Sellu v The Crown

On 5 November 2013 in the Central Criminal Court before Nicol J and a jury, David Sellu, a consultant specialising in colorectal medicine, including surgery, was convicted by a majority verdict (10:2) of the manslaughter by gross negligence on 16 February 2010 of James Hughes (then aged 66 years); he was acquitted of perjury in relation to evidence given to the coroner about the death. He was sentenced to 2½ years’ imprisonment. No appeal either against conviction or sentence was then mounted but detailed further consideration of the case has led new counsel (instructed by different solicitors) to pursue a wide-ranging appeal. Having granted an appropriate extension of time, the appeal proceeds by leave of the full court (Hallett LJ, McGowan and May JJ): see [2015] EWCA Crim 1980. In the event, none of the counsel in the appeal appeared at the trial.

On 5 February 2010, Mr Hughes underwent elective total knee replacement surgery at the Clementine Churchill Hospital, a private hospital, under the care of a consultant orthopaedic surgeon, Mr Alan Hollingdale. Prior to the operation a number of standard checks were made on him and nothing unusual or concerning was discovered. Although there is now an issue as to Mr Hughes’ general health, at the time, he was deemed to be quite healthy for his age. He was assessed as low risk for the general anaesthetic that was required for the operation.

The operation, which was routine, seemed to go well. Following the surgery, Mr Hughes remained as an in-patient at the hospital and was prescribed painkillers, an antibiotic (Cefuroxine), and an anticoagulant called Dabigatran. It is not suggested that Mr Sellu should have been aware of the administration of the anticoagulant at the relevant time and its effect was not considered either by any of those involved in Mr Hughes’ treatment at the time or, indeed, by any of the expert witnesses at trial.

By Wednesday 10 February, Mr Hughes was deemed well enough to be discharged but he elected to remain at the hospital until the following Monday in order to have his stitches removed. Early in the following morning, his progress was not maintained because he complained to the nursing staff for the first time of having a pain in his abdomen and of not having opened his bowels for four days (that is, since the operation). A note was made asking for him to be reviewed by the Resident Medical Officer, Dr. Georgiev (“the RMO”). He had qualified in Bulgaria and first came to England in January 2010; he had been working at this hospital for three weeks. He was on duty from 8:00 am on 11 February on call for 24 hours a day for seven days (with a bed in the hospital where he could sleep when not needed).

Dr Georgiev saw the deceased at some time after 8:00 am, prescribed him Buscopan, and ordered blood tests, the results of which were available by 10:30 am. Dr Georgiev analysed the results and found nothing significant but, throughout the day, the pain experienced by Mr Hughes in his abdomen became increasingly severe (noted as 9/10 at 3:00 pm). Mr Hughes spoke to various medical staff but when no effective action was taken, adopted the unusual step of seeking to contact his general practitioner, Dr Lauder (with whom he was friendly). He spoke to Mrs Lauder (herself a retired G.P.) told her he was in “terrible pain” and that he could not make himself understood to the doctor. He then spoke to Mrs Hollingdale who acts as her husband’s secretary: she was taken aback by the call. That any of this was necessary is obviously a cause for concern.

As a result, Mr Hollingdale saw Mr Hughes at 6:00 pm. He was in evident pain and, upon examination, had a tender abdomen. He ordered an x-ray examination which was marked ‘urgent’ by a nurse. The x-rays were first viewed by a Dr. Shah and she rang Dr Georgiev to inform him that the x-rays showed that there was free air in the abdomen and she suspected that this was due to a perforation of the stomach or bowel. Mr Hollingdale asked for further blood tests (which it does not appear were taken). In passing, it is appropriate to add that Dr. Shah’s written report on the abdominal x-ray, although not the chest x-ray, referred to her finding of free air in the abdomen but was not verified (following the appropriate practice) until 15 February and was thus not available on the hospital’s computer system until after Mr Hughes had died.

A free perforation of the stomach or bowel can be distinguished from a localised event and is the result of a burst diverticulum resulting in the contents of the bowel leaking out into the abdomen. This would cause pain in the general abdomen. The body would attempt to combat this by secreting a liquid to dilute what constitutes poison. However, there then existed a fluid which was a good culture for bacteria coupled with the possibility of further faecal matter being pumped out through the perforation in the colon. As the bacteria increased in number they would be absorbed into the bloodstream producing a generalised sepsis. This in due course would result in septic shock during which the blood vessels become dilated, blood pressure falls, the circulation around the major organs falls, ultimately leading to multiple-organ failure. Left untreated this was the course that a free perforation would follow.

Treatment for the condition should have included the administration of broad-spectrum antibiotics and optimisation of the patient’s condition by effective resuscitation and source control (i.e. repairing the source of infection by surgical intervention).

Because this development did not fall within the expertise of an orthopaedic surgeon, Mr Hollingdale referred Mr Hughes to Mr Sellu, who had practised at Clementine Churchill Hospital since 1997 and been a consultant surgeon at Ealing NHS hospital since 2000. The prosecution alleged that Mr Hollingdale (who was angry about what had occurred and complained to senior nursing staff the following morning) would have provided the history to Mr Sellu; Mr Sellu, however, said that he was unaware of the calls to the GP or Mr Hollingdale, or of any difficulty in communication between Mr Hughes and Dr Georgiev.

Mr Sellu examined Mr Hughes at about 8:30-9:00 pm on 11 February, his notes indicating that he was aware of the length of time during which the deceased had been in pain and describing him as being unwell. He was recorded as having a slightly increased breathing rate and that there were abnormalities with the abdomen upon examination. The readings for his C-reactive protein levels were high, which was indicative of inflammation. The blood test results that the appellant reviewed had in fact been taken that morning. There was no documentary record of any blood gas test results that evening.

Mr Sellu had not previously worked with Dr Georgiev (who accepted that he may have described himself as a fully trained surgeon who specialised in abdominal surgery) and both viewed the abdominal and chest x-rays between 9:05-9:45 pm. Mr Sellu wrote in the records: “? free gas under the right hand diaphragm” and recorded his opinion as “? Perforated [viscus]”. He left a management plan that included intravenous fluids, blood tests, and a CT scan for the following morning that he marked “urgent”. Radiologists would have been available to perform an emergency scan if the appellant had so directed it. The deceased was left in the care of Dr Georgiev.

Of critical importance in the subsequent analysis of what happened, no antibiotics were administered to address the possibility of abdominal infection. The medical notes made no mention of the appellant having ever given instructions for the deceased to be placed on antibiotics. Mr Sellu’s first statement to the coroner was silent as to the administration of antibiotics. In his answers to the hospital’s internal investigation, he stated that he instructed the RMO (Dr Georgiev) to give antibiotics following a taking of bloods and later when they spoke on the phone he gave instructions to continue with the antibiotics. In a further statement to the coroner he stated that he “would have asked the RMO” to begin treatment with antibiotics. In both his subsequent oral evidence to the coroner and in police interview he unequivocally stated that he had discussed antibiotics with Dr Georgiev and given instructions that these were to be commenced after the blood tests: Dr Georgiev denied that he had ever been given any such instruction. In his police interview, Mr Sellu did, however, accept that ultimately it was his responsibility to ensure that the patient had received antibiotics and he had failed in that regard.

On 12 February, upon the RMO’s instructions, Nurse Sarota telephoned the appellant at about 6:00-6:30 am in respect of Mr Hughes’ condition which had seemingly deteriorated through the night. His urine output was low, there was substantial dark brown fluid output from his nasogastric tube, his abdomen was distended, and he was administered intravenous painkillers given his on-going pain.

The earliest time at which the appellant next possibly examined Mr Hughes briefly was at 10:00 am. It was disputed at trial as to whether this examination took place at all as there was no record of such a visit in the medical notes. Mr Sellu remained consistent in his account that this visit did occur. The recorded observations at this time showed the deceased to have a high pain rating, an increased respiratory rate and low blood pressure and urine output.

The CT scan was not performed until 11:20 am as the imaging department declined to treat the matter with the urgency requested by the ward sister. The significant results, which confirmed a perforated colon, were communicated by the radiologist, Dr. Kantor, to Dr Sellu at 1:45 pm. Mr Sellu did not see the actual scans until 9:51 pm that evening. From 1:45 pm onward, it became common ground that Mr Hughes required major abdominal surgery to prevent his condition progressing to fatal septicaemia. The timing of such surgery was a careful balance between the degree of resuscitation required before surgery and the risks of further deterioration due to further delay.

Mr Sellu then took steps to secure the attendance of an anaesthetist and continued his own engagement with a number of elective surgical procedures on other patients. The hospital did not have a rota of anaesthetists on call. There was, however, a policy in place allowing emergency surgery to be interposed between elective cases, albeit this was an infrequent occurrence. There were periods throughout the day when one or more operating theatres were available for the necessary time period. At about 3:30 pm the appellant was informed that Dr. Viswanatha, the anaesthetist who would eventually assist, was involved in another procedure which the appellant was told was due to finish at no later than 19:00. As it happened, Dr Viswanatha’s earlier operation overran significantly and did not finish until 8:50 pm, at about which time he ordered that the deceased be brought down to theatre.

In the interim, during the afternoon, nurses on Mr Hughes’ ward became increasingly concerned about his condition and took the view that he should be moved to the intensive care ward and that an emergency operation should be performed. An attempt was made to contact Mr Sellu. The Director of Nursing was contacted and she visited Mr Hughes at about 2:40 pm and noted a deterioration in his condition as indicated by his observations. She understood that an operation was planned for 6:00 pm.

Mr Sellu saw the deceased at 4:00 pm so as to obtain his consent for the operation. He then had an outpatients’ clinic between 5:00 and 7:20 pm. During the evening, Mr Hughes’ condition which was calibrated against the Modified Early Warning Score (“MEWS”), which was used to rate the condition of the patient (the higher the score the worse the underlying condition) dropped from four to three before rising again to four at 7:00 pm. The nurse on the ward had been instructed to notify Intensive Care Unit (ITU) if the score stayed at four, but this he failed to do.

Mr Hughes was not admitted to theatre until about 10:00 pm. Post-anaesthetic, the signs were that the deceased was going into sceptic shock but the decision was taken to proceed with the operation nonetheless. It is important to underline that neither that decision nor the surgical performance of Mr Sellu, formed part of the allegation. Neither was criticised. The operation finished at 4:30 am on 13 February but Mr Hughes never regained consciousness and, tragically, he died at 3:38 pm on the following day, 14 February. The cause of death then provided was “1(a) multiple organ failure; (b) faecal peritonitis; (c) perforation of the diverticulum”.

