smith discussed with the Claimant and her parents the pros and cons of the proposed major spinal surgery to correct scoliosis of her spine. They decided to accept his advice and to proceed. It is not necessary for me to decide whether the family were warned about the risk of paralysis posed by the operation as Mr Featherby QC helpfully made clear at the outset of the hearing that nothing turns on consent.

[9]Mr Smith was an experienced orthopaedic surgeon. Each year he carried out 6 to 7 operations to correct scoliosis. He also carried out other surgery using instrumentation in the spine. He was the only surgeon at the Royal Preston Hospital who carried out operations to correct scoliosis. He would refer cases of congenital early idiopathic scoliosis to other hospitals. He did not consider that the Claimant’s spine had deteriorated to the stage which he considered required referral to another centre. No complaint is made that he should have referred the Claimant to another orthopaedic surgeon. Hers was not a case of ‘cord at risk’.

[10]The Claimant was admitted to the Royal Preston Hospital on 21st February 2005 and underwent the necessary pre-operative assessments.

[11]The Claimant and her family went home and were asked to return on the afternoon of Thursday 24th February prior to the operation on 25th February 2005.

[12]The operation took place on 25th February 2005. Mr Roger Smith was assisted by Mr Robert Smith who was then in training. Dr Azavedo was the anaesthetist.

[13]From the witness statements, peri-operative care notes, radiography records, anaesthesia notes and blood transfusion records it appears that the following material events occurred during the operation.

[14]Surgery started at 9.20am and the first incision was made at 9.25am. A radiographer at Royal Preston Hospital, Tina Armstrong, gave evidence that a request from Mr Roger Smith for an image intensifier for a planned scoliosis procedure was logged on 25th February 2005 at 9.44am. The image intensifier is plugged into a bank containing two television screens. One screen shows the last image taken and previous images can be called up onto the second screen. The camera of the intensifier is on a ‘C’ arm and can be swung into position as required.

[15]Tina Armstrong explained that when carrying out lateral imaging, the C arm could go over the patient. It could go under the bed to carry out anterior/posterior imaging but it would become desterilised. Lateral but not anterior/posterior imaging was carried out during the operation on the Claimant. Ms Armstrong did not know how many exposures were taken. The total exposure time was 24 seconds. If images were needed they could be printed out. They were stored for a time and then overprinted. Nobody asked for the images taken during the Claimant’s operation to be preserved and they no longer exist. There are therefore no records of what and when images were taken.

[16]Mr Rob Smith recalled that pedicle screws were inserted from the lumbar spine upwards so that screws were inserted in the thoracic spine after those in the lumbar spine. On this basis pedicle screws were inserted at first at L2 and L1 and then T12, T10, T8 and T6, although the operation note sets them out in the reverse order. Twelve pedicle screws were inserted although an additional four were recorded on notes as having been used. Pedicle hooks were inserted into T4 and T2 and a transverse process hook on the right of T2.

[17]In his witness statement and oral evidence, Mr Roger Smith described his usual practice in inserting a pedicle screw. When he had located the pedicle he would use an awl to make a hole in the soft pedicle bone and would advance along the hole with a ball ended probe to assess its direction and depth. He would gradually create a hole for the screw. He would feel the discretion of the bone direction and check the hole regularly with the probe. He could be reasonably confident that the hole created is in line with and contained within the pedicle. Care would be taken when placing the screws in the holes as the screws can divert from, or go beyond, the hole. Mr Smith understands that this occurs in about 5% of screws placed.

[18]Mr Smith was confident of accurate placement of screws in the larger pedicles, at lumbar levels and at T12. He used imaging at these levels after the placement of screws. He explained at paragraph 27 of his statement that:

“Once I started to work on thoracic level 10 and up to and including thoracic level 6, the position and depth of the screw hole was checked as it was created and before the placement of the screws. By leaving the probe in place during imaging, I could see the alignment and depth of the hole being prepared. Imaging was then repeated after placement of the screws, to check their position in the lateral plane.”

However in his oral evidence Mr Smith said that he could not be 100% certain that imaging took place when he was placing the probe into the pedicle at T6.

[19]Mr Smith prepared the pedicles at T4 and T2 for insertion of pedicle hooks by performing a laminectomy, trimming the facet joint and the lamina.

[20]Mr Smith inserted pedicle hooks at T4 and T2. He also placed a hook on the transverse process on the right at T2. This usually takes 5-10 minutes. During the placement of pedicle hooks Mr Smith asked Dr Azavedo to start waking the Claimant up for her wake-up test. Dr Azavedo stopped anaesthesia at 12.20pm.

[21]At paragraph 30 of his statement Mr Smith explained:

“A rod was then put in place in the concavity of the curve. I began to correct the curve by gradually moving the hooks and screws along the rod, which applied pressure on the vertebrae to come into reasonable alignment. Laura had a very significant scoliosis and I aimed during this procedure for only a moderate partial correction. The rod on the convex side of the curve was then positioned and compression was applied to the hooks and screws, which were then fixed to the rods. I was satisfied with the corrected alignment of the vertebrae, after both rods had been fixed in position.”

[22]Mr Robert Smith described the manipulation of the spine carried out by Mr Roger Smith as gentle.

