Stephen Niland v R [2025] EWCA Crim 945

[2025] EWCA Crim 945

Case No 202401229 B2

IN THE COURT OF APPEAL (CRIMINAL DIVISION)

ON APPEAL FROM THE CROWN COURT SITTING AT SHEFFIELD

T20060267

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

Date 25/07/2025

Before

THE LADY CARR OF WALTON-ON-THE-HILL

LADY CHIEF JUSTICE OF ENGLAND AND WALES

MR JUSTICE JAY

MR JUSTICE GOSS

Between

STEPHEN NILANDApplicant

THE KINGRespondent

Anya Lewis KC and Dale Harris (instructed by GWB Harthills Solicitors) for Applicant

Louise Oakley (instructed by CPS Appeals and Reviews Unit) for Respondent

Hearing Hearing date: 17 July 2025

Approved Judgment

This judgment was handed down remotely at 10.00am on Friday 25 July 2025 by circulation to the parties or their representatives by e-mail and by release to the National Archives.

.............................

The Lady Carr of Walton-on-the-Hill, CJ :

[1]On 1 June 2006 in the Crown Court sitting at Sheffield Mr Stephen Niland (“the Applicant”) pleaded guilty to one count of causing grievous bodily harm to his three-month daughter, Molly, on 10 October 2005, contrary to s. 20 of the Offences against the Person Act 1861 (“the s. 20 offence”) (“the 2006 conviction”). He was later sentenced to (and served) 2 years 3 months’ imprisonment.

[2]On 27 February 2020 Molly died as a result of a chest infection linked to the brain injury and associated disabilities that she had sustained in 2005. The Applicant was charged with her manslaughter and was tried for that offence in the Crown Court sitting at Sheffield (Mrs Justice Lambert DBE (“the Judge”) sitting with a jury) between 5 and 27 March 2024 (“the manslaughter trial”). Causation was not in issue. Rather, the Applicant denied committing any assault on Molly. He stated that he had only pleaded guilty in 2006 because he was told that the medical evidence against him was overwhelming and he would receive a shorter custodial sentence. At the conclusion of the trial, he was acquitted.

[3]He now applies for leave to appeal against the 2006 conviction on the basis that it is unsafe (for the purposes of section 2 of the Criminal Appeal Act 1968). He also applies, as he must, for a very substantial extension of time (6,488 days) in which to do so. He makes a further application to vary his Grounds of Appeal so as to be able to advance the following Grounds: i) Ground 1: he entered a guilty plea on the basis of advice that the medical evidence against him was overwhelming and in order to get a shorter sentence. ii) Ground 2: there was no other evidence which unequivocally supported his acceptance of guilt at the time. iii) Ground 3: there have been developments in medical science since the time of the original proceedings, in particular, as to the extent to which the “triad” alone can be said to be diagnostic of abuse. iv) Ground 4: he was subsequently tried before a jury in circumstances where he bore the burden of proving, on the balance of probabilities, that he had not assaulted Molly. The jury, having heard all the evidence relied upon by the prosecution at the time of the original proceedings, and further evidence obtained subsequent to those proceedings, acquitted him. Thereby the jury were satisfied on the balance of probabilities that his plea was equivocal and that he did not cause Molly grievous bodily harm. (This was the single original Ground of Appeal.)

[4]In her oral submissions, Ms Lewis KC for the Applicant made it clear that her primary submission was contained in Ground 4 alone: the simple fact that the jury acquitted the Applicant at the manslaughter trial is sufficient to render the 2006 conviction by way of guilty plea unsafe. However, in the alternative, Ms Lewis submitted that the circumstances as a whole, including the factors underpinning Grounds 1 to 3, render the 2006 conviction unsafe. She did not contend that Grounds 1 to 3 either together, or cumulatively with each other, were sufficient to allow the court to quash the conviction. They were relied on only in support of Ground 4, if necessary.

[5]The relevant principles applicable to the application for an extension of time and to vary are well-known: see R v Thorsby

;

and R v James

;

respectively. It is convenient to assess the merits of each application in the context of a consideration of the merits of the substantive grounds as a whole.

The Facts

[6]We may summarise these from the Agreed Facts in the manslaughter trial.

[7]Molly Niland was born on 21 July 2005, along with her twin brother. She was delivered vaginally with ventouse assistance at 40 weeks gestation. Her birth weight was within the normal range. The medical notes documented no other complications of labour. Thereafter, Molly lived with her parents, the Applicant and Sally Ashmore, and her twin brother at an address in Rotherham.