Before parting from the surgery, it is important to add that Mr Sellu reported that, during surgery, there had been macroscopic signs of cirrhosis of the liver. No post-mortem was performed and thus there was no histological evidence as to the degree of cirrhosis. The presence of cirrhosis falls to be considered as a risk factor for any surgical intervention, although its existence or extent has been the subject of argument on appeal.

Not surprisingly, there was an internal hospital inquiry (the ‘Empey’ inquiry) and, on 18 October 2010, an inquest: Mr Sellu gave evidence to both. As a result of his answers at the inquest, the coroner suspended the hearing and referred the matter to the police. In police interview the appellant answered the questions put to him. There were various inconsistencies of detail between his different accounts, not least in the impression he had given to the coroner that he had viewed the CT Scans at lunchtime on 12 February (although he was acquitted of the subsequent allegation of perjury). These investigations led to the subsequent prosecution.

In short, the prosecution alleged that Mr Sellu’s care of Mr Hughes fell very far below an acceptable standard and that a series of missed opportunities and serious errors of judgement on his part combined to cause the deceased’s premature death. There were a catalogue of breaches of the duty of care but there were three specific failings on the part of the appellant alleged to be grossly negligent and which it was said substantially contributed to Mr Hughes' death. These were as follows:

The principal case advanced on Mr Sellu’s behalf was that the actions that he had undertaken were reasonable in all the circumstances and did not amount to gross negligence. In short, it was contended that Mr Hughes’ death was not caused by substandard care.

The issues for the jury as articulated by Nicol J were, first, whether the appellant had behaved negligently, in that no reasonable consultant colorectal surgeon in his position would have behaved as he did and, if so, whether that negligence was gross or severe; and, second, whether any gross negligence caused or significantly contributed towards the deceased’s death.

In the light of the way in which this appeal has relied upon fresh material not placed before the jury and the other complaints about the trial, it is appropriate to summarise the evidence and, in particular, the medical and expert evidence that was deployed during the trial. First, as a matter of background, it is right to underline that Dr Whitehead, a consultant anaesthetist, said that she had done many operations with Mr Sellu including complex, high-risk cases. He was the first choice for extremely sick patients with emergency problems in the abdomen. His decision making was extremely sound and very professional. He was calm, emotionally controlled, and acted in the best interests of the patient and the medical team. The agreed evidence of Dr Ali, another consultant anaesthetist, was also to the effect that he was a faultless surgeon who exercised a high degree of care.

Reverting to Dr Whitehead, she explained that Mr Sellu had contacted her about operating on Mr Hughes but she told him that she was not available after 7:00 pm. She would have been willing to operate that afternoon but that was not proposed. She made efforts on his behalf to identify an available anaesthetist and eventually Dr. Viswanatha put himself forward. She informed Mr Sellu that Dr. Viswanatha would be available from 7:00 pm.

Dr Whitehead also dealt with the question of interrupting an elective list for an emergency operation and said that it was possible, but, in practice it was harder to achieve. It would require the surgeon on the emergency operation speaking to the elective surgeon. She had known examples where the latter had tried to dissuade the former from interrupting their list, but she had never had a case where the surgeon had flatly refused.

Dr Georgiev, the RMO, said that as he and Mr Sellu looked at the x-rays together (between 9:05 and 9:45 pm on 11 February), he told Mr Sellu that in his opinion it was most likely that the deceased had suffered either a stomach/duodenal ulcer or a perforation of the bowel. Dr Georgiev said that he kept asking the appellant whether Mr Hughes was to be given antibiotics but was told that he was not to be given antibiotics that night. He did not remember any further telephone conversation with Mr Sellu that night and specifically denied that he had been given any instruction to start administering antibiotics. He did not remember having any conversation with the appellant about blood gas test results.

Dealing with the other evidence touching the administration of antibiotics, Mr. Hollingdale could not recall whether or not Mr Sellu had said anything about prescribing antibiotics in their discussion that evening; as far as he was concerned, Mr Hughes was now Mr Sellu’s responsibility and it would not have been appropriate for him to question his expert judgement. Nurse Sarota, who was on duty on the night of 11-12 February said that she was given no instructions by Mr Sellu, or anyone else, to give Mr Hughes antibiotics: her concern as to his condition led to her contacting Mr Sellu early on the morning of 12 February. It is worth adding that any consideration of the notes that night or the following morning would have revealed that the precaution of administering antibiotics had not been taken.

As for the other tests undertaken, Dr. Bharti Shah, the radiologist who reported, said that she could tell from the x-ray that there was free air in the abdomen and all the signs were telling her such. There was no doubt in her mind as to her conclusion. This meant to her that there was probably a perforation somewhere along the gut.

Ms. Wilkins, the ward sister on duty on the morning of 12 February, said that she made efforts to stress to the imaging department the urgency of the obtaining a CT scan in respect of Mr Hughes, but they declined to prioritise him. She said that there would have been nothing to prevent Mr Sellu from contacting the department himself had he thought that the scan request was particularly urgent and some consultants took that step.

In the event, when the CT scan was undertaken, Dr Kantor, who compiled the report, said that he gave it the highest category of urgency. His findings (of perforated colon) were significant and he wanted a minimum of delay. Having compiled the report, however, at 1:45 pm, he happened to chance upon Mr Sellu and informed him verbally of his findings. Given the new evidence now sought to be adduced, we add that the CT scan revealed no abnormality with the deceased’s liver.

The question of the availability of an anaesthetist was the subject of detailed evidence. Jan Hale, the hospital’s Executive Director, explained that if Mr Sellu had difficulties in arranging an anaesthetist, he could have spoken to the theatre manager who would have had a list of possible anaesthetists and would have started ringing around to locate one who was available. There was a nationwide policy to break into an elective list if it was necessary to interpose an emergency operation; this policy, it was contended, would have been well known to all consultants. Mr Hollingdale confirmed that there had been several occasions in his twenty years of experience when elective surgery lists had been broken to accommodate an emergency operation.

This was also the evidence of Dr Wrigley, the consultant anaesthetist who ultimately enrolled Dr. Viswanatha to take part in Mr Hughes’ operation. He said that he would have acted that afternoon had he been asked and, had the request ever been made, he and Mr Hollingdale would have postponed their elective list.

In the event, when Dr. Viswanatha was approached, he told Dr Wrigley that he would not be available until after 8:00-9:00 pm because of an earlier commitment. When Mr Sellu called between 6:00 and 6:30 pm, he repeated that he could only help out after he had finished his earlier engagement (which he did not specify). Mr Sellu agreed that was acceptable.

Before embarking on his other commitment, Dr Viswanatha visited Mr Hughes and, given his condition, described the operation as an ‘emergency’. Concerned that the operation needed to be performed as soon as possible, he went to find Mr Sellu and told him, although also he expressed the view that Mr Hughes’ condition was not dangerous. He suggested that Mr Sellu find another anaesthetist given that he was about to embark upon another operation. Mr Sellu commented that as it was Friday night it would be difficult to find a replacement and he would therefore wait for him to finish and would carry on with his paperwork in the meantime. We confirm that there is not the slightest criticism of the way in which Mr Sellu conducted the operation when it was finally undertaken.

We turn to the expert evidence which was a crucial part of the prosecution case. In summarising that evidence, it is important to underline the large number of references to gross negligence (as opposed to some other formulation of the extent to which the experts considered that Mr Sellu had fallen short of his duty of care to Mr Hughes). Without a full transcript, it is impossible to confirm how these assertions came to be made, but it appears from the notes that we have seen that, at least in the main, prosecuting counsel asked the leading question about the topic then under discussion: was this gross negligence? We shall deal with the extent to which these questions were appropriate and the impact likely to flow from the way in which the case was summed up, as part of the ground of appeal concerned with the direction on gross negligence manslaughter.

Returning to the evidence, the prosecution called Mr Michael Kelly, a colorectal surgeon. He said that Mr Hughes’ presentation was absolutely typical of a free perforation of the colon. The available diagnostic tools were blood tests, x-rays, and CT-scans. If a patient had free gas then he needed an operation. The finding in Dr Kantor’s report that there was free air both in the abdominal cavity and behind the back lining of the abdomen was a sign that the situation was very bad. The operation carried out by the appellant (a laparotomy combined with a Hartmann’s procedure) was one of two possible alternatives. It was a challenging operation for the patient and it was desirable to get the patient in as good a shape as possible prior to the operation, a process known as resuscitation or optimisation.

He said that treatment with both aerobic and anaerobic antibiotics was necessary. If that had been mentioned in the medical notes, then the nurses would have been alerted to the instruction and to the fact that they were not being delivered. Furthermore, he believed that it was important to involve the Intensive Treatment Unit (“ITU”) from an early stage and that the failure to do so was very bad practice: any reasonable surgeon would have done so. The minimum required was to let the ITU know of the case and make a joint plan.

Mr Kelly thought that the x-rays probably showed free gas. Given the circumstances, he said that Mr Sellu should have spoken to a radiologist as soon as the initial x-rays were available so as to try to confirm a diagnosis. If a CT Scan was considered necessary, then it should have been done within four or five hours; it was not a good idea to wait until the following morning. The finding of free gas in the abdomen strongly suggested a generalised, rather than localised, perforation. The x-rays would have looked quite different had it been the latter.

Mr Kelly considered that there were a number of deficiencies with the treatment notes made by Mr Sellu. It was grossly negligent of him to have failed to put in place a plan with goals, which would have picked up on the deceased’s deterioration, rather than simply leaving him in the care of nurses. He also should have gone into hospital to see Mr Hughes when contacted by the nurse on the morning of 12 February as the low urine output was a sign that resuscitation was not working.

At the point where Mr Sellu claimed to have seen the deceased in the morning he would have (or should have) looked at the various charts. His view was that from the recorded levels of urine output, nasogastric fluid output, and respiratory rate, no reasonable surgeon could have concluded that Mr Hughes’ condition had not deteriorated overnight. As for the antibiotics, it would have been reasonable for Mr Sellu to assume that the RMO had carried out any instructions to administer antibiotics, but only until it became clear that the deceased was not prospering at which point he should have checked.

On the morning of 12 February, the results of the blood tests that morning were such that any reasonable surgeon would have commissioned immediate blood gas tests, contacted ITU and made a plan for an operation. It was grossly negligent not to have done so. If the results of the blood tests that had been done were not available to Mr Sellu then he should have chased them up. Furthermore, Mr Sellu’s failure to demand that the CT scan be done that morning was grossly negligent. The CT scan result report, coupled with the blood test results, was such that the need for an operation was urgent. The appellant’s failure to both view the scan and to go and see the deceased himself were grossly negligent.