[23]The Claimant had tachycardia during the operation. At 11.40am the Claimant’s heart rate rose from 95bpm to 120bpm where it remained at least until recording ceased at 12.10pm. At 12.00 noon the recording of diastolic blood pressure ceased. The Claimant received the following infusions of blood: 300cc at 10.25am, 300cc at 11.00am, 300cc at 11.30am and 250cc at 11.50am. Dr Azavedo records that three further infusions of blood were given of 305cc, 240cc and 300cc although he does not record the time when they were given. From the blood transfusion records it appears that after blood used at 11.50am, the next amount was signed for at 12.25 pm and the following at 13.12pm. Dr Azavedo gave evidence that the final transfusion was of autolagous blood as had been the first three transfusions. Other intravenous fluids were also given. Dr Azavedo gave the Claimant two litres of Hartmann solution. According to his chart during the operation this was given at hourly intervals: 9.28am, 10.30am, 11.30am. He said that these were needed because it was important to achieve a correct balance when giving infusions of blood.

[24]Dr Azavedo gave evidence that he attributed the tachycardia which was noted at 11.40am to what he described as a massive blood loss. He said that blood would not be given immediately on such an event occurring as the results of necessary tests would first have to be looked at. His record shows blood loss of 3.4 litres. Mr Smith said that blood loss of between 2 and 3 litres is not unusual.

[25]Mr Smith stated that as he was placing the rods the Claimant was beginning to wake but not yet moving her arms to commands. He took the opportunity to undertake the planned costotomy which takes about five minutes. Dr Azavedo, the consultant anaesthetist, records that the wake up test took from 12.45pm to 13.20pm. When the Claimant was sufficiently awake to respond to his commands Dr Azavedo found that she could squeeze his fingers. He asked her to wriggle her toes. Dr Azavedo’s contemporaneous note records that there was no movement in the toes and no movement in the lower limbs. Paragraph 13 of his statement of 12th January 2009 states that ‘when asked to wriggle her toes, although there was some movement in one of her limbs, it was brief’. In oral evidence Dr Azavedo said that there was a flicker. He could clearly remember that there was movement in the toes. The distraction (stretching of the spine using the instrumentation) was reduced. Dr Azavedo gave evidence that after Mr Smith reduced the distraction he asked the Claimant to move again. She could not. Anaesthetic was restarted and Mr Smith took the metalwork out.

[26]Mr Smith’s note made at 2pm after the operation reads as follows:

“Wake up test – slight movement (L) Foot, then stopped .˙. all metal work removed.”

Foot, then stopped .˙. all metal work removed.”

Foot, then stopped

[27]In a letter dated 26th February 2005 to a doctor at Alder Hey hospital to which the Claimant was to be transferred on that day Mr Smith wrote:

“…the wake up test showed only minimal L sided movement, so all instrumentation was removed. We experienced no problems with screw or hook insertion, no CSF leak, and the rods were contained to the manually corrected curve. Post-operatively there was only a flicker of foot movement which has diminished and a sensory level at T4/5. ”

In his oral evidence Mr Smith stated that he was relying on the observation of the Claimant’s movements by someone positioned by the Claimant’s feet. That person was usually a staff nurse. It was not his observation. He said that he knew straight away there was not enough movement. He would have waited a while before taking any action. The movement deteriorated over a few minutes. When he did not see movement he thought it absolutely essential to release tension on the cord. This was done by a small turn of a few screws. The Claimant was put to sleep again at around 13.20pm whilst Mr Smith removed all the metalwork and closed the wound.

[28]There is a difference between the contemporaneous note made by Dr Azavedo on the one hand and his written statement and oral evidence on the other as to whether there was initial movement in the Claimant’s left foot. Dr Azavedo explained that neither he nor Mr Smith saw for themselves whether the Claimant could move her toes or feet. Dr Azavedo was by her head and no doubt Mr Smith was standing by the spine area. Mr Smith was relying on information which had been given to him at the time. There is no direct evidence from the person who was observing the Claimant’s feet during the wake up test.

[29]There are no records of the time at which Mr Smith inserted each screw, the pedicle hooks and the transverse process hook. When cross-examined about why for the purposes of providing an answer to questions from the Claimant’s solicitors he had said that the pedicle screws and rods and hooks were inserted between 11am and 12 noon and removed shortly after 12 noon he said that this was an estimate at the time. To judge by Dr Azavedo’s notes this timing was not accurate. He said that he thought he simply got his timing wrong.

[30]All that is certain is that twelve screws, two pedicle hooks and one transverse process hook were inserted between 9.20am and 12.20pm. The rods were then put in place.

[31]Soon after 12.20pm, when Mr Smith asked Dr Azavedo to start the wake up test, the spine was manipulated and costotomy performed. In making his operation note at 2pm, just after the operation, Mr Smith mistakenly lists the costotomy after the removal of metalwork which was after the wake-up test.

After the operation

[32]After the operation the Claimant had no movement below the level of T6. She had an MRI scan of the spine on 25th February 2005. Mr Smith stated that the radiologist, Dr Gunawardena reported that within the limits of this examination, he could not identify the presence of a haematoma causing compression of the spinal cord. However in evidence Mr Smith said that he thought he had seen a shadow haematoma in the cord. Both parties’ expert radiologists were of the view that the image quality of this scan was sub-optimal. By the time of the trial they both agreed that an epidural haematoma could be seen on the scan taken on 25th February 2005 as well as those of 28th February and 16th March 2005.