[8]At a “birth visit” which took place on 4 August 2005, the health visitor recorded no concerns regarding Molly’s feeding and her passing stools and urine. There was evidence of slight jaundice. Molly was said to be “alert when awake”. It was also noted that the house was clean and tidy and no apparent concerns were expressed.

[9]On 6August 2005, Molly’s recorded weight had risen. On 10 August the same health visitor noted that Molly (and her brother) were still slightly jaundiced but after a telephone call to the GP, she was advised that all was fine and there was no need for concern.

[10]At 23:21 hours on 10 September 2005 Molly attended the out-of-hours service in Doncaster and was seen by a doctor. The examination note recorded “slightly pyrexial, good colour, some grunting present while breathing. Chest sound clear”. The diagnosis was recorded as “Viral URTI only” and drugs were prescribed. A follow up message recorded “Admitted to hospital”. Molly left the out-of-hours service at 23.25 hours.

[11]At 23.40 hours, Molly attended Doncaster Royal Infirmary with her parents. The reason for admission, as recorded in the admission notes, was a suspected “pyrexia” and “breathing difficulties”. Molly was examined and an audible grunting was noted. She was admitted overnight for observations. There was no raised temperature, and no abnormality of breathing noted. The following morning, Molly was discharged home.

[12]On 12 September 2005, at the GP surgery, Molly was examined for her routine six weeks old child health check-up. No concerns were noted. Three days later Molly had her first injections.

[13]On 4 October 2005, Molly and her twin brother attended the baby clinic with both her parents. She was now over double her birth weight. Two days later, Molly was taken to her GP surgery and presented with “nasal catarrh/acute croupy cough since last night”. Nasal drops were prescribed.

[14]On 7 October 2005 Molly was taken by her parents to Doncaster Royal Infirmary having reported concerns about a non-fading rash to the GP out-of-hours service. After being reviewed by paediatric doctors, Molly was discharged home on the following day.

[15]On the 17 October 2005, Molly was taken to the GP surgery. Her mother expressed some concern about Molly in that she had a cough and was chesty. A GP examined Molly’s chest, which was said to be clear. Reassurance was given as well as the second stage of routine vaccinations.

[16]On 18 October 2005, Molly, her twin brother, mother and maternal grandmother attended the baby clinic where a nursery nurse was also present. They were seen by a heath visitor. Molly seemed a little tearful, but her mother explained that Molly was hungry, and that this was normal. Molly’s head circumference was recorded at 40.4cm, which one of the experts instructed at the manslaughter trial said was in the 90th percentile. The health visitor had no concerns with the family and was more than happy with the coping mechanisms displayed.

[17]On 19October 2005, the Applicant was on his own looking after Molly and her twin brother. At 11.14 hours the Applicant made a 999 call in which he reported that Molly had turned white and stopped breathing. At 11.46 hours Molly arrived at Barnsley Hospital where a CT scan was performed. The findings were of “[r]ecent subdural and subarachnoid haemorrhage on the left side extending from the left temporal region to the occiput and vertex. No fracture identified. Relative hypoplasia of the right side of the brain.” A scan of Molly’s abdomen and chest was also undertaken. No fractures were seen. Molly was then transferred to Sheffield Children’s Hospital where she was admitted at 17.00 hours.

[18]Later that evening, Molly’s subdural space was drained to relieve raised intracranial pressure by inserting a tap into the subdural space. At 02.34 hours on 20 October 2005 a further CT scan was undertaken. This scan was reviewed by a radiologist who concluded there was no evidence of a skull vault fracture or of an overlying soft tissue injury. Internally, “[t]here are bilateral supra-tentorial shallow subdural haematomas larger over the left convexity than the right and extending into a parafalcine distribution particularly posteriorly. There is also probably some posterior fossa subdural haemorrhage.”