Dealing with the difficulty in arranging another anaesthetist, Mr Kelly considered that it would have been perfectly possible for Mr Sellu to have done so as a matter of urgency. It was not necessary to use a general bowel specialist, if such a category of anaesthetist even existed. Mr Sellu should have broken into an elective list. Finally, failure to review the observations noted in respect of Mr Hughes at 16:00 on 12 February would have been grossly negligent. Had Mr Sellu prepared a summary at this point he might have realised that Mr Hughes was not receiving antibiotics.

Dr Kelly also dealt with the extent to which the lapse of time had affected the risk to Mr Hughes’ life. He referred to a statistical tool which is the Portsmouth predictor equation for the Physiological and Operative Severity Score for Enumeration of Mortality and morbidity (colloquially known as “P-POSSUM”). He calculated mortality risk as approximately 4% in the early hours of 12 February, rising to around 11% by 10:00 am as a consequence of the failure to administer antibiotics. In his view (exercising clinical judgment) the mortality rate increased to about 25% by 2:00 pm and between 5:00-6:00 pm it would have risen to about 50%. At the time of the actual operation, he believed that the chances of survival were about 1%. The use of P-POSSUM is criticised and the subject of an application to admit fresh evidence.

In summary, Dr Kelly’s overall view was that on the night of 11 February and the following day the appellant embarked upon “a bizarrely slow, laidback, and inadequate treatment and diagnosis regime which, if proposed by a candidate for a basic doctor’s exam would have resulted in a fail”. Because of his gross negligence he had missed quite a wide window of opportunity, which ended around 4:00 pm on 12 February, to perform a standard lifesaving operation. His performance amounted to recklessness and gross negligence that had a decisive influence on the deceased’s death.

We turn to the second expert called by the prosecution, Dr Dominic Bell. He is a consultant in critical care and anaesthesia. He said that there was a balance to be struck between optimising the patient prior to the operation, which was desirable, and the need to bring the soiling and toxicity in the abdominal cavity under control. The balance between the two objectives was not precise and was a matter for the professional judgement of the surgeon and anaesthetist. Treatment with both aerobic and anaerobic antibiotics was necessary.

He also said that there was an important role for the ITU which was not limited to admitting the patient to the unit. It was established practice for treating doctors to involve the ITU as soon as they identified features of a critical illness. An early intervention increased the chances of a positive outcome. He considered that an operation should have been performed within 2 to 4 hours of the finding of free gas.

He said that a reasonably competent practitioner would have undertaken an assessment of the patient on the morning of 12 February including a review of the observations. If such a meeting did take place, the failure to note it was a serious departure from good practice. Asked how he characterised these failures, he said that Mr Sellu either did not understand the implications of his findings or, if he did, had failed to take appropriate action in a timely fashion. Either way, he had been grossly incompetent.

Dr Bell also used the P-POSSUM statistical tool and assessed the mortality risk at 2.6% on the morning of 11 February, around 9% in the early hours of 12 February, and rising to 15% by later that morning. Dr Bell took the view that the tipping point (i.e. when the chances of survival dipped below 50%) occurred at about 3:00 pm on 12 February although expeditious action even as late as 8:00 pm may have saved his life.

Dr Bell denied that there were any signs that Mr Hughes was particularly susceptible to the effects of the perforated bowel and the fact that decompensation took place as late as it did showed that he had considerable physiological reserves. There were significant errors of judgment on multiple occasions. The deviations from good medical practice, comprehensive assessment and documentation fell below the level of a reasonable practitioner on multiple occasions.

Turning to the defence case, Mr Sellu described his background and experience which we have set out above. He gave evidence that Dr Georgiev had not proffered an opinion as to what could be seen on the x-rays, although it was likely he had mentioned that there was a suspicion of free gas under the diaphragm. The observations he saw recorded on the evening of 11^th February did not cause him any alarm, assuming that the blood tests he had looked at had been taken that evening and not in the morning as was in fact the case. In his view, at that time there was merely a suspicion of free gas.

Mr Sellu agreed that he could have spoken to a radiologist about their views of the x-rays. He did not do so: even if he had contacted a radiologist the treatment would have been the same, as he would still have wished to embark upon the optimisation of the deceased and resolve any uncertainty by way of CT scans. He considered that his findings on the evening of 11 February were consistent with localised perforation and peritonitis which was not inevitably associated with widespread contamination of the abdominal cavity if untreated. It was not uncommon to encounter patients who had suffered abdominal pain for the same period as the deceased.

Mr Sellu said that he had spoken with Mr Hollingdale that night about prescribing antibiotics and had instructed the RMO, Dr Georgiev on the night of 11 February to start the deceased on a course of antibiotics once the blood tests had been completed. He accepted that any failure to recommend such a course at that point would have been a serious error and recognised that ultimately the responsibility was his.

He did not set any goals or parameters for the resuscitation of the deceased as the MEWS system provided adequate means for deciding when senior medical staff needed alerting to a patient’s condition. Although it was not in the notes, he told nurses that frequent, regular checks should be made on Mr Hughes. As for an out-of-hours CT scan, he had delayed the CT scan on the night of 11 February because Mr Hughes first needed to be resuscitated, which itself could take up to twelve hours or longer. There was no indication that an operation needed to be done that night. He marked the request for a CT scan as urgent and expected it to be done first thing in the morning.

He had also instructed the RMO to obtain blood gas results. He spoke to Dr Georgiev about the results and was told that they were normal. He did not notice the following day that there were no such results on file. He was also told by the RMO that blood culture samples had been taken. That evening, he had assessed that surgery the following day was a possibility. With this in mind he made a number of calls to an anaesthetist at Ealing Hospital to obtain his help in finding an anaesthetist for the following day, but nothing came of it. He did not ask Dr. Whitehead that evening whether she could act the following day.

He spoke of the telephone call from Nurse Sarota at 6:00 am on 12 February and said that he was simply informed of a decline in urine output and the aspiration of some brown fluid, neither of which gave him cause for concern. He did not recall being told that Mr Hughes had deteriorated and the level of concern that was communicated to him did not make him worry. He did not consider it necessary to attend at the hospital straightaway and that decision would not have been different if he had been given a full summary of the night’s observations.

Having said that, he did visit Mr Hughes on the morning of 12 February and had a specific memory of doing so, although he was unable to recall details of the visit such as what observation and test results he saw. He was at fault for not recording the visit as he had been rushing to another commitment. He was still awaiting the results of the CT scan before altering the deceased’s management plan; he assumed and trusted that Dr Georgiev had administered the antibiotics.

Initially in evidence, Mr Sellu said that Mr Hughes’ condition had improved that morning but he subsequently clarified that he simply meant that it had not deteriorated. He did not speak to anyone who had been responsible for his care during the night. He did not recall Mr Hughes complaining of being in pain. A single parameter, such as the observed increase in respiratory rate, did not mean an overall deterioration. The deceased’s readings at this point in time showed that resuscitation was not complete. Contrary to his account in police interview, he did not see the RMO that morning and he did not request or review any further blood tests.

Contrary to his recollection when he gave evidence to the Empey investigation, he had not attempted to find an available anaesthetist on the morning of 12 February as he was still waiting for the outcome of the CT Scan. When he became aware that the CT scan had not yet been performed, he contacted the imaging department and, upon learning of the CT scan results he realised that Mr Hughes required an operation. However, he thought it would be reasonable if surgery took place within 4 to 6 hours, i.e. between 6:00 pm and 8:00 pm, and that it was no more urgent than that.

As he did not have an anaesthetist available Mr Sellu could not make a firm theatre booking. He had previously raised his concerns about the absence of a rota of anaesthetists at the hospital. He tried contacting a number of anaesthetists, including Dr. Whitehead with whom he was due to perform an operation that afternoon, but none were available. He did not try to break into the list of scheduled operations as he did not think it was clinically necessary and he was trying to arrange anaesthetists for 7:00 pm as he believed the operation could wait until then. That decision had nothing to do with his pre-existing appointments and he was able to reach his conclusion without seeing the deceased or the blood test results.

He had a preferred list of anaesthetists who he knew he could rely upon and some were specialised in colorectal surgery. He had never seen the hospital’s ‘break-in’ policy. He was aware that this was something that could be done but only if the operative need was immediate and in his view the deceased did not fall into that category. He had tried unsuccessfully in the past to liaise with theatre staff and ITU to find a suitable anaesthetist. He denied that there was an obligation to set goals or parameters for the deceased. The function of the MEWS scores was to do just that and very few surgeons would write out goals.

Mr Sellu disagreed with the suggestion that he was obliged to involve the ITU. In the NHS, patients such as the deceased were managed on wards. Dr. Kantor’s report did not contain anything significant over and above what he had been told orally. He thought there was no significance in the finding of free air in two different places.

When Mr Sellu saw Mr Hughes at 4:00 pm, he did not notice any change in his overall condition. He denied that he had failed to review the observations noted in the records. He commented that although urine output and blood pressure were slightly down, the readings overall did not show significant deterioration. He recognised that the drop in both his white blood cell count and readings relating to the blood’s clotting ability were both significant indicators that his body was being overwhelmed, but they did not indicate the need for an immediate operation.

Mr Sellu was clear that his time frame of 7:00 pm remained reasonable. Mr Hughes needed more fluid resuscitation and did not require an immediate operation. Although isolated indicators, such as the very high c-reactive protein figure, showed deterioration, there was not an overall deterioration in his condition. He agreed that the assessment he made at this time was important because he was planning an operation. He did not check whether the deceased was receiving antibiotics.

Around 7:30 pm, he learnt that the anaesthetist, Dr. Viswanatha, would not be available until later than originally anticipated. Although he made one unsuccessful attempt to obtain the services of an alternative anaesthetist, there was in fact no practical possibility of finding an alternative anaesthetist or of transferring Mr Hughes to a NHS hospital. Had there been any further decline in the deceased’s condition he would have expected the nurses to contact him but they did not. When he viewed the CT scans he could see no difference between the scan and the report of Dr. Kantor that he had read at 4:00 pm.

Mr Sellu emphatically denied deliberately trying to mislead the internal hospital investigation or the coroner and any errors in his earlier accounts were simple misunderstandings.

Mr Sellu’s account was supported by the expert evidence of Mr Peter Sagar, also a consultant in colorectal surgery. Dealing with the first meeting between Mr Hughes and Mr Sellu, Mr Sagar agreed that, although they could have been delayed until after the blood samples were taken, treatment with antibiotics was necessary within a few hours, regardless of whether there was a free perforation or a localised infection. The consultant needed to specify the types of antibiotics to be prescribed.

He disagreed with the experts called by the prosecution and said that it would not be usual to involve the ITU for a case such as that of Mr Hughes. He considered that the quality of the chest x-ray was poor and there was no convincing image of free gas. One of the two abdominal x-rays might show free gas, but it was subtle and Mr Sellu could not be criticised for failing to spot it. If there was free gas then a perforation was the most likely explanation. It was not mandatory to get a radiologist’s opinion immediately. The right thing to do was begin hourly observations.