[33]As a result of the further deterioration and concerns over the Claimant’s condition, it was decided by Mr Smith and Dr Azevado that she should be transferred to the Alder Hey Children’s Hospital. This was done at 09.30am on 26th February. On 28th February 2005 the claimant had another MRI scan of her spine. This showed signal change at T3 and T4 levels. Dr Evans, the Claimant’s consultant radiologist, expert was of the view in his report of 1st June 2009 that there was no specific feature to indicate a cause for the abnormal signal. Dr Connolly, the Defendant’s consultant radiologist expert, considered that ‘The axial T1 imaging confirms the presence of a short segment posterior haematoma (axial imaging of the upper thoracic spine images 13 and 14) but without evidence of compression of the cord’.

[34]The Claimant was an inpatient at Alder Hey for some weeks. On 1st March her mother noticed wetness around her daughter’s back which on 9th March 2005 was confirmed to be a leak of cerebral spinal fluid (CSF). MRI scanning on 16th March showed a very extensive spinal cord infarction which Dr Evans described as extending over ‘possibly 8 contiguous segments of the thoracic spine’ and Dr Connolly as being ‘approximately 3 vertebral bodies long’. On 12th April, the Claimant was operated upon by Mr Dorgan, consultant spinal surgeon at the Alder Hey.

[35]An MRI scan of the Claimant’s spine on 27th August 2009 demonstrated a large syrinx (fluid filled cavity) within virtually the entire thoracic cord.

Expert evidence on the aetiology of the injury

[36]The opinions of the radiologists before trial are set out in a report from Dr Evans dated 8th October 2009 and from Dr Connolly of 6th October 2009. There are also notes of a telephone discussion on 3rd September 2009 to the Claimant’s agenda (CA) and minutes of a meeting of that date to the Defendant’s agenda (DA). The radiologists gave their opinions on the three scans of the Claimant’s spine carried out on 25th and 28th Februaryand 16th March 2005 and the significance, if any, of the scan of 27th August 2009.

[37]The expert radiologists agree that the scan carried out on 25th February was sub optimal. This may have been because of movement. In the meeting to the CA the experts agreed that the scan excluded major extrinsic compression and major trans-section of the thoracic cord. By the time of the trial Dr Connolly agreed that an epidural haematoma was to be seen on all the scans carried out on 25th and 28th February and 16th March 2005. At the trial Dr Connolly explained that the medical term ‘primary cord infarction’ referred to an infarction caused by a vascular event including a constriction of blood vessels by manipulation of the spine. The note of the CA telephone conference of 3rd September 2009 records:

“4. Dr Connolly feels the imaging is most consistent with primary cord infarction. The cord signal abnormality is demonstrated at the T3/4 levels on the T2 imaging. There is no evidence of direct trauma to the cord at the T3/4 level or the T6 level as witnessed by the lack of a significant epidural haematoma, the lack of cord T2 signal abnormality on the initial MRI scan of 25th February 2005 and the lack of evidence of haemorrhage within the cord on the MR scans. Dr Connolly does not feel that the presence of a small epidural haematoma and a CSF leak are direct evidence of instrumentation or [sic] the cord. Rather these appearances are interpreted by Dr Connolly as in keeping with the nature of the surgery undertaken and/or the need for removal of metalwork. Dr Evans feels there is supporting evidence of a preceding penetrative event. Dr Evans emphasised the presence of both epidural haematoma and subsequent CSF fistula as supporting evidence for cord injury preceding infarction.”

The experts agreed that:

“…if direct instrumentation of the cord was to have occurred we would expect to see cord contusion/haemorrhage and also damage to the dura, a leakage of CSF and an epidural haematoma. We agree that on MRI studies subsequent to the date of 25.02.2005 an epidural haematoma (which is non compressive) and fluid (which may or may not be CSF (see 10) were demonstrated. We also agree that extensive signal change appearing on later MR imaging is consistent with cord infarction. ”

[38]The notes of the meeting to the DA record that Dr Connolly:

“...thinks that on a balance of probabilities the patient has had a vascular cord infarct…”

By the time he gave evidence at trial Dr Connolly was of the view that the infarct was caused by manipulation of the spine.

[39]Dr Evans said that most cord infarcts are secondary to an extrinsic event. He said that primary cord infarction is rare and it is rare for these to progress to extensive infarction which was seen in this case. He could only recall two cases of primary cord infarction in 28 years. These were not in adolescents. He said that the formation of a syrinx after primary cord infarction is unusual. Dr Evans was asked about the listing of ischemic infarction as a primary etiological factor of syrinxes in an article by Thomas Milhorat in Journal of Neurosurgery Vol 82 May 1995 at page 809 Table 3. Comment was made on behalf of the Claimant that the primary ischemic infarction there referred to was thrombosis or embolism and not manipulation of the spinal cord as was being alleged by the Defendant in this case. Thrombosis or embolism are not conditions associated with young people.

[40]Dr Evans gave evidence that it is very rare to see epidural haematomas and even rarer to see a CSF leak. The breach of the dura causing a leak of CSF is not necessarily at the level in the spine where the CSF comes to the surface. Fluid takes the line of least resistance, such as through a scar. CSF is unlikely to come to the surface more than three or four vertebra away from where the dura was pierced.