[19]According to a medical record timed at 09:10 hours on 20 October:

“I have several conversations with mum and dad over the night shift. Both appear to be very worried and distressed by Molly’s condition. I explained that she had a serious injury to her head and brain and this has led to swelling and bleeding. I said that it was likely that this could affect her brain for some time and that she may be left with permanent brain injury but I could not predict the extent of it. Mum repeatedly said that they wouldn’t have deliberately done anything to harm Molly. She kept going over her actions of the previous few days and tried to think if she could have done anything or if anything had happened that could have led to this. She mentioned a soft toy that was hanging near the swing seat she had been in yesterday – asked if that could have hit her head. I explained that small knocks or gently rocking wouldn’t cause this level of damage. Mum received a phone call from Molly’s maternal grandmother this morning and learnt that they would not be able to have Molly’s twin brother, back at home at present. Mum became very distressed at this. She said she could not bear to lose her children and didn’t understand why we were doing this. Dr Stack was also present at this time and explained that something must have happened to have caused the injury and that it would probably have happened yesterday although [he] could not say at what time. After he left, I continued to talk to mum about this. She went over all of her activities with Molly until she left and could not think of anything that could have hurt her. However, she said, “it must have been me” several times. She said that Steve (Molly’s dad) had only left her for a second to go to the toilet. She said that Steve couldn’t have hurt [Molly] because he loves her so much is a great dad and would have said if he had accidentally knocked her head. At the bedside mum and dad again discussed the events with me, focusing on the fact the injury would have happened yesterday. Steve said that when he came back into the room Molly was lying on a blanket on her mat on the floor but was curled up, was white and wasn’t breathing. He phoned 999 and then shook her to try and start her breathing. He phoned 999 and then shook her to try and start her breathing again. Mum asked him to describe how he shook her, he said “he was in shock, it wasn’t gentle but it wasn’t really rough either”

. He also said if I’d knocked her head or dropped her I would have told them on the 999 call so I could help to make her better.”

[20]A MRI scan was conducted at 18.30 hours on 20 October 2005. This showed “a tiny shallow subdural in the posterior fossa and further subdural and subarachnoid blood over the left cerebral hemisphere.”

[21]On 21 October 2005, a full skeletal survey was undertaken. There was no evidence of skeletal injuries, and the skull x-ray suggested some sutural widening consistent with increased intracranial pressure.

[22]A further CT scan on 23October 2005 was reviewed by a radiologist. This noted “still acute left sided subdural and subarachnoid blood.”

[23]The Applicant’s account to Social Services on 20 October 2005 was that all had been usual, Molly was hungry, crying and refusing her bottle. The Applicant said that he had placed Molly on the floor with a blanket over her and had fed Molly’s brother. He then returned to Molly and made attempts to feed her again. She would not feed and continued crying. He placed her back onto the mat and went upstairs to the toilet. He was away for no more than three minutes and on his return saw Molly on her side pure white and gasping for breath. He picked her up and shouted her name and then phoned for medical assistance.

[24]The Applicant, then aged 22 and with no previous convictions, was interviewed on 20 and 21 October 2005. He maintained his account that Molly had been crying since she woke for her feed and had struggled to “latch on” when feeding. He had left her for just a couple of minutes to go to the toilet and when he returned her back was arching, her toes pointed, and she was struggling to breathe. He added that he may have shaken her gently to try and revive her, but he did not assault his daughter.

The Medical Evidence in 2005/6

[25]The prosecution obtained medical evidence from five experts, only four of whose reports we need summarise.

[26]Dr Neil Wright was the Consultant Paediatrician who had care of Molly after she was admitted to Sheffield Children’s Hospital on 19October 2005. He spoke to Molly’s parents who confirmed there had been no neonatal problems. He noted that the initial CT scans showed acute intracranial bleeding and subdural haemorrhages. Dr Wright observed that “the combination of retinal haemorrhages, a macular fold/retinal tear, and evidence of acute subdural haemorrhages and fresh intracranial bleedings suggested significant trauma to the brain”. He further noted that neither of Molly’s parents had offered any explanation as to how she sustained her brain injuries and that both of them stated Molly was fit and well on the morning of her admission to hospital. In his opinion, the clinical picture and injuries were consistent with non-accidental injury, and he thought it most likely Molly was shaken. As to the timing of the injuries, given their serious and extensive nature he felt that it was probable Molly would have become ill within a very short space of time after the injuries were inflicted.

[27]Mr Patrick McMullan, a Consultant Neurosurgeon, examined Molly on 19October 2005. He confirmed that Molly had suffered a widespread death of her brain tissue, together with severe and irreversible brain damage. In his opinion, the likely mechanism was shaking.