Although a CT scan might be useful it was not mandatory to have one done that night and a delay of about twelve hours was reasonable. The MEWS scores, which were assessed hourly, set the parameters and trigger points at which someone more senior should have been contacted. Mr Hughes’ urine output was not so low as to trigger a warning.

Mr Sagar disagreed that resuscitation should have been limited to two to four hours. This was not the only way forward and he considered that Mr Sellu’s treatment plan of resuscitation overnight was more reasonable. The overnight observations on Mr Hughes were not such as to make it obligatory for him to visit the deceased on the morning of 12 February, as readings, other than his urine output, were “okay” and Mr Sellu was intending to come in later in the morning in any event: that was sufficient. Having said that, he agreed that it should have been Mr Sellu’s first priority that morning (as, indeed, Mr Sellu had said that it was). In the event, the observation charts did not indicate an overnight deterioration and there was nothing in them that would have mandated a change of management plan.

Mr Sagar recognised that the blood tests produced some concerning results, but overall they did not mean that Mr Sellu could not properly wait for the CT scan which was pivotal. Having requested an urgent CT scan, there was nothing more that he could do, and there was no duty at that stage to seek advice from the ITU or involve an anaesthetist.

Mr Sagar was of the view was that it was routine for a surgeon to rely upon an expert report (such as that Mr Sellu had received from Dr Kantor) rather than see the underlying images for themselves. The results, showing air in two places, were not of significance to the severity of the deceased’s condition. In the circumstances, he did not think that the appellant’s timeframe of 4-6 hours for an operation was unreasonable and there was nothing in the available information to demand a more urgent operation. It was exceedingly unusual to break into another surgeon’s operating list. Even at 4:00 pm there was nothing to suggest that the deceased was about to decompensate. Had a theatre and anaesthetist been available, however, then it would have been negligent not to operate immediately.

Dealing with P-POSSUM, Mr Sagar objected to its use to predict mortality or morbidity risks in an individual case. It was a statistical tool to analyse clinical performance post-operatively rather than a predictive tool into which results pre-operatively could be fed with any meaningful result but, on the basis that it was used, he broadly agreed with the P-POSSUM scores arrived at by Dr Bell (which provided only the starting point figures).

In short, Mr Sagar was of the view that Mr Hughes had died because he had not been able to withstand the septic insult he had experienced around 10:00 pm as surgery began. It did not mean that he had received substandard care but could be explained by other factors such as age, obesity, and the suggestion of sclerosis of the liver, which may not have been picked up on the CT scan because the focus had been on the acute matter of the perforation. He did, however, agree that the period of time before the deceased received his operation and the lack of antibiotics contributed to the outcome. Overall the treatment plan (assuming it included prescription of antibiotics) had been reasonable.

We add only that, when cross-examined, Mr Sagar agreed that Mr Sellu should have required that antibiotics had been given to the deceased and that had they not been, that would amount to a serious error. He accepted that such an error could be characterised as gross negligence. When re-examined, however, he said that Mr Sellu was entitled to assume that the RMO had prescribed the antibiotics when directed to do so.

Nicol J provided the jury with 7 pages of written directions which defined gross negligence manslaughter as being committed by a person who (i) owes the victim a duty of care, (ii) breaches that duty in a particularly serious or gross manner and (iii) where the negligence causes or significantly contributes to the death of the victim. He explained what did and did not constitute negligence, making it clear that competent doctors could have different schools of thought as to the right approach and that the luxury of hindsight should not be taken into account. He went on:

“But your task is not just to decide whether Mr Sellu fell below the standard of a reasonably competent consultant colorectal surgeon, but whether he did so in a way that was gross or severe. Start with what Mr Sellu knew or ought reasonably to have known about the risk to Mr Hughes’ life if the proper standards were not observed. Then ask yourselves, did Mr Sellu’s behaviour or failure to act fall so far below those standards that his conduct and omissions deserves to be characterised as gross? When we want to weigh a physical object we can use scales marked in ounces or grams. There is nothing similar which I can give you to measure or weigh whether any negligence was ‘gross’. As in many other contexts we leave it to juries to apply their own common and good sense to decide whether the line has been crossed. Using that good and common sense, it is for you to decide whether Mr Sellu acted in a way that was grossly negligent. If you conclude he was then it will mean that his behaviour was potentially criminal.”

The word ‘potentially’ led into a direction on causation which, having explained the prosecution contention, went on:

“You may decide that, even if an earlier operation would not have been bound to succeed, the effect of Mr Sellu’s negligence was to deprive Mr Hughes of a significant chance of survival and in that sense was a significant contributory cause of Mr Hughes’ death. Once again, how big a contribution has to be in order to qualify as ‘significant’ is left to your good sense. So, if you decide that Mr Sellu was grossly negligent in his care of Mr Hughes you must ask yourselves whether the failure to treat him in a proper way significantly contributed to Mr Hughes’ death.”

The only other part of the direction to which it is necessary to refer concerns the judge’s treatment of expert evidence. Nicol J made it clear that expert evidence was the same as other evidence for the jury to assess and evaluate for its strength and weakness. He went on:

“It is you who are trying Mr Sellu, not Mr Kelly, Dr Bell or Mr Sagar.”

Perhaps surprisingly, the judge did not provide the jury with a “Route to verdict” document. There was no obligation under the Criminal Procedure Rules to do so (although such a course is now described: see 25.14(3)(b) CrPrR 2015). It was, however, common practice in complex cases.

After deliberating for a day, the jury sent a note to the court. It read:

“Two questions: one, could we please be reminded of what we must or are to be deliberating on (evidence)? Two, are we to be deliberating legalities or are to be judging as human beings, lay people?”

Dealing with the questions in reverse order, the judge reminded the jury that they had to apply the law to the facts and that he had summarised the evidence. He had explained what constituted evidence but went on that sympathy and any other emotion was not and that there was a difference between that and using the common sense which jurors brought to the task of finding facts and deciding whether the elements of the offences had been established. This direction was approved by counsel in the case. On the afternoon of the following day (after a majority direction), the verdicts were returned.

The first ground of appeal pursued by Mark Ellison Q.C. (now appearing for Mr Sellu) concerns fresh evidence that has been assembled in the period which has elapsed since conviction. It has four limbs all of which it is contended impact on causation. Two of these limbs concern areas of the case which were covered by evidence at the trial. The first of the two revolves around what is said to have been the mis-use of P-POSSUM and, thus, the accuracy and reliability of the prosecution experts’ evidence at trial as to a broadly linear and dramatically decreasing percentage chance of Mr Hughes surviving over the period that he was in the Mr Sellu’s care and the corresponding need for him to have operated earlier than he did. The second revolves around the issue whether Mr Hughes suffered from cirrhosis of the liver and the impact of that on causation.

The two limbs of fresh evidence not ventilated at the trial relate, first, to the realisation (unidentified by any of the experts who considered the papers prior to the trial) that Mr Hughes had been treated with the new anti-coagulant Dabigatran, which were not factored into the P-POSSUM scores and which it is argued would have created a risk of death if he had been operated upon earlier. The second limb is based on research published after the trial (in 2014) which it is argued shows a far higher rate of mortality for patients with faecal peritonitis than was stated at trial and which also questioned the relationship between delayed operative intervention and poor outcome (the GenOSept Report).

For the Crown, Mark Heywood Q.C. responded robustly to the fresh evidence. As for the new material in relation to P-POSSUM, those acting for Mr Sellu had known of the limitations of the statistical device (and Mr Sagar had spoken about it): no adequate explanation has been provided for the failure to obtain and adduce that evidence which could have been available at trial. In any event, it did not afford any potential or sufficient ground for concluding that the verdict was unsafe.

As for the other two heads of fresh evidence, Mr Heywood accepted that the evidence relating to Dabigatran (although potentially available at the trial) had not been addressed by the Crown (or anyone else) and that the GenOSept report had not been available and was therefore completely new: he did not resist its consideration de bene esse but contended that it made no significant difference and went nowhere near undermining the safety of the conviction. To the extent that the evidence had any effect on risk of mortality on surgery, it had no material impact on the other limbs of the prosecution case at trial, namely what he contended was the overwhelming evidence of Mr Sellu’s grossly negligent failure to secure prompt and effective treatment by the administration of antibiotics and by planned and managed resuscitation.

With that introduction and, having regard to the terms of s. 23 of the Criminal Appeal Act 1968 (“the 1968 Act”), we deal with each of the four limbs of fresh evidence in turn.

As we have explained, POSSUM is an acronym derived from a “physiological and severity score for the enumeration of mortality and morbidity”.  It can also be described as a scoring system for surgical audit.  Known since publication in 1991, it has acquired a development, or variant, known as P-POSSUM, based on a refined version of the scoring system developed in Portsmouth University.

In the course of pre-trial discussion and evidence at trial, experts for the prosecution and defence engaged with the prospects of survival for Mr Hughes, depending on the nature and timing of treatment, by the application of antibiotics, effective resuscitation and/or surgery.  Part of that debate, even before the trial, centred on the proper use of POSSUM/P-POSSUM in predicting mortality for Mr Hughes.  The position of Dr Bell and Mr Kelly for the prosecution was that P-POSSUM was a useful tool, set alongside clinical judgment, when assessing mortality for the individual patient.  Mr Sagar for the defence was more hesitant, and Professor Aitkenhead, the consultant anesthetist instructed but not called on behalf of the defence, entered a real note of caution.  In his report of 1 June 2013, he emphasised that the POSSUM score:

“is a system which has been developed to estimate risk in a surgical population…” 

and that system 

“is not intended to calculate risk for an individual patient and cannot predict which patients will survive and which will not.  It simply gives a statistical probability of death or morbidity.”

In his later report of 10 September, Professor Aitkenhead wrote:

“I have remarked previously on the stability of observations during the day, which could have misled clinicians and which renders the P-POSSUM predictions relatively valueless.  Clinically, I think that most clinicians would agree, with the benefit of hindsight, that the risk of death was increasing progressively during the afternoon of 12 February on the basis of experience of the progression of sepsis in patients who have not received antibiotics and who are not receiving resuscitation guided by the results of invasive monitoring.”

When it came to trial, Dr Bell did say that POSSUM was of predictive value, but he placed limited reliance on it in the crucial period. Having acknowledged that POSSUM was an audit tool, he said that the refinement allowed predictions for an individual patient.  He keyed in information for Mr Hughes if diagnosed and treated on the morning of 12 February and it produced a “2.263 per cent mortality rate”.  However, as his evidence proceeded, bearing on prospects during the rest of that day he moved from reliance on any precise figure derived from POSSUM and simply stated that there was an escalating risk of mortality:  a precise figure was “exceptionally challenging and clinical information is confusing in terms of the observations”.  The thrust of Dr Bell’s evidence as the matter unfolded was that Mr Hughes’s prospects were to be judged as the result of clinical judgment.  