[41]Dr Connolly said that he sees about three primary cord infarcts a year. These were in adults. In seven years he was not aware of seeing paediatric primary cord infarcts. He believed that the most likely explanation for the Claimant’s paraplegia was cord contusion. The cord was not necessarily pierced. It could have been compressed. Dr Connolly interpreted the shadow around CSF on the scan of 25th February 2005 as blood. That was not evidence of compression of the cord. The second scan of 28th February 2005 showed no evidence of compression of the cord.

[42]In his report dated 1st June 2009 at paragraph 4.3 Dr Connolly wrote that personal experience and discussions with other senior spinal colleagues indicates that apparently normal adolescent females with idiopathic scoliosis and normal MRI scans of the spinal cord may nonetheless have vulnerability to spinal cord infarctions. In this case there was no evidence of undue stretching of the spinal cord or precipitate hypotension contributing to the cord infarction. Dr Connolly considered that the most likely location of the breach of the dura causing the CSF leak would be T2 or T3. He agreed in cross examination that CSF was under pressure and if it leaked it would take the line of least resistance. He said that it was possible but unlikely that the breach of the dura causing the leak was at T6.

[43]Mr Thompson’s opinion was set out in reports of 21/10/06, 26/05/08, 09/07/08, 17/08/08 and 08/10/09 and letters dated 14/12/07, 19/06/08 and 02/06/08. Mr Grevitt’s opinion was set out in reports of 01/06/09 and October 2009. There are notes of joint discussions of the orthopaedic experts to the Claimant’s agenda (‘CA’) and the Defendant’s agenda (‘DA’).

[44]Mr Thompson has an honorary contract with Birmingham Children’s Hospital having retired in October 2007. Before his retirement Mr Thompson operated on about seventy scoliosis cases a year. Since his retirement in 2007 the number has dropped to three or five a year. He has had a close professional association with Dr Evans.

[45]Mr Thompson’s first report dated 21st October 2006 contains the following passage in the section entitled “Opinion on Causation and Liability”:

“The operative treatment has resulted in a paraplegia due to a spinal cord injury. Spinal cord injury in patients undergoing surgery for spinal deformity is a recognised hazard and requires special measures to prevent its occurrence. The quoted incidence world-wide is approximately .25% but the statistics may vary. This relates principally to posterior spinal instrumentation. It is believed that manipulation of the vertebral column adversely affects the blood supply of the spinal cord resulting in vascular occlusion and cord infarction. In Laura’s case there appears to have been a violation of the theca in that penetration of the dura occurred and this was manifest in the substantial cerebro-spinal fluid leak, which was evident after transfer to the Alder Hey Hospital.....On the balance of probabilities this occurred during the insertion of or the preparation for the insertion of the pedicle screw at T6.”

It was pointed out by Miss Bowron that in this passage Mr Thompson first stated that the cause of spinal cord infarction was manipulation of the spinal column. This was what the Defendant says occurred in the Claimant’s case. What Mr Thompson wrote later in the passage is inconsistent with that view.

[46]At trial Mr Thompson drew a diagram of what he considered to be the likely path of the pedicle screw at T6 causing a piercing of the dura. The diagram shows the screw passing outside the concavity of the pedicle through the dura and into the spinal cord. Three arteries are shown, two very near the supposed path of the screw. The pedicle is shown as being 4mm wide on the concavity and the screw, 5-6mm in diameter. The cord is 12mm across.

[47]At the time of her surgery the apex of the curvature of the Claimant’s spine was at T8 and T9. The pedicle would be most distorted at that point. It was Mr Thompson’s belief that the pedicle screw must have impinged on, distorted or compressed the spinal cord resulting in the death of the cord. It remained his view that this occurred at T6. It was his opinion that the damage to the Claimant was done by some compression of the spinal cord rather than a penetrative event. Mr Thompson said that the cord will tolerate some displacement. There was definitely no trans-section of the spinal cord.

[48]Mr Thompson explained that the spinal cord would be compromised by 20 minutes of pressure. Before that time it would not be permanently damaged. He said that the tachycardia from 11.40am evidenced damage to the cord.

[49]Mr Thompson considered that on the balance of probabilities the piercing of the dura occurred during the insertion of or the preparation for the insertion of the pedicle screw at T6. It was his opinion that the piercing of the dura was much less likely to have been caused by insertion or removal of a pedicle hook, as suggested by Mr Grevitt. Mr Thompson said that he had never used pedicle hooks. He was surprised that laminectomies were not recorded if they were used in inserting the hooks. He said it was very unusual to pierce the dura when placing a pedicle hook. When it was suggested to him on behalf of the Defendant that because the CSF leak was in the area of T2-T3 it was likely that the piercing occurred there he said that the penetration of the dura could be remote from where the CSF came out from the skin.

[50]Mr Thompson agreed that if there had been a tear of the dura this would be obvious to the surgeon. While clearly reluctant to comment, Mr Thompson agreed that a small nick or pin prick in the dura possibly may not be obvious to a surgeon.