[28]Dr Jane Marr, a Consultant Ophthalmologist, examined Molly on 20October 2005. Her findings were of four distinct retinal haemorrhages to the right eye and extensive multiple haemorrhages throughout the entire retina to the left eye. There was also a large macular fold to this eye. Dr Marr opined that, in the absence of infections or known disorders, the findings were “strongly indicative” of a forceful injury or repetitive shaking type injury. Dr Marr subsequently confirmed that macular folds, together with retinal haemorrhages and brain trauma, could lead to a “confident diagnosis” of shaking injury.

[29]Professor Christopher Milroy, a Home Office Pathologist, reviewed Molly’s medical records in 2005. In his opinion, Molly was apparently well before being admitted to hospital where she was found to have suffered subdural haemorrhage with associated retinal haemorrhage in the macular fold. There were no external injuries. In the absence of a medical cause, these findings had to be considered traumatic in origin.

[30]It may be understood that these experts did not express their conclusions with identical conviction. In relation to a traumatic cause, in other words shaking, the gradations of opinion were from “likely” to “confident diagnosis” to “had to be considered traumatic”. The familiar “triad” of(1) subdural haemorrhages and intracranial bleed,

(2) retinal haemorrhages, and

(3) encephalopathy (or, as Dr Wright put it, severe encephalomyelopathy) was not being advanced by the Crown as absolute proof of guilt.

[31]According to the pre-sentence report, which was not in evidence in the manslaughter trial following an objection made by Ms Lewis:

“2.8 In interview Mr Niland explained that he had fed both children at approximately 10.30am. They continued to cry after feeding and he put Molly on the floor and went upstairs to use the toilet. When he came downstairs two minutes later Molly was lying on her side and her back was arched. He tells me that he did not know what was wrong. She appeared lifeless and he half picked her up and moved her back and forth as if to rouse her gently. He phoned for an ambulance and continued to follow instructions from the operator. He stated that whilst on the phone, Ms Ashmore arrived home. He denies shaking the baby vigorously. However, he has acknowledged that his response to her distress was to shake her. He disclosed having very limited experience in caring for babies and I believe that he would have felt at a loss as to what to do. 2.10 [There is no para 2.9] In committing the offence Mr Niland demonstrated naive behaviour as he failed to consider how his actions could impact upon the health of his daughter. Mr Niland now recognises that he was reckless as to the harm that could result and that he had a “momentary loss of control” due to his own frustration and inexperience. 2.11 To his credit, Mr Niland has admitted committing the offence and in my assessment takes some responsibility for causing the injuries sustained by the victim. In response to victim awareness issues the defendant admits the injuries to the victim are gravely serious and shows significant remorse for his actions and how these have, and will continue to, impact upon his daughter. However, he continues to struggle with accepting that his actions caused such extensive injuries.”

[32]Although it is correct, as Ms Lewis submitted, that on the Applicant’s account the shaking of Molly occurred after her collapse, what he told the probation officer is inconsistent with the case he advanced to the jury. The Manslaughter Trial The Applicant’s evidence: the Guilty Plea in 2006

The Manslaughter Trial

[33]There is no longer any documentary evidence of the circumstances surrounding the Applicant’s guilty plea in 2006. The Applicant himself has some limited memory. On 1 June 2006 the Applicant was represented by a barrister who has since died. The barrister who represented him when he was sentenced on 30 June has no recollection of the case. The solicitor’s firm who acted for him no longer exists. No transcript of the sentencing hearing is available.

[34]The Applicant’s evidence at the manslaughter trial was that his memory of the court proceedings was not too clear because he had “tried to block a lot of stuff out”. He said that he had met his solicitor at her office before he went to Court on an occasion before 1 June 2006. When he went to Court on that date he thought that he would be pleading not guilty. He then met his solicitor outside Court and she advised him to plead guilty as there was a lot of medical evidence stacked against him. He was told that if he pleaded guilty and went to trial he would be found guilty and imprisoned for five to six years. That he would be wasting everyone’s time and money. On the other hand, if he pleaded guilty he would get nine to twelve months. He therefore pleaded guilty. The Applicant said that he did not speak to his partner before he pleaded guilty and that the discussion was only ten minutes before he went into Court. He could not recall any conversation with a male barrister on that occasion. He did not remember the case being adjourned for sentence but thought that he was sentenced on the same day. In cross-examination he said that he was told that he did not have a chance because of the medical evidence. He agreed that it was his decision to plead guilty. He denied that he pleaded guilty because he was guilty; he said that he pleaded guilty to something he did not do to get less jail time and get back to his family. When the statutory maximum for a s. 20 offence was put to him and it was suggested that no lawyer would have advised him that he would get five to six years on a not guilty plea were he to be convicted, the Applicant said that he did not know and that he was not educated in the law.