Essentially the same picture emerged from the evidence of Mr Kelly.  He described POSSUM as a mathematical formula, commonly used by anaesthetists.  He agreed that the calculations were “approximations”.  Although he gave the POSSUM figures through the morning of 12 February, the remainder of his evidence as to the prospects of survival were given in round figures based on “the educated guess of the experts”.  

In the course of his evidence, Mr Sagar for the Appellant said that POSSUM scores should not be used predictively for mortality but were a tool for measuring the rate at which illness or abnormalities and death occurred over a population.  That point was repeated to the jury by the judge in his summing up, although the judge noted that Mr Sagar had engaged with the POSSUM scores in the course of giving his own views.

As we have observed, this whole question was addressed directly by Professor Aitkenhead before the trial, but the defence chose not to call evidence from him.

Mr Ellison seeks to call evidence from three witnesses on this issue:  Mr Smith, a consultant surgeon with a special interest in risk prediction; Mr Faiz, a consultant colorectal surgeon; and Dr Poloniecki, an expert in statistics and the application of risk models such as POSSUM and P-POSSUM.  In each case, these experts would wish to cast doubt on the utility of POSSUM and P-POSSUM as predictive tools for the mortality risk for an individual patient.  In advancing this application, Mr Ellison submitted that this evidence might afford a ground for allowing the appeal: it would change the context in which the jury reached their decision.  In particular, it might serve to undermine the proposition, coming from Dr Bell and derived from his reading of the POSSUM score, that the prospects of survival for Mr Hughes if there was surgical intervention on the morning of 12 February were around 3 per cent. 

This figure was likely to have placed in the jury’s mind the idea that there was a very steep decline in the prospects for Mr Hughes as they fell towards an approximately 50 per cent chance of survival at some time in the late afternoon, whereas the truth was his prospects declined more gradually from a poorer starting point.

We declined to admit this “fresh” evidence.  There are two main grounds.  Firstly, it is highly questionable whether this evidence might afford any basis for allowing the appeal.  Even if it was to be accepted, it would not alter the picture given to the jury, except as to the prospects for Mr Hughes early in the morning of 12 February.  There was no serious attempt by the Crown to say that Mr Hughes had to have been undergoing surgery as early as that.  The real debate turned on surgical intervention in late morning or promptly in early afternoon.  As we have indicated, by that time all the evidence before the jury was based on clinical judgment, perhaps informed to some degree by the POSSUM scores, but only to a limited degree.  We thus reject the submission that this evidence could, or would, have a significant effect on the outcome.

Moreover, there is no reasonable explanation for the failure to adduce this evidence, or other evidence with the same impact, at the trial.  The issue had been raised squarely by Professor Aitkenhead in his pre-trial reports.  He could have been called to give evidence precisely to this effect.  If the very experienced defence team at trial had thought that further evidence was needed, then one or all of the proposed new witnesses could have been consulted.  We imply absolutely no criticism of the defence team at trial in this regard.  However, we consider that Mr Ellison has failed to establish that the evidence might have formed a ground for allowing the appeal (s. 23(2)(b) of the 1968 Act) and has failed to establish that there was any reasonable explanation for the “failure” to adduce the evidence at trial within the meaning of s. 23(2)(d).  For those reasons we reject this application.

In the course of his contemporaneously recorded surgical note, Mr Sellu wrote that Mr Hughes’s liver appeared to be cirrhotic, with bleeding from a sub-capsular haematoma.  He gave evidence of that to the jury.  No other witness was able to speak about the appearance of the Mr Hughes’ liver from direct sight because no post-mortem was undertaken.  It follows that there were no slides of the liver from which the question could be definitively answered.  In the course of evidence to the jury, the defence expert Mr Sagar mentioned possible cirrhosis as a factor pre-disposing Mr Hughes to react poorly to a septic insult, and the judge reminded the jury of that evidence in his summing up.  The prosecution adduced evidence from the radiologist that the appearance of the liver on CT scan was normal and the prosecution questioned whether there was definitive evidence of cirrhosis.  Mr Kelly described Mr Sellu’s note and evidence as “guesswork”.

On this issue, the Appellant seeks to introduce the evidence of Profession Brian Davidson, consultant surgeon specialising in liver transplant surgery.  His evidence, if admitted, would suggest that Mr Hughes had an elevated bilirubin level, as demonstrated by liver function tests carried out on the morning of 12 February:  a picture “compatible with someone with underlying liver cirrhosis”.  In Professor Davidson’s view, Mr Hughes had “well-compensated chronic liver disease”, which he would grade at “Childs-Pugh A grade” liver cirrhosis, meaning that the abnormality was mild.  Impaired liver function might have made a contribution to the death by reducing Mr Hughes’s capacity to clear lactic acid.  In a refinement of his evidence, Professor Davidson agreed that the literature is poor as to the effect of cirrhosis on mortality.  There is one UK paper with a relatively small study which makes no distinction as to the degree of disease present in the cohort (Tapper, EB et al 2014, in press).  Professor Davidson also refers to a US study of adverse outcomes for colorectal surgery with chronic liver disease (Ghaferi, August 2010).  However, the qualifiers for that study include significant physical signs of cirrhosis which were not present in Mr Hughes.

Once again we decline to admit this evidence.  We are not persuaded that the evidence would afford a ground for allowing the appeal.  Even if we accept for present purposes that it is demonstrated that Mr Hughes had mild cirrhosis of the liver, the consequences are too uncertain and unquantifiable in our view to make any significant difference to the jury’s consideration.  Moreover, this issue was well-known at the time of the trial.  No real explanation has been offered as to why the matter was not developed.  The evidence contained in the report from Mr Davidson could perfectly well have been deployed at trial as coming from him or from other suitably qualified experts.  Once again, we emphasise we imply no criticism of the trial team on this point. This application is also rejected.

The GenOSept study (Patients with faecal peritonitis admitted to European Intensive Care Units: an epidemiological survey of the GenOSept cohort: Tridente & others, 2014), which it is agreed was published after the trial and could not have been available to the experts giving evidence, is a large-scale epidemiological survey of outcomes and risk factors for mortality for patients with faecal peritonitis admitted to intensive care units in 16 countries across Europe. The sample was large, at just under a thousand patients. The “primary outcome measure was six months’ mortality”, meaning death from any cause within six months of admission. The median age of the patients was 69.2 years and the most common cause of peritonitis was perforated diverticular disease, the same condition as Mr Hughes.

The strongest independent risk factors associated with death included age (within a central spectrum of 58-77), a higher APACHE II score (a measure of sickness), acute renal and cardiovascular dysfunction, hypothermia, lowered haematocrit and bradycardia (reduced heart rate) on the first day in the ICU.

Mr Faiz introduced this study in his second report. His purpose in doing so was that:

“The study challenges a long held (and intuitively logical) belief that duration of sepsis is a determinant of mortality. The relationship between sepsis duration, surgical intervention and mortality was not the intended aim of the GenOSept study and this finding should be seen within the context of the wider literature. GenOSept is however a large influential investigation, the findings of which raise some controversy regarding the relationship between duration of sepsis (i.e. time to source control) in faecal peritonitis and mortality.”

Mr Faiz’s view is that the study reflects the uncertainty that exists as to why some patients’ response to sepsis is catastrophic, whilst others are able to survive. The study did not reveal the expected association between earlier surgical intervention and better survival.

In the course of his careful evidence before us, Mr Faiz did not appear to overstate the effect of this study. He readily conceded that the current guidelines for surgical intervention have not been altered in the light of this study. Mr Faiz himself has not altered his own surgical practice: to do so would, he said, represent a “natural test we dare not explore”.

Dr Bell in his report for the Crown of March 2016 emphasised that the authors of the GenOSept study themselves observe that:

“This finding contrasts with previously published studies of secondary peritonitis in which time to reoperation, source control and indices of physiological derangement have been the strongest outcome predictors.”

It seems to us there are a number of reasons, even on the face of the study, why any application to the individual case of Mr Hughes would be tentative and limited at best. The mortality recorded in the study overall was 32%, with age as the strongest associated factor. Mr Hughes’s age was below the median of the cohort studied. He did not have acute renal dysfunction, hypothermia or lowered haematocrit. He was not treated in ICU, although perhaps he should have been. The parameters of the study do not confine relevant mortality to the direct consequences of the faecal peritonitis, much less to septic shock from acute untreated peritonitis. Since all the GenOSept patients were admitted (alive) to ICU, all will have been given active antibiotic treatment, and the paper considers some of the different combinations of such treatment employed.

In his submissions, Mr Ellison was careful not to lay much weight on the impact of the GenOSept study. We consider that he was correct in that approach. This study would not be likely to alter the agreement amongst the experts that timely surgical control of the source of contamination was called for, and would probably be critical in preserving life. As Mr Faiz himself said, most patients will die without surgical treatment. This conviction is not rendered unsafe by the admission of this fresh evidence.

Dabigatran (Dabigatran Etexilate Mesylate) was a relatively new anticoagulant which had been administered to Mr Hughes to prevent the formation of clots after his knee surgery. It is given as a single daily dose orally. Dr Patel, Consultant Haematologist, instructed on behalf of Mr Sellu, stated in his report of 29 September 2014 that the maximum anticoagulant effect of the drug is achieved 2-3 hours following intake. The drug is largely cleared by the liver and the anticoagulant effect is dependent on liver function. Dr Patel writes that:

“The risk of surgical bleeding following a Dabigatran dose is poorly characterised … The majority of data on Dabigatran associated haemorrhage comes from studies using higher dose (twice daily) long term Dabigatran for other purposes than used in this case.”

Mr Hughes was prescribed a final dose of Dabigatran on 11 February at approximately 10:00 pm. Mr Hughes’ renal function, it is agreed, was normal before 13 February and therefore he would have been likely to clear the Dabigatran at a normal rate. Dr Patel therefore suggests that Mr Hughes’s increased risk of surgical haemorrhage is likely to have been high if performed within 12 hours of the Dabigatran dose (that is before 10:00 am on 12 February), and to have been moderately increased over the next 12 hours with a progressive fall in bleeding risk over that period. The recommendation from the manufacturer states that:

“The risk of bleeding may be increased if an acute (surgical) intervention cannot be delayed for 12 hours after the last dose of Dabigatran”.

It follows that, had surgery been performed before 10:00 am, there was a high risk of surgical haemorrhage but thereafter a moderate and declining risk.