[51]Mr Thompson stated that the presence of a large syrinx shown on the scan of 27th August 2009 was indicative of trauma damage to the Claimant’s spine. When asked about the relative figures of trauma and ischemic infarction as causes of syrinx Mr Thompson pointed out that adults were more likely than children to have vascular problems which were the type of ischemic infarction referred to in the Milhorat paper. In his opinion it is likely that a pedicle screw penetrated the space below the dura and pressed on the spinal cord, although Mr Thompson agreed that this was pure conjecture.

[52]Mr Grevitt gave evidence for the Defendant. Since 1998 he has been operating on scoliosis cases. He explained that the Harrington method was used in the past to correct scoliosis. Harrington rods were more rigid than those used in the Mehdian method in operating on the Claimant. Using the Harrington rods the spine was stretched or distracted significantly with higher risks, than the milder manipulation of the spine using the Mehdian method.

[53]Mr Grevitt commented that if a screw had intruded into the spinal cord as suggested by Mr Thompson’s diagram it would have had to occupy one half to one third of the cord. He considered it highly improbable that arteries would have been compressed. He observed that the ligaments supporting the cord are quite intolerant. To nudge aside the cord without doing it damage was highly improbable. A surgeon using the free hand technique would feel the difference between pedicle bone and the spinal cord. Mr Grevitt had conducted a search of the literature to find a documented incidence of a cord injured by a screw. He could not find any. All the papers showed screw misplacements beyond the pedicle but outside the cord. These did not cause paraplegia.

[54]Mr Grevitt thought that the theory that damage was caused by compression of the cord was fanciful. Probing with an awl to make a track for the screw one would feel graduated resistance. The corticle margin is of tough bone. You would notice that the implement was plunging into soft tissue when rotating the implement 360 degrees around the inside of the bone. The surgeon would feel a softness on the inner aspect of the probed track.

[55]Mr Grevitt took issue with the suggestion that a gradual compression of the spinal cord would cause tachycardia. He said tachycardia was more likely to be caused by a percussive injury. When the spinal cord is struck, the heart rate goes up but then it goes on strike. There is a decrease in blood pressure and bradycardia.

[56]Mr Grevitt considered the extent of the infarction to be extremely important in determining what event had occurred. With manipulation, the infarction would extend over a large area. If the damage were localised the cord would swell and the bleeding could propagate. However a localised infarction would not spread to the extent seen on the scan. Mr Grevitt’s evidence of the difference in spread of infarction depending on its cause was based on his experience of trauma cases and the one case on which he operated where there was a very extensive infarction from a non traumatic event.

[57]Mr Grevitt pointed out that there was no radiological evidence of damage to the spinal cord from a misplaced pedicle screw.

[58]In his report of 1st June 2009 Mr Grevitt referred to three possible causes of the Claimant’s paraplegia. He dismissed malposition of instrumentation. In his then opinion the cause of the Claimant’s paraplegia was interrupted blood supply to the spinal cord (infarction). He stated that the paralysis arising from cord infarction has a characteristic temporal sequence. There may be some initial lower limb movements which then cease. As for the third possible cause Mr Grevitt stated that manipulation damage is unusual but if the scoliosis is bad even a mild manipulation puts a strain on the spinal cord. He did not believe the presence of the syrinx indicates the cause of the paralysis.

[59]By the date of the joint statement of 27th August 2009, the opinion of Mr Grevitt was that the third hypothesis was the most likely. He said that:

‘the most probable cause for the Claimant’s paraplegia is a spinal cord infarction caused indirectly by the required manipulation of the spine resulting in disruption to the blood supply to the cord at numerous levels.’

[60]After exchange of reports Mr Grevitt confirmed the finding of Dr Evans that the CSF leak occurred by the piercing of the dura during the operation. He said that inserting a pedicle hook in T4 was difficult and could cause perforation of the dura. The dura could have been perforated by preparation for insertion or insertion of the hook itself. Removal of pedicle hooks may have caused the CSF leak. He now considered that perforation of the dura and consequent CSF leak could have occurred during the placement or removal of a pedicle hook.

[61]Mr Grevitt commented on the syrinx seen on the scan of 27th August 2009 in paragraph 3 of his report of 16th October 2009:

“3.2… the extent of the syrinx coincides with the very extensive amount of spinal cord infarction seen on the previous scan at Alder Hey hospital. 3.3 These changes of the syrinx formation extent, in my opinion, favour spinal cord infarction, arising from the necessary manipulation of the spinal cord during the corrective manoeuvres to straighten the scoliotic spine, as the cause of the Claimant’s paraplegia.”

Conclusions

[62]The tragic outcome of the operation on the Claimant was obviously distressing and concerning to Mr Roger Smith and Dr Azavedo. The doctors taking part in the operation on the Complainant did their best to recollect what happened. However Dr Azavedo, a careful and experienced anaesthetist, inexplicably ceased recording the Claimant’s heart rate at just after 12 noon and diastolic blood pressure just before 13.00pm. His contemporaneous note records that there was no movement in the Complainant’s toes or lower limbs on the wake up test. When he gave oral evidence he said that there was movement initially. No explanation was given for this difference. Mr Smith’s operation note made at about 2pm just after the operation was not accurate in that it lists the costotomy after removal of the metalwork and therefore after the wake up test. This was plainly an error as the costotomy was carried out before the wake up test.

[63]As is understandable and commendable in such a difficult case the experts modified their opinions when they had reason to do so.