The Applicant’s Evidence: the Events of 19 October 2005

[35]The Applicant told the jury that he was alone with the twins and Molly woke up “crying like she was in pain”. He said that Molly would not take the bottle. He tried to give her the bottle, but she would not latch on. The Applicant said that he placed her on a changing mat and then went to the toilet. After about two minutes, he returned to see Molly with her back arched, her toes were pointed as if she was stretching, her eyes were rolling in the back of her head, her lips were purple, and she was pale. The Applicant then held the top of her arms with both of his hands and tried to revive her. He agreed that he shook her just for a few seconds in order to get a reaction. It was “a gentle movement backwards and forwards, and not a violent shake.”

The Crown’s Expert Evidence

[36]We propose to summarise only some of the evidence that was called. Transcripts of the oral evidence are not available, but we have been able to study the witness statements and reports of these various experts as well as the Judge’s exemplary summing-up.

[37]Professor Milroy, who had provided a statement in 2005, gave oral evidence in March 2024. He said that there were no documented external injuries, and no suggestion of any soft-tissue swelling. Although often seen in cases of non-accidental injuries of babies, there was no evidence in Molly’s case of either rib or lower limb (metaphyseal) fractures. Professor Milroy agreed that in 2005 this was a pure “triad” case with no external injuries and no other injuries which might support this being an abusive head injury. But the gist of Professor Milroy’s evidence was that it was not just the elements of the “triad” that could prove a traumatic cause but rather their combination with the absence of another explanation. He accepted that in the world of pathology, of which he is part, there is much more scepticism as to whether these three groups of injuries are caused by shaking alone. Finally, Professor Milroy did not accept that Molly’s ventouse delivery could have been the cause of her injuries.

[38]Dr Wright also gave evidence in March 2024. He said that there was no good evidence of pre-incident health problems and he repeated his opinion that these were non-accidental injuries. In reaching that conclusion, he said that he had looked at the whole picture. He agreed that there was no evidence of external bruising and that a full skeletal survey disclosed negative findings. Asked about the findings on CT scan of a larger than expected subarachnoid space on the right side, Dr Wright accepted that a benign enlarged subarachnoid space (“BESS”) caused by excess fluid can increase the risk of a subdural bleed, but he rejected the hypothesis that this space could have been caused by a bleed at birth: there was no evidence of a vestigial membrane. Finally, Dr Wright accepted that tests had not been carried out in 2005 to exclude a number of rare conditions including Von Willebrand disease, Factor VII deficiency and Ehlers-Danlos Syndrome.

[39]Dr Roger Malcolmson, a Consultant Paediatric and Perinatal Pathologist, provided a report in August 2020 concerning the histopathological examination of Molly’s eyes. He concluded that there was no evidence of congenital or other pre-existing natural disease processes affecting her eyes. He found that the eye pathology findings in the case could be summarised as showing phthisis of the left eye, gliosis of the right retina and bilateral atrophy of the optic nerves. These findings were consistent with historic head trauma.

[40]In a further statement in response to a report from Dr Scheller (a board-certified Paediatrician and Child Neurologist in America), Dr Malcolmson opined that he considered it highly unlikely that the pattern of retinal haemorrhages described shortly after events on 19October 2005, associated with a uniliteral peri-macular retinal fold, was the result of a natural intracranial pathological event. In cross-examination, Dr Malcolmson was prepared to accept that “there may be very rare examples” of instances where a peri-macular fold is caused by something other than trauma. Further, the findings described by the treating ophthalmologist at the time of the index incident, while asymmetrical, were otherwise typical of those seen in Abusive Head Trauma (“AHT”) in infants. In his oral evidence, Dr Malcolmson added that the fact that the damage to the eyes was not symmetrical did not help him in determining whether or not this was AHT. Overall, in his opinion AHT would therefore seem to be the most likely explanation for the eye pathology findings.