It is agreed that there was no obligation on Mr Sellu to be aware of these risks at the time. The drug was relatively new. As we have said, it is noteworthy that none of the experts who advised at trial on this case on either side raised the administration of Dabigatran as a complicating factor at all. It follows that the potential risks associated with this drug could have played no part, and can play no part, in establishing any poor practice by Mr Sellu. The matter is only potentially relevant to causation.

We address elsewhere the question of whether the Crown case was clear as to the last time at which surgical intervention should have taken place. Realistically, it was not a time before 10:00 am on 12 February.

Mr Sellu took his decision to proceed to surgery in ignorance of any bleeding risk associated with Dabigatran. On this scenario, he would and should have paid regard to the serious mounting risks of delay, including coagulopathy as a consequence of sepsis. Even if Mr Sellu had been advised of the risks associated with Dabigatran by another, most likely an anaesthetist, it is clear from the expert evidence he would and should have been calibrating the mounting risks associated with delay, against the moderate and declining risk of surgical bleeding from Dabigatran. In neither scenario is it likely this consideration should or could have delayed surgery beyond the time when Mr Hughes was more likely to die than to survive, that is to say, on one view of the evidence, after 3:00 pm on 12 February.

Equally, the risk of surgical bleeding played no part in Mr Hughes’s death. Dr Patel is clear that the coagulopathy which Mr Hughes did develop at or around the time of his operation at 2:00 am on 13 February was not as a consequence of Dabigatran, but was the outcome of the sepsis and septic shock. At its highest, all that can be said is that there might have been a risk of surgical bleeding had Mr Hughes undergone surgery in late morning through to early or mid-afternoon on 12 February. Given the normal renal function which was maintained throughout this period, the jury would have been told that such a risk was moderate and declining throughout this period. Therefore, it seems to us inevitable that even if this new evidence was before a jury, they would have had little regard to it when considering the causation flowing from the gross negligence which was alleged. Mr Hughes did not die from surgical haemorrhage derived from Dabigatran. The risk that he might have had such haemorrhage was limited during the relevant period and, of course, there is no basis for assuming that such haemorrhage would have been fatal.

For those reasons, we conclude in relation to this fresh evidence, as with the other strands, that it would have made no material difference to the consideration of the jury. None of the material renders the conviction unsafe.

The second and third grounds of appeal concern the question of causation. The case advanced by the prosecution (and summarised by the judge) was that there were a number of stages during the chronology when Mr Sellu could and should have taken action, up to and including the evening of 12 February, and that, in relation to each stage, it was grossly negligent not to do so. That action could have taken the form of arranging for the administration of antibiotics, ensuring that tests were undertaken, or obtaining the services of an anaesthetist and embarking on an earlier operation. Prior to 3:00 pm on 12 February, it was contended that, had Mr Sellu taken appropriate steps, Mr Hughes’ prospects of survival exceeded (or well exceeded) 50%. After that time, however, Dr Bell expressed the opinion that Mr Hughes’ prospects of survival reduced below 50% and although prompt treatment even thereafter could have saved his life, the prospects by the time of the operation had reduced to 1%. Thus, 3:00 pm. was the so-called ‘tipping point’.

The judge provided the jury with written directions which made it clear that Mr Sellu would only be guilty if his gross negligence caused or significantly contributed to Mr Hughes’ death, so that if the jury decided that he was grossly negligent in his care of Mr Hughes “you must ask yourselves whether the failure to treat him in a proper way significantly contributed to Mr. Hughes’ death”. He elaborated in the following terms:

“Mr Sellu will be guilty of the offence ...only if his gross negligence caused or significantly contributed to Mr Hughes’ death. ...Mr Hughes died of diverticulitis, where the pouch or diverticulum perforated and which then caused infection to spread throughout his body and that led to multiple organ failure. Here though, the prosecution contend that Mr Sellu did significantly contribute to Mr Hughes’ death in the sense that he failed to take various steps which could have led to an earlier operation, which in turn, would have had a significant chance of saving Mr Hughes’ life. No operation under general anaesthetic is completely safe. But you may decide that, even if an earlier operation would not have been bound to succeed, the effect of Mr Sellu’s negligence was to deprive Mr Hughes of a significant chance of survival and it that sense was a significant cause of Mr Hughes death.”

As for the tipping point, the judge said:

“Beyond about 15.00 in Dr Bell’s view the chance of Mr Hughes surviving the operation had started to enter the tipping point – that is it was starting to be less than 50%. His chances of survival continued to deteriorate. But in Dr Bell’s opinion, expeditious action even as late as 8.00pm may have saved Mr Hughes’ life....by the time Mr Hughes went into the operation he had about 1% chance of survival....”

The judge had raised a concern with counsel as to whether the jury all had to be satisfied about “some sub-category of negligence” so that if six jurors considered that it was grossly negligent not to prescribe antibiotics and six believed that there was a failure to treat in a timely manner, it would be sufficient to conclude that there was gross negligence and causation. Prosecuting counsel argued that the breaches of duty could be considered individually or accumulatively so that no Brown direction (see R v Brown (1984) 79 Cr App R 115) was necessary. After consideration, defence counsel did not consider such a direction was necessary in this case.

On appeal, Mr Ellison argued that some jurors might have been sure of gross negligence only at such a late time in the chronology (i.e. after 3:00 pm) that the conviction might have been returned without a consideration of the fact that, by then, the likelihood was that Mr Hughes would still have died. Thus, it was left open to them to convict in relation to a failure to act at a stage in the chronology where they were no longer sure that Mr Hughes would have survived in any event i.e. when causation could no longer be proved. Furthermore, he contends that the judge erred in not directing the jury in accordance with Brown that they must all agree as to any particular negligent act or omission before they could move on to decide whether there was gross negligence causative of death.

In relation to these grounds of appeal, Mr Heywood accepted that if death was more likely than not to result even if a duty to act is properly performed, then a conclusion beyond reasonable doubt that the failure to act at that time then contributed to death is not open and causation would not be established. He suggested that this was absent ‘special circumstances’ although it is entirely unclear what he considered such circumstances might be. He argued, however, that the combination of the judge’s directions and treatment of the evidence made the jury’s task clear: they could not convict unless they were sure that the omissions they were considering materially contributed to the death.

In that regard, he submitted that a Brown direction was fact specific and not required in this case. The relevant ingredient to be established was not simply one or more examples of negligent omissions but, rather, the fact that Mr Sellu had failed promptly to deliver life-saving treatment at a time when serious risk to life could have been averted. Mr Heywood argued that the written directions correctly and clearly identified the material ingredient to be established; the disputes of fact that did exist for the jury to resolve neither stood alone nor were likely to be determinative of guilt. Furthermore, both parties had agreed that the direction given was appropriate.

The classical statement of the law in relation to omissions is R v Morby (1882) 8 QBD 571 which concerned a father who, due to religious views, did not employ a doctor to treat his son, who later died of small pox. The medical evidence at trial had been that proper medical attention might have saved or prolonged the child’s life, and would have increased his chance of recovery, but might have been of no avail. Following a conviction for manslaughter, the case was referred to the Queen’s Bench Division, as a Crown Case Reserved. Quashing the conviction, in a pithy judgment, Lord Coleridge CJ put the matter in this way:

“It is not enough to shew neglect of reasonable means for preserving or prolonging the child's life, but to convict of manslaughter it must be shewn that the neglect had the effect of shortening life. The medical witness called for the prosecution gave his evidence clearly and well, and under a high sense of his duty and responsibility, and what he stated was, that in his opinion the chances of life would have been increased by having medical advice, that life might possibly have been prolonged thereby, or, indeed, might probably have been, but that he could not say that it would, or indeed that it would probably, have been prolonged thereby. In order to sustain the conviction affirmative proof is required.”

Mr Ellison also referred to the application of this principle in cases involving a breach of duty arising out of an omission to act, where there has been evidence of a decreasing chance of survival as time passed. In R v Sinclair (1998) WL 1044437, 21 August 1998, giving the judgment of this court, Rose LJ said:

“The judge...had properly and fairly directed the jury that, in relation to causation they should concentrate on the duty of care before 10 pm, because thereafter, they could not be sure that what the defendants did could have been a significant contributory cause to death: this was because Dr Challand’s evidence was that at 10 pm, the chances of recovery with medical treatment would have been of the order of 50%.”

The point also figured in the passages in the summing up in R v Misra [2005] 1 Cr App R 21, [2004] EWCA Crim 2375 which concerned breach of duty arising out of the omission to treat post-operative infection. Commenting on the summing up of Langley J in relation to this aspect of the case, Judge LJ said (at [70]):

“Langley J correctly directed the jury that one of the matters about which they had to be sure before the appellants could be convicted was that such failure or failures as were proved against each individually was a substantial, even if not the sole or the major cause of death. His directions include this passage:

"The last element is the element of causation. If the prosecution has made you sure that either or both of the doctors did fail so grossly in their duty of care, then you must consider whether it has also made you sure that the failure or failures were a substantial cause of Sean Phillips' death. If you are not sure that Sean Phillips would have survived at all, either however well he had been treated or because he might not have received appropriate treatment, then the prosecution has failed to prove its case on this aspect and that is the end of the matter. You must find both defendants not guilty. Equally, if at some point in the events of the Saturday or the Sunday you reach the conclusion that you are not sure that Sean Phillips would have survived beyond that time, then from that time onwards the prosecution will fail to prove that anything Dr Misra or Dr Srivastava did or failed to do was a cause of Sean Phillips' death, and, whatever you think of the subsequent events, they cannot lead you to a verdict of guilty. If you have any reasonable doubt about when Sean's condition became irreversible, I repeat that you must give the defendants the benefit of those doubts."

What was critical was that the jury reached conclusions as to such failings as they were sure constituted gross negligence and, in the light of those findings, went on to consider the question of causation, understanding that causation would not be established if that gross negligence was after the time that they could be sure that Mr Hughes would have survived. In some cases, depending on the way the case is put, a Brown direction might be necessary; in others, it will not.

Theoretically, Mr Ellison is correct in his submission: the judge did not specifically direct the jury to that effect (which would inevitably have driven them to have to reach common ground as to the time at which they were agreed that Mr Sellu had been grossly negligent if such was their conclusion). In the circumstances of this case, however, the facts would not support the proposition that the jury could have decided that the only operative gross negligence was after the time that Mr Hughes would not have survived. Further, in the light of the way in which the case was put, we do not consider that it was incumbent on the judge to give a Brown direction. In our judgment, the way in which the judge left causation to the jury does not lead to the conclusion that this verdict is unsafe and these grounds of appeal (both in relation to causation and Brown)fail.