[64]Immediately after the operation Mr Smith no doubt realised the seriousness of the situation yet did not call for the radiography record of when images were taken during the operation nor ask for hard copies to be retained. He explained that he thought a static film existed of the procedure as stated in the Defendant’s Solicitor’s letter of 19th January 2007 but this was not in fact the case. T6 was near the apex of the curvature. Progressing up the spine the insertion of pedicle screws becomes increasingly difficult which is why Mr Smith would usually use the imaging intensifier during rather than after placement of the screws from T6 upwards. Mr Smith was commendably frank in saying that he could not be 100% certain that he used it at this stage in the operation.

[65]It is not clear that the damage to the Claimant’s spine occurred at 11.40am when there was a rise in her heart rate as is suggested on behalf of the Claimant. I accept the evidence of Dr Azavedo that the Claimant was loosing more blood than is usual in such an operation and that if this occurs it is to be expected that the heart rate increases.

[66]There is no evidence as to the time when Mr Smith inserted pedicle screws in T6. However by reason of the order in which the metalwork was inserted and the work which was carried out by Mr Smith after the screws were inserted at T6 it is likely that this would have been some time before the manipulation. The manipulation must have taken place at about 12.20pm when Mr Smith asked Dr Azavedo to stop giving anaesthetic so that the wake up test could be carried out.

[67]The evidence of the doctors was that the manipulation of the Claimant’s spine was gentle. On the expert evidence the risks from manipulation of the spine using the Mehdian method, as in this case, were less than using the Harrington method.

[68]Mr Smith relied upon a nurse to tell him whether the Claimant’s left foot initially moved at the beginning of the wake up test. The nurse was not called to give evidence and I was not told whether a statement had been taken from him or her. Those involved in the operation must have appreciated that there was likely to be some investigation if not litigation as a result of the tragic outcome for the Claimant. Mr Smith did not check for himself whether there was movement in the Claimant’s foot and Dr Azavedo’s contemporaneous note records that there was none. There is doubt as to whether the Claimant had some movement in her left foot at the start of the wake up test. By reason of my conclusions based on other evidence I do not make a finding of fact as to whether the Claimant has established that there was no movement in her foot at the beginning of the wake up test.

[69]Trans-section of the spinal cord would cause irreversible immediate loss of movement in the lower limbs. On the evidence of Mr Thompson I find that compression would compromise the spinal cord after about twenty minutes.

[70]Scans taken on 25th and 28th February and 16th March 2005 showed some epidural haematoma but no evidence of piercing of the dura.

[71]As for whether the location of where a CSF leak reaches the surface indicates where the dura has been pierced, in my judgment the proposition advanced on behalf of the Claimant that CSF finds the path of least resistance to make its way to the surface so that it can emerge through a healing cut some way from the puncture was not convincingly challenged. The suggestion that this could not have occurred because Mr Thompson thought it likely the Claimant was lying on a bed does not affect my conclusion. No doubt the Claimant would have been moved into different positions from time to time.

[72]The literature contains examples of penetration of the theca or dura by pedicle screws (Diab Spine Vol 32 Number 24 2007 p2759). Lenke in Spine Volume 29 Number 3 2004 records at page 340:

“For thoracic pedicle screws used in the treatment of spinal deformities, the incidence of screw misplacement ranges from 3% to 44.2%, with screw-related neurological complications in the 0% to 0.9% range. A few reports have described complications caused by overpenetraton related to the placement of thoracic pedicle screws with major visceral injury. Although many studies reported medial wall violation of the thoracic pedicle between 1.4% and 14% from 1mm to 8.0mm, there were no permanent neurologic, cardiovascular, or pulmonary complications associated medial wall violation in any cases. ”

Suk records in Spine Volume 26, Number 18 p2053 that one patient suffered neurologic injury ‘because of medial perforation of the pedicle by the screws causing delayed epidural haematoma.’ There were three dural tears as evidenced by gushing out of CSF. Diab Spine Vol 32 2007 at p2762 recognises that there may be as much if not more of a safety margin medial to the pedicle than lateral, where the ‘great vessels and viscera’ being ‘impacted by an extraneous screw tip.’

[73]Diab notes at p2759 that factors associated with neural complications include ‘type of procedure, curve magnitude, type of instrumentation’. The ‘highest rates of neural injury are seen in distraction and sublaminar wire instrumentation.’ Distraction is the term used of the more vigorous stretching of the spine using Harrington instrumentation. The gentler stretching using the Mehdian method is referred to as manipulation.

[74]The two competing theories as to how the damage was caused to the Claimant’s spine each have unusual features. The decision as to how the damage occurred is difficult and finely balanced. I have taken into account against the Claimant’s proposition that the damage to her spine was caused by a pedicle screw the following factors:(a) The literature gives accounts of misplaced screws outside the pedicle but not inside the dura causing permanent damage to the spinal cord.

(b) The MRI scans did not contain any indication of penetration of the dura by a screw.

(c) Penetration of the dura by a screw 5mm in diameter would have made a large hole and occupied ½ or ⅓ of the spinal canal.

(d) It would be unlikely for the spinal cord to be compressed but not torn.

(e) No CSF leak was seen until many days after the operation.

(f) The CSF was seen coming from the Claimant’s skin at the area of T2 and not T6.

(g) In his first report dated 21st October 2006 Mr Thompson refers to the incidence of spinal cord injury and cord infarction by manipulation of the spinal column.