[41]Dr Daniel du Plessis, a Consultant Neuropathologist, examined at post mortem Molly’s brain, spinal cord and spinal dura. He confirmed the presence of an old, extreme hypoxic ischaemic brain injury caused by sudden unexpected respiratory arrest. In his opinion, this injury happened on 19October 2005 at around 11.00am when Molly suddenly and unexpectedly collapsed into a state of respiratory arrest. He noted that the distribution of this injury (what he called a multicompartment bleeding) correlated with the scans taken at hospital on the day of Molly's admission on 19October 2005. Dr du Plessis opined that the cause of the unexpected respiratory arrest was a traumatic head injury, more specifically a non-accidental head injury event. He noted that the so-called “triad” of injuries was present and provided some discussion as to the controversy concerning whether the “triad” provided absolute proof of non-accidental injury. He noted, however, that the “triad” was not an isolated finding in this case and went on to describe a number of additional injuries that further supported a finding of non-accidental injury. He concluded, looking at the whole picture including the absence of a natural disease process, that the evidence confirmed a traumatic head injury as being the cause of Molly’s respiratory arrest on 19October 2005, and that the collective findings were consistent with a non-accidental event involving forceful shaking with or without a head impact on a soft surface.

[42]Looking beyond the “triad” of injuries, Dr du Plessis said that there was evidence of spinal subdural bleeding which he described as a key feature. He accepted that the blood which had left scar tissue in the spine may well have been due to cranial bleeding, but said that in his opinion it did not matter what the source of the bleeding was. The fact of a spinal subdural bleed is indicative of non-accidental trauma. Dr du Plessis said that there were many research papers, the vast bulk of which lay in the field of radiology, which supported that strong association. In cross-examination, however, Dr du Plessis accepted that there was a need for more research in this area.

[43]On a similar theme, Dr du Plessis also relied on the evidence of bleeding in the nerve roots in the spine. This evidence was not part of the “triad”. These bleeds may have been the direct result of trauma but they may have been caused by the hypoxic-ischaemic injury to the brain. On this second hypothesis, this evidence could not logically add to the “triad”. Dr du Plessis conceded that what he described as tract selective axonal injury in the spinal cord in the neck could have been caused by the global brain injury.

Defence Experts

[44]Dr David Ramsay, a Neuropathologist and Emeritus Professor in the Departments of Pathology at the London, Ontario Health Sciences Centre and Western University, London, Ontario, was also instructed to review the evidence on behalf of the Applicant. He concluded that the cause of Molly’s acute illness was undetermined. It might have been the result of an inflicted injury or, as suggested by some atypical features, an unidentified non-traumatic cause, such as the progressive expansion of a post-natal subdural haematoma and/or subdural bleeding associated with benign enlargement of the subdural spaces. Dr Ramsay accepted Dr du Plessis’ contention that the non-triad features of Molly’s case may provide independent support for the traumatic head injury hypothesis, but he posited other non-traumatic explanations. The lapse of time between injury and death rendered the interpretation of these features more difficult.

[45]Dr Joseph Scheller’s evidence was that Molly did have a pre-existing condition that was apparent on her CT head scan carried out on 19 October 2005. In his view, the scan demonstrated the presence of excess fluid between her brain and inner skull. In his oral evidence, Dr Scheller clarified that this fluid was subdural rather than subarachnoid. It was likely responsible for her head circumference being greater than the 90th percentile at the time. Although Molly could have suffered mild head trauma at any stage during the first three months of life, Dr Scheller’s opinion was that it was chronic and that “the only head trauma we can be sure she suffered was that of birth via a vacuum delivery.” Dr Scheller considered that the most significant injury that Molly suffered on 19 October was a subarachnoid haemorrhage, and that her severe retinal haemorrhage in the left eye could have been a complication of that, brought about by the excess pressure in the subarachnoid space – Terson’s syndrome. Dr Scheller also considered it to be significant that there was no evidence of external marks on Molly’s body when she was brought to hospital.

[46]In cross-examination, Dr Scheller agreed that the presence of all three elements of the “triad” was suspicious, “but you must look for other evidence”. He disagreed with Dr du Plessis’ analysis of the spinal findings. As he put it:

“If someone throws a grenade into a house, 15 years later you can’t say what happened because too much time has elapsed between the grenade going into the house and all the damage being caused and your investigation 15 years later. … To bring a pathologist into the case simply makes no sense when the damage was all caused 15 years earlier.”

The Judge’s Legal Directions

[47]Section 74 of the Police and Criminal Act 1984 (“PACE”) governed the admissibility of the Applicant’s 2006 conviction in the manslaughter proceedings. Under sub-section (3), there was a rebuttable presumption that the Appellant by virtue of his guilty plea was guilty of the offence of manslaughter unless the contrary was proved by him. That meant proof on the balance of probabilities, and no issue is taken with the Judge’s direction on this issue.