The final ground of appeal relates to the adequacy of the direction relating to gross negligence manslaughter. In that context, as part of the submission, but not suggested as a free-standing complaint, Mr Ellison developed an argument relating to the repeated extent to which the experts, Mr Kelly and Dr Bell, had been allowed (and, indeed, encouraged) to proffer an opinion at various stages of the chronology as to whether, in the particular regard then being discussed, Mr Sellu’s conduct had been grossly negligent (which was, of course, the ultimate issue). This question was repeated to both on a number of occasions and thereby required them to answer what was the ultimate question which the jury had to address, it being their task (and not the task of the experts) to determine where the line should be drawn.

Mr Heywood argued that there was no authority to suggest that a fuller direction than provided by the judge as to the meaning of gross negligence was mandatory; no one synonym or alternative formulation to the word “gross” was required. It was also contended that the judge’s directions (in writing and, although the recording was then defective, as counsel recollected, repeated verbatim) were unimpeachable as to the necessary level of negligence to be established. It was entirely appropriate for the experts to articulate their view as to whether the negligence of which they spoke was ‘gross’. In the circumstances, we deal first with the proper approach to the “ultimate issue” and then to the direction on gross negligence manslaughter.

The law has developed to the point where an expert has been permitted to give his opinion on what has been called the “ultimate issue” but, in such a case, the judge is required to make it clear to the jury that they are not bound by the expert’s opinion. In R v Stockwell [1993] 97 Cr App R 260, the court was concerned with identification. A facial mapping expert had been permitted to say “the photographs strongly support the view that the suspect and the robber are the same man”. Lord Taylor CJ said (at 265-266):

“Whether an expert can give his opinion on what has been called the ultimate issue, has long been a vexed question. There is a school of opinion supported by some authority doubting whether he can (see Wright (1821) Russ & Ry 456, 458). On the other hand, if there is such a prohibition, it has long been more honoured in the breach than the observance (see the passage at page 164 in the judgment of Parker LJ in Director of Public Prosecutions v A. and B.C. Chewing Gum Ltd (1968) 1 Q.B. 159 and the cases cited at page 501 of Cross on Evidence (7^th ed.) ………. The rationale behind the supposed prohibition is that the expert should not usurp the functions of the jury. But since counsel can bring the witness so close to opining on the ultimate issue that the inference as to his view is obvious, the rule can only be, as the authors of the last work referred to say, a matter of form rather than substance. In our view an expert is called to give his opinion and he should be allowed to do so. It is, however, important that the judge should make clear to the jury that they are not bound by the expert’s opinion and that the issue is for them to decide. Here the judge did just that.”

“Strong support” provides the jury with the view of the expert without positive assertion of identification which could depend on other evidence (either way) to prove or disprove the opinion of the expert. Rather nearer this case, however, is R v Brennan [2015] 1 WLR 2060, [2014] EWCA Crim 2387 which concerned the expression by psychiatrists of qualitative opinions in relation to diminished responsibility and, in particular, whether the impairment of a defendant in understanding the nature of his conduct or forming a rational judgment or exercising self-control (or any combination) was “substantial”. Davis LJ said (at [51]):

“51.

The third stage involves, as did the former version of the section, reference to “substantial impairment”. That does, we accept involve a degree of evaluation potentially indicative of being a jury question (what the phrase actually connotes has recently been the subject of further discussion by a constitution of this court in Golds [2014] EWCA Crim 748). But overall the provisions of s.2 as amended are altogether significantly more structured than the former provisions, in particular by reference to “substantial impairment of mental responsibility” as contained in the original version of s.2 of the 1957 Act. As we see it, most, if not all, of the aspects of the new provisions relate entirely to psychiatric matters. In our view it is both legitimate and helpful, given the structure of the new provisions, for an expert psychiatrist to include in his or her evidence a view on all four stages, including a view as to whether there was substantial impairment. As Professor Ormerod explains in his paper: “Since the question of whether there is impairment of ability is a purely psychiatric question, it would also seem appropriate for the expert to offer an opinion on whether there is ‘substantial’ impairment”. We agree. Moreover, where the expert is able to and does express a view on all four matters we can see no legal or other objection to such expert, if willing and prepared to do so (as here), making explicit in evidence his or her opinion on what is called “the ultimate issue”: the more so when such a view will in any event probably have been implicit from his or her stated opinion on the four matters. It is difficult to see how the expression by an expert of such a view in a given case could contravene any principle of deference to the jury as the ultimate decision makers.”

It is important to put that decision into the context of the legislation. Under s. 2 of the Homicide Act 1957, substantial impairment was in relation to “his mental responsibility for his acts and omissions in doing or being a party to the killing”. As amended by s. 52 of the Coroners and Justice Act 2009, the substantial impairment relates to understanding the nature of his conduct, forming a rational judgment and exercising self control. These are all medical issues upon which psychiatrists are qualified to express an opinion. The final limb of diminished responsibility is that the abnormality provides an explanation for the act or omission in doing or being a party to the killing, defined by s2(1B) as causing or being a significant contributory factor in causing the carrying out of the conduct. That might be a medical question or it might not. It would be for the psychiatrists to explain and the jury to assess.

In the instant case, for the jury to reach an adverse conclusion in relation to negligence, they had to be sure that Mr Sellu’s standard of care fell below what should reasonably have been expected from a competent consultant colorectal surgeon. In order to consider that question, the jury were entitled to receive evidence from appropriate medical experts as to their opinion as to what should reasonably have been expected from a competent colorectal surgeon and so to decide whether, on the facts as the jury found them to be, Mr Sellu fell below the requisite standard. However, whether any such negligence was “gross” (which is not a medical term) involved an evaluation for the jury. Medical opinion may be better informed on that point but it is not and could not be determinative. Experts might be able to place negligence on a spectrum (and examples can given of that spectrum) but this assistance needs to be considered by the jury in the context of all the circumstances as the jury find them to be, rather than as evaluated by the experts.

The importance of the jury as the ultimate decision maker has been recently reaffirmed in Pora v R [2016] 1 Cr.App.R. 3 [2015] UKPC 9 where, in a case involving a psychologist giving evidence about an unreliable confession, Lord Kerr put the matter in this way:

“24.

……..It is the duty of an expert witness to provide material on which a court can form its own conclusions on relevant issues. On occasions that may involve the witness expressing an opinion about whether, for instance an individual suffered from a particular condition or vulnerability. The expert witness should be careful to recognise, however, the need to avoid supplanting the court’s role as the ultimate decision-maker on matters that are central to the outcome of the case. …

27.

The dangers inherent in an expert expressing an opinion as an unalterable truth are obvious. This is particularly so where the opinion is on a matter which is central to the decision to be taken by a jury. There may be cases where it is essential for the expert to give an opinion on such a matter but this is not one of them. It appears to the Board that, in general, an expert should only be called on to express an opinion on the “ultimate issue” where that is necessary in order that his evidence provide substantial help to the trier of fact…”

Reverting to the present case, Nicol J gave the following direction about the expert evidence:

“Both prosecution and defence have relied on expert witnesses. The witnesses have assisted with some of the technical terms which are used. But, they have also been called to give their views on how a reasonable colorectal surgeon would have responded at various stages to the information that was available about the condition of Mr Hughes and what, if any, further information a reasonable colorectal surgeon would have gathered. They have given their views as to the chances of survival if an operation had been carried out on Mr Hughes at different times through 11 and 12 February.

In one sense expert evidence is in a special category. Other witnesses give evidence about what they did, what they intended, what they saw or otherwise perceived – in other words evidence about the facts. Experts are allowed to give evidence about their opinions. They are allowed to do so because they have an expertise in areas which you and I do not.

But in another sense, expert evidence is precisely the same as other evidence. Their evidence, like that of all the witnesses, is laid before you to assess and evaluate for its strength and weaknesses. It is you who is trying Mr Sellu, not Mr Kelly, Dr Bell or Mr Sagar.

Remember, too, that while the experts deal with their particular parts of the case, you receive all of the evidence and it is on all of the evidence that you make your final decisions.”

That direction clarified that it was the jury, not the experts, that were tasked with making the final decision as the ultimate decision maker. As it was pointed out in Pora, however, it was the duty of an expert witness to provide material on which a court can form its own conclusions on relevant issues. The question arises in this case as to what material was presented to the jury upon which it was to base its conclusion.

For the purpose of the appeal, it was agreed that the expert evidence was accurately summarised by the judge in his summing-up. In the course of the summing-up, as already referred to in our summary of the evidence, the judge referred to numerous different expressions of expert opinion as to the quality of Mr Sellu’s performance.

Collecting them together, the judge reminded the jury of six occasions when Mr Kelly referred to Mr Sellu’s “gross negligence” or to his being “grossly negligent”. These were when, on the evening of 11 February, he did not set an initial plan with goals; when, the following morning he did not commission blood tests and blood gases, contact the ITU and make a plan for an operation; when, thereafter, he did not go down to the imaging department and beg them to do a CT scan; when, after the CT scan had been done, he did not look at the CT scan himself; when he saw Mr Hughes at 4 pm, he did not review Mr Hughes’ observations; and when, in his conclusion, he said that, because of his gross negligence, he had missed quite a wide window of opportunity, which probably ended at 4 pm, to undertake any lifesaving, standard operation on Mr Hughes. The judge also reminded the jury that Mr Sagar accepted that it would have been grossly negligent for Mr Sellu not to have required antibiotics to be given to Mr Hughes.

The judge additionally referred to a number of different expressions used by the experts to describe Mr Sullu’s performance. He referred to Mr Kelly stating that his failure to liaise with the ITU was “very bad practice”; that “no reasonable surgeon” could have reached a conclusion that Mr Hughes had got no worse overnight; that he “embarked on a bizarrely slow and laidback and inadequate treatment and diagnosis regime which if proposed by a candidate for a basic doctor’s examination would result in a fail”; such that his overall performance amounted to “recklessness” as well as being “grossly negligent”. He referred to Dr Bell stating that Mr Sellu’s failure to operate within 2-4 hours from the time of diagnosis of free gas was, in his opinion, “grossly incompetent”, and “there were significant errors of judgment on multiple occasions and deviations from good medical practice, comprehensive assessment and documentation falling below the level of a reasonable practitioner on multiple occasions”. The synonym he used from Mr Sagar’s evidence if Mr Sellu had not required antibiotics to be given to Mr Hughes, was that it would also have been “a serious error”.

The above analysis demonstrates the variety of different ways that the experts described Mr Sellu’s actions and inactions. Some of them do provide a yardstick against which the jury could consider whether the criminal test had been met. Others are little more than assertion. How were the jury to assess these opinions on what were, ultimately, legal issues? How was the jury to differentiate between and evaluate how much weight should be given to the assertions as opposed to other qualitative descriptions? Contrary to the specific assistance the jury was given in relation to the bibliography of medical terms that was within a section of the jury bundle, there is nothing to suggest the experts provided any explanation as to the terminology of many of their opinions. Further, the jury was given no additional guidance by the judge other than that which we have cited from the summing-up as to expert evidence generally.