(h) The size of the infarction indicated generalised damage to the cord not local trauma.

(i) It is possible that a syrinx could be caused by ischemic infarction which was not thrombosis or embolism.

(j) Mr Smith is an experienced surgeon who at the time carried out five or six scoliosis operations a year.

[75]However on all the evidence I conclude that on a balance of probabilities the Claimant has established that the damage to her spinal cord was caused by a pedicle screw penetrating the dura at T6 and compressing the spinal cord. I reach my conclusion for the following reasons:(a) Mr Smith did not use intra operative imaging at the time of preparing for the insertion or inserting the pedicle screw at T6. The pedicle at T6 is smaller than the lower pedicles and there is a greater chance of misplacement of the screw. The greater the curvature of the spine the greater the difficulty in achieving a correct placement.

(b) The Mehdian method and not the Harrington method of correction of the spine was used. The Mehdian method carried less risk of damage on manipulation of the spine than distraction using the Harrington method.

(d) The pedicle screws were released during the wake up test. If the paraplegia were not caused by the misplacement of a pedicle screw at T6 but by manipulation of the spine it is somewhat surprising that the reduction of the tension in the rods did not prevent paraplegia. If the Defendant is correct in saying that some slight movement was seen in the Complainant’s left foot at the beginning of the wake up test, having regard to the fact that insertion of the rods took place shortly before the wake up test, it is somewhat surprising if release of tension in the rods would not reverse or halt any neurological damage.

(e) It is more likely that the CSF leak and the damage to the spine resulted from one unusual event, the penetration of the dura by a screw, than from two unusual events, the piercing of the dura by a pedicle hook or other instrumentation and unconnected damage caused by manipulation of the spine. (d) The syrinx seen on the scan of 27th August 2009 was more indicative of traumatic compression injury to the spinal cord than of damage caused by manipulation.

Was Mr Smith negligent in any material respect?

[76]The parties were agreed that the test to be applied to determine whether Mr Smith had been negligent is that set out in Bolam v Friern Hospital Management Committee

[1957] 1 WLR 583

“I myself would prefer to put it this way, that [a medical practitioner] is not guilty of negligence if he has acted in accordance with a practice accepted as proper by a responsible body of medical men skilled in that particular art. ... Putting it the other way round, a man is not negligent, if he is acting in accordance with such a practice, merely because there is a body of opinion who would take a contrary view.”

Anterior/posterior imaging Contentions of the parties

Anterior/posterior imaging

[77]The Claimant contended that Mr Smith was negligent in that he failed to use bi-planar imaging. In his written closing submissions Mr Featherby QC submitted that the ‘C’ arm of the image intensifier available to Mr Smith could have been used to obtain anterior/posterior images. Mr Smith was responsible for selecting the operating table. He selected a Gardner frame which was radio-opaque whereas he could have used a radio-lucent frame and table which would have enabled images to be taken from an anterior/posterior angle. Anterior/posterior imaging would have shown whether the screw was going too far sideways and thus impinging on the spinal canal. In his written closing submissions Mr Featherby contended:

“(e) The only reason Mr Smith gave for not using AP imaging was that he found the images taken in that plane “difficult to interpret”. (i) If that is the case, he should not have attempted the surgery. He should have referred it to someone who could. (ii) Mr Smith did not say why he found the images difficult to interpret beyond saying that it was because the spine was rotated. If the vertebrae were not symmetrical to the AP plane (and assuming a radio-lucent table was used), it was only a matter of slightly adjusting the aim of the columnator to ensure that true AP images of the vertebrae were obtained. Mr Grevitt agreed this was possible.” (i) If that is the case, he should not have attempted the surgery. He should have referred it to someone who could. (ii) Mr Smith did not say why he found the images difficult to interpret beyond saying that it was because the spine was rotated. If the vertebrae were not symmetrical to the AP plane (and assuming a radio-lucent table was used), it was only a matter of slightly adjusting the aim of the columnator to ensure that true AP images of the vertebrae were obtained. Mr Grevitt agreed this was possible.”

[78]In her written closing submissions for the Defendant Miss Bowron QC contended that imaging is used ‘on an as required basis during the operation to provide some assistance with that which the surgeon cannot assess by eye or expert feel’. The Kim paper refers to imaging after making holes and insertion of screws and not while this is being done. Mr Grevitt and other scoliosis surgeons used and taught the use of lateral imaging alone. Miss Bowron contended that the practice of Mr Grevitt and the University Hospital Queens Medical Centre Nottingham establishes that there is no reasonable basis upon which it could be found that not using anterior/posterior imaging was negligent applying the Bolam test.

Discussion

[79]It is to be noted that whilst Mr Grevitt and his colleagues at Nottingham used lateral imaging alone they did have the safety net of spinal cord monitoring (‘SCM’) which was not provided by the Defendant.

[80]As Mr Smith wrote in paragraph 27 of his statement of 30th March 2009 from T6 upwards it was his practice to check the position and depth of screw holes as they were created and not just after placement. This was because it was more difficult to insert screws at these levels that at lower levels of the spine.

[81]The reason that he did not use anterior posterior intra-operative imaging was not because it was not usual to do so or because it was not available at Royal Preston Hospital but because he found the images difficult to read.