[48]The jury were further directed, accurately, that the issue for decision was as follows: “Has the defendant proved that he is not guilty of the offence of inflicting grievous bodily harm, that is say that it is more likely than not that:-a. He did NOT deliberately use unlawful force on Molly on 19th October 2005.

b. At the time of doing so he did NOT have either: i. the intention of causing Molly some injury, however slight; or ii. was NOT aware of a risk that he might cause Molly some injury, however slight, but took that risk; or

c. that the force by him did NOT cause Molly grievous bodily harm (which means really serious harm).” (para 16 of the written legal directions containing the Route to Verdict)

[49]The Judge directed the jury that they must consider all the evidence they had heard and not just the expert evidence. In relation to the medical evidence, the Judge added this: “In this case you have heard a good deal of expert evidence about the so called “triad”. … You have heard that the significance of these findings has been, and continues to be, the subject of vigorous controversy and debate by medical professionals throughout the world. When considering the issues in this case you must take into account all the evidence which you have heard concerning the significance of the medical findings in this case. Given the controversy and debate you will need to exercise special caution when considering the medical expert evidence in this case.” (para 27 of the written legal directions)

[50]In crystallising the issues between the parties arising on the medical evidence, the Judge explained that the Crown’s case was not limited to the “triad” – Dr du Plessis had identified other relevant findings, and its experts had also relied on the absence of a possible alternative cause. It was the defence case that Molly’s brain findings were the consequence of an ongoing complication which had arisen either from the circumstances of her birth or some naturally occurring event which was not diagnosed or capable of being diagnosed. The Judge pointed out: “You will need to be careful to take into account here that current medical science cannot always explain everything and that medical science develops over time.” (para 29 of the written legal directions)

The Authorities

[51]Our attention has been drawn to a number of authorities. We need address only two.

[52]The leading authority is now R v Tredget

[2022] EWCA Crim 108

;

[2022] 4 WLR 62

(“Tredget”), a decision of this Court (Fulford LJ, VP-CACD, Hilliard J and Lord Hughes). The appellant had entered guilty pleas in 1981 to various offences of arson and manslaughter. He appealed on the basis of fresh evidence which was said to cast doubt on the veracity and reliability of his confessions. This Court dismissed the appeal on the facts, but the following two points of principle may be drawn from its judgment.

[53]First, a guilty plea is a public confession of guilt. As a matter of logic and policy, the circumstances in which a guilty plea can be re-opened on appeal are heavily circumscribed. As Lord Hughes (sitting in this Court) explained in R v Asiedu

[2015] EWCA Crim 714

;

[2015] Cr App R 8

(referenced in Tredget at [152]):

“19. A defendant who pleads guilty is making a formal admission in open court that he is guilty of the offence. He may of course by a written basis of plea limit his admissions to only some of the facts alleged by the Crown, so long as he is admitting facts which constitute the offence […]. But ordinarily, once he has admitted such facts by an unambiguous and deliberately intended plea of guilty, there cannot then be an appeal against his conviction, for the simple reason that there is nothing unsafe about a conviction based on the defendant's own voluntary confession in open court. A defendant will not normally be permitted in this court to say that he has changed his mind and now wishes to deny what he has previously thus admitted in the Crown Court. … 31. […] Of course a defendant who is confronted by a powerful case may have difficult decisions to make whether to admit the offence or not. He will of course be advised that if he does plead guilty that fact will be reflected in sentence, but that general proposition of sentencing law does not alter his freedom of choice in the absence of an improper direct inducement from the judge, such as there was in R. v Inns (1974) 60 Cr. App. R. 231. He will always have it made clear to him that a plea of guilty, should he choose to tender it, amounts to a confession. Only he knows the true facts, which usually govern whether he is guilty or not and did so here. If he is guilty, the fact that the choice between admitting the truth and nevertheless denying it may be a difficult one does not alter the effect of choosing to admit it. […].”