The inherent dangers referred to by Lord Kerr in Pora do not appear to have been appreciated (or guarded against) in the context of the evidence given by the experts. The jury was left on its own to trawl through the differing descriptions, which were adduced in evidence essentially by leading questions, essentially asking whether the behaviour under discussion was or was not gross negligence. In circumstances where those conclusions were not subject to any more detailed explanation and sat alongside a series of other descriptions which were also not expanded upon, the danger existed that the jury merely may have accepted without more that Mr Sellu had been grossly negligent on those occasions where he was stated to have been grossly negligent. Thus, notwithstanding the judge’s direction, the jury’s role as the ultimate decision maker may have been supplanted.

It is not suggested that the way in which the experts gave their evidence was itself a sufficient ground for allowing the appeal but it is against that background that the directions on gross negligence manslaughter fall to be considered.

Before analysing the present case, it is worth recounting the development of articulation of this offence. Thus, the modern authoritative analysis of the law of gross negligence manslaughter is generally considered to be contained in the unanimous opinion of the House of Lords, expressed by Lord Mackay LC in R v Adomako [1995] 1 A.C. 171 in these terms (at page 187)

“...in my opinion the ordinary principles of the law of negligence apply to ascertain whether or not the defendant has been in breach of a duty of care towards the victim who has died. If such breach of duty is established the next question is whether that breach of duty caused the death of the victim. If so, the jury must go on to consider whether that breach of duty should be characterised as gross negligence and therefore as a crime. This will depend on the seriousness of the breach of duty committed by the defendant in all the circumstances in which the defendant was placed when it occurred. The jury will have to consider whether the extent to which the defendant's conduct departed from the proper standard of care incumbent upon him, involving as it must have done a risk of death to the patient, was such that it should be judged criminal.”

Lord Mackay observed (at page 189) that he “did not wish to state the law more elaborately than [he had] done” or “to take expressions used in particular cases out of the context of the cases in which they were used and enunciate them as if applying generally”. He went on:

“The task of trial judges in setting out for the jury the issues of fact and the relevant law in cases of this class is a difficult and demanding one. I believe that the supreme test that should be satisfied in such directions is that they are comprehensible to an ordinary member of the public who is called to sit on a jury and who has no particular prior acquaintance with the law.”

On the face of it, this comparatively straightforward approach was supported by this court in R v Misra (supra): see [61]. Going into the authorities further, however, it is important to underline what was at issue in these cases. Thus, Adamako was concerned with whether it was a necessary requirement to refer to the definition of recklessness (as set out in R v Lawrence [1982] AC 510). On the contrary, it was held that it was sufficient to adopt the gross negligence tests following R v Bateman 19 Cr App R 8 and Andrews v DPP [1937] AC 576.

Going back to those decisions, in Bateman,Lord Hewart CJ put the test in this way (at page 11):

“In explaining to juries the test which they should apply to determine whether the negligence in the particular case amounted or did not amount to a crime, judges have used many epithets, such as ‘culpable’, ‘criminal’, ‘gross’, ‘wicked’, ‘clear, ‘complete’. But, whatever epithet is used and whether an epithet be used or not, in order to establish criminal liability the facts must be such that, in the opinion of the jury, the negligence of the accused went beyond a mere matter of compensation between subjects and showed such disregard for the life and safety of others as to amount to a crime against the state and conduct deserving punishment.”

In Andrews, Lord Atkin (at page 583) quoted this passage from Bateman and spoke of the requirement for “a very high degree of negligence”, the word reckless being “of all the epithets” the one that “most nearly covers the case”. It was that concept of the use of the term recklessness that led to the review in Adomako, rejecting the proposition that the use of that term was required. The case did not concern how the concept of gross negligence should be explained.

Similarly, in Misra. The issue for the court was, in one sense, not the articulation of the test but the question of circularity and uncertainty (relevant to the application of Articles 6 and 7 of the ECHR) on the basis that the jury had to assess whether (in the case of a doctor) there had been a departure from the proper standard of care such that it should be judged criminal (which Lord Mackay in Adomako recognised “involved an element of circularity”). Dealing with that question, Judge LJ said (at [62]):

“The decision [in Adomako]whether the conduct was criminal is described not as ‘the’ test, but as ‘a’ test as to how far the conduct in question must depart from accepted standards to be ‘characterised as criminal’. On proper analysis, therefore, the jury is not deciding whether the particular defendant ought to be convicted on some unprincipled basis. The question for the jury is not whether the defendant’s negligence was gross, and whether, additionally, it was a crime but whether his behaviour was grossly negligent and consequently criminal. This is not a question of law, but one of fact, for decision in the individual case.”

Read in its proper context, Judge LJ was not suggesting that asking the jury to decide whether the conduct was grossly negligent, albeit a necessary direction, was itself sufficient definition of gross negligence manslaughter. Indeed, judges have generally not considered it to be such. Thus, in Misra itself, the court described as “a fair and balanced summing up” the approach of Langley J, again extensively setting out the terms of his direction:

“… duty and breach of duty - … will be the starting point to establish civil liability to pay damages. But as you would expect, and is the law, the prosecution must make you sure of something much more, and much more serious, than that before a person can be convicted of the crime of manslaughter. That is why you see in the indictment the words 'gross negligence'. Mistakes, even very serious mistakes, and errors of judgment, even very serious errors of judgment, and the like, are nowhere near enough for a crime as serious as manslaughter to be committed. If you do conclude that you are sure that either or both of the defendants have been in breach of their duty of care in their treatment of Sean, you must therefore go on to consider the nature of that carelessness or negligence, as you find it to be. Over the years, the courts have used a number of expressions to describe this vital element of the crime, but the key is that it must be gross in the perhaps slightly old-fashioned sense now of the use of that word. So in this case, when you are considering the conduct of each doctor, I think you will find it most helpful to concentrate on whether or not the prosecution has made you sure that the conduct of whichever one you are considering in all the circumstances you have heard about and as you find them to be, fell so far below the standard to be expected of a reasonably competent and careful senior house officer that it was something, in your assessment, truly exceptionally bad, and which showed such an indifference to an obviously serious risk to the life of Sean Phillips and such a departure from the standard to be expected as to amount, in your judgment, to a criminal act or omission, and so to be the very serious crime of manslaughter.”

This articulation is not very different from other first instance directions to the jury which we have seen. We do not identify them as precedents (and any directions must obviously be tailored to the individual circumstances of the case then under consideration) but merely quote them as examples of the type of assistance which juries have, in fact, received and which illustrate the type of guidance that will assist a jury to resolve this issue. Two examples suffice. The first refers to “something which was truly exceptionally bad which showed such an indifference to an obviously serious risk of death of the deceased and such a departure from the standard to be expected as to amount to a criminal act and omission and so to be the very serious crime of manslaughter”. The second similarly expresses the test in clear terms as follows:

“It is not enough to found guilt that Dr [X] was negligent. You must be sure of gross negligence.

Mistakes, even very serious mistakes, and errors of judgment, even very serious errors of judgment, and the like, are not enough for a crime as serious as manslaughter to be committed. You must go on to consider the nature of the carelessness or negligence, as you find it to be. …

Over the years, in relation to the crime of gross negligence manslaughter, the courts have used a number of expressions to describe the additional element which is encompassed in question five. The key is that the breach of duty must be gross. It must have been so bad, so obviously wrong, that, having regard to the risk of death involved in it, it can properly be condemned as criminal, not in some technical sense of the word like somebody might be regarded as a criminal if they didn't have a light on the back of their bicycle, but in the ordinary language of men and women of the world. So, in this case, when you are considering the conduct of Dr [X], you may find it helpful to concentrate on whether the prosecution have made you sure that the conduct of Dr [X], in all the circumstances you have heard about and as you find them to be, fell so far below the standard to be expected of a reasonably competent General Practitioner that, in your assessment, his breach of duty – his negligence – should be characterised as gross in the sense that it was truly exceptionally bad and was such a departure from that standard that it consequently amounted to it being criminal and thus the criminal offence of gross negligence manslaughter.”

These directions (with an overall view of the range of conduct from mistake upwards) contrast with the way in which Nicol J left the issue to the jury (at [76] above) referring only to whether Mr Sellu fell below the standard “in a way that was gross or severe”, and whether in the light of what he knew or ought reasonably to have known about the risk to Mr Hughes’ life if the proper standards were not observed, “his conduct and omissions deserves to be characterised as gross”, for which purpose the jury had to use their common sense. Mr Heywood is correct to say that no particular formulation is mandatory, but what is mandatory is that the jury are assisted sufficiently to understand how to approach their task of identifying the line that separates even serious or very serious mistakes or lapses, from conduct which, to use the phrase from the above direction, was “truly exceptionally bad and was such a departure from that standard [of a reasonably competent doctor] that it consequently amounted to being criminal”.

Mr Heywood also points to the fact that, in the opening words of her closing speech to the jury, prosecuting counsel said that the death of Mr. Hughes was avoidable and would have been avoided had not Mr Sellu’s treatment of him been “so truly exceptionally bad”, thereby echoing the language in the examples to which we have referred. Had the judge directed the jury in those terms, there could have been no complaint. But Mr Sellu was entitled to have that formulation underlined and reinforced by a direction in law from the judge, carrying the weight of the instruction that his approach had to be followed. As it is, what the jury had was the opinion of the experts expressed when asked the ultimate question without further elaboration.

We are reinforced in the view that the judge’s directions were insufficient to inform the jury as to the correct approach because of his response to the assistance that they sought, asking whether they were “deliberating legalities or … judging as human beings, lay people”. He dealt with the question by repeating the directions he had given, underlining that sympathy and emotion played no part, again without identifying where the line that they were seeking to identify should be drawn. That counsel did not submit that he need do no more is not to the point.

Furthermore, the judge did not repeat the important direction that what was gross negligence was a matter for them and not the experts. That was particularly important given the manner in which evidence had been repeatedly adduced from Mr Kelly and Dr Bell, at various times during the recitation of the chronology, in relation to that ultimate question of gross negligence in the way that we have described. In the circumstances, we do not believe that Mr Sellu had the benefit of sufficiently detailed directions to the jury in relation to the concept of gross negligence contained within the offence of gross negligence manslaughter. This failure was underlined by the way in which the experts had asserted gross negligence and aggravated by the absence of a route to verdict which would have focussed their minds on the various stages to be considered.

Although we reject the grounds of appeal based on fresh evidence and causation, we have come to the clear conclusion that the way in which the issue of gross negligence manslaughter was approached (and, in particular, the consequential direction to the jury) was inadequate. As a result, the conviction is unsafe and is quashed.