[82]The Kim paper records at page 340:

“Thoracic pedicle screw fixation is potentially dangerous because of the maximum permissible translational error of less than 1mm and rotational error of less than 5° at the normal midthoracic spine using a geometric model due to a small pedicle diameter and little space between the spinal cord and medial pedicle. There are several methods of thoracic pedicle screw insertion to enhance the safety, such as guide pins into the pedicles, intraoperative C-arm image intensifier, direct visualization of the medial wall after laminotomy, and image guided systems based on CAT scan or fluoroscopy. ”

At page 341 the authors comment:

“Our results document that we have been able to create a safe method of thoracic pedicle screw placement without use of these other intraoperative methods/devices, but we acknowledge that this method may not be the best for many surgeons. Surgeons must use their best judgment for creating the safest environment as possible when placing thoracic pedicle screws.”

[83]The Kim paper makes it clear that the free hand technique may be safe for some surgeons to use. However Mr Smith was performing only five or six scoliosis operations a year. He recognised that inserting pedicle screws at T6 and upwards was increasingly difficult. He therefore used fluroscopy whilst placing the pedicle screw.

[84]Lateral imaging can only show the depth of the screw. Anterior posterior imaging shows the lateral position of the screw.

[85]The evidence establishes that whilst it may be safe for some surgeons not to use imaging during the placement of screws, the actions of the reasonably careful surgeon will depend upon his skill and experience. Mr Smith prudently recognised his need to use imaging during placement of screws from T6 upwards. Having regard to the gravity of consequences for the patient from a misplaced screw even if he used lateral imaging I find that Mr Smith was negligent in not using both lateral and anterior/posterior imaging during the placement of the pedicle screw at T6 having regard to the fact that unlike other centres including Nottingham where Mr Grevitt practiced, SCM was not available at the Royal Preston Hospital.

Spinal cord monitoring

[86]It is contended on behalf of the Claimant that the Defendant was negligent in that Mr Smith performed the operation without using SCM.

[87]The most commonly used form of spinal cord monitoring is SSEP (somato-sensory evoked potentials). The benefit of SCM is that it provides continuous assessment of the function of the spine so that if compromise is suspected or detected the surgeon can and should investigate. On behalf of the Claimant it is said that false negatives, (and, presumably false positives) are no reason for not using SCM. It is a matter of balancing risk and benefit.

[88]Mr Smith said in cross-examination that before the operation on the Claimant he had written twice to his colleagues in neurophysiology asking for SCM equipment to be provided. He had not mentioned this in his statement. No such letters were found despite enquiries made by the Defendant during the trial.

[89]There are many references in the literature to the use of SCM. The note of the joint conference of 27th August 2009 records that Mr Grevitt did not know whether there were any units in England which did not use spinal cord monitoring in 2005. Mr Grevitt’s colleagues at Nottingham produced a paper (Spine Vol 25 Number 19 Papastefanou) in which the authors comment at page 2471:

“The cost of monitoring is fully justified, as compared with the financial implications of even a single case of postoperative paraplegia, to say nothing of the human cost to the patient and family.”

Mr Grevitt said that the now universal use of SSEP is defensive medicine.

[90]On the evidence it appears that the SCM was available and no doubt used for carrying out spinal operations at most if not all centres apart from the Royal Preston Hospital. Applying the Bolam test in my judgment the Defendant Trust fell below a reasonable standard of care in failing to provide SCM for use in the operation on the Claimant. If SCM had been provided it is clear from his having previously asked for it to be made available that Mr Smith would have used it.

Planned wake-up tests during surgery

[91]The Claimant’s case is that, in the absence of SCM, there should have been wake-up tests at earlier stages of the surgery. It is said that the obvious times would have been:(i) during the insertion of the screws in the mid-thoracic vertebrae (because that is where the danger of misplacement is at its greatest);

(ii) after the insertion of the screws but before manipulation.

[92]Mr Grevitt did not know of any report which describes the experience or accuracy of multiple wake up tests after repeated insertions of pedicle screws. He questioned its practicability. A recent poll of scoliosis surgeons revealed that no-one had heard of or seen the use of multiple wake-up tests.

[93]I am not satisfied on the evidence that Mr Smith was negligent in failing to plan wake up tests as contended on behalf of the Claimant.

Did the negligence of the Defendant cause the Claimant’s paraplegia?

[94]Having found that the Claimant’s paraplegia was caused by compression of her spinal cord by the misplacement of a pedicle screw inserted at T6 I find that on the balance of probabilities the failure of Mr Smith to use bi-planar imaging together with the failure of the Trust to provide SCM caused the Claimant to suffer damage to her spinal cord which resulted in her paraplegia. The use of bi-planar imaging during the preparation for and the placement of the pedicle screw at T6 would have been likely to have prevented its misplacement, the piercing of the dura by the screw and its compression of the spinal cord. If SCM had been provided, even if the screw had pierced the dura, compression of the cord may well have been detected and swift action could have been taken. On the evidence that infarction of the spinal cord by compression takes about 20 minutes, paraplegia may thus have been prevented. Accordingly I find that it was the Defendant’s negligence in Mr Smith failing to use bi-planar imaging and the Trust failing to provide SCM which together caused the Claimant’s paraplegia.

Conclusion

[95]The Claimant has established that the Defendant was negligent and caused her paraplegia.