[54]Secondly, Tredget identifies three categories of cases in which an appellant may submit that a conviction on the basis of a guilty plea is unsafe, albeit that the list of categories is not “necessarily closed” (see [153]). The three categories are as follows:(1) Category 1: where the guilty plea is vitiated in some way. This can occur in several circumstances, including where the plea is equivocal; brought about by judicial oppression or some other unfair pressure; an incorrect legal ruling; or incorrect legal advice which deprived the defendant of a defence which would probably have succeeded. The question is whether a clear injustice has been done (see [158] of Tredget and R v PK

[2017] EWCA Crim 486

; [2017] Crim LR 716 at [12]).

(2) Category 2: where it was an abuse of process to try the appellant. This category is not relevant here.

(3) Category 3: where it is established that the appellant did not commit the offence; in other words, where the plea is a false one.

[55]Category 3 was described by this Court as a “small residual third category” (see [162]). Its scope requires examination. In R v John Verney

[1909] 2 Cr App R 107

, a conviction for sacrilege was quashed on the basis that the appellant was in prison on the given date and could not have committed the offence. In R v Barry Foster

[1985] 1 QB 115

;

[1985] 79 Cr App R 61

, the subsequent admissions of another man showed conclusively that he and not the appellant had committed the offence the subject of one of the counts, and the Crown conceded the appeal on a linked count. Watkins LJ indicated that the court should intervene in a case of this kind only if the grounds were sufficiently compelling. In R v Noel Jones

[2019] EWCA Crim 1059

, the appeal was allowed against the appellant’s conviction for manslaughter on the basis that later DNA evidence wholly exonerated him.

[56]This Court also examined cases such as R v Lee (1983, unreported) and R v Brady

[2004] EWCA Crim 2230

where convictions were set aside on the application of less stringent criteria. Although these cases were held to have been rightly decided on their own particular facts, the Court identified (at [169] of Tredget) a “significant difficulty” with an approach which did not accord sufficient weight to the effect of a guilty plea as an informed public admission of the offence.

[57]Drawing these strands together:

“171. It can nevertheless exceptionally occur that a reasoned legitimate doubt may be entertained by this court about the verdict reached by the jury following disputed evidence, and this may be sufficient to establish that the conviction is unsafe. But following a freely made guilty plea, the conviction does not depend on the jury's assessment of disputed evidence. The evidence has never been heard, still less tested. It cannot be appropriate to enquire how it might have emerged and might have been assessed if there had been a trial. A submission that the evidence leaves a doubt about the guilt of the defendant is simply inappropriate. In such a case, of a free and informed plea of guilty, unaffected by vitiating factors, it will normally be possible to treat the conviction as unsafe only if it is established that the appellant had not committed the offence, not that he or she may not have committed the offence. Therefore, the test is not that of "legitimate doubt", still less a "lurking doubt", but instead it must be demonstrated that the appellant was not culpable. This is essentially consistent with four of the authorities set out above. In summary, the decision in Verney was based on the court's conclusion that the appellant could not have committed the offence because he had been [in] custody at the relevant time. In Barry Foster, although Watkins LJ did not describe the approach in precisely these terms, he nonetheless set a high test when he suggested that no jury could be sure of the appellant's guilt, adding that the court should only intervene in a case of this kind if the grounds were sufficiently compelling. In Saik, fresh evidence demonstrating the appellant was not guilty of the offence was said to represent a classic example of material that potentially undermined the safety of the verdict. The DNA evidence in Noel Jones wholly exonerated the appellant. 172. As Lord Salmon observed in DPP v Shannon [1975] AC 717 at page 769, "a plea of guilty is equivalent to a conviction", where entered, we would add, by an individual who knows whether he or she committed the offence. It would be wrong in principle for a defendant to be entitled freely to enter a guilty plea, thereby convicting himself or herself, only later to seek to appeal that conviction simply by producing evidence that might have led a jury to doubt his or her guilt if there had been a trial, or by subjecting the evidence which might have been led at trial to a theoretical paper analysis in the absence of the witnesses. The objectionable nature of such a course is demonstrated in the instant case where many features of the evidence have never been and are now incapable of being tested. Therefore, although we consider the decisions in Lee and Brady were no doubt correctly decided on their facts given the strength of the evidence demonstrating the appellants had not committed the offences in question, the test applied by the court in both cases was incorrect. In consequence, with respect to the editors of Archbold, the observation at 7-46 concerning Brady is in our view unjustified and fails to reflect the correct approach. 173. An important common element across the three categories, therefore, is that the circumstances relied on by the appellant need to be established by him or her. That is merely an application of the normal rule that it is for an appellant to demonstrate that his conviction is unsafe. By way of summary, for the first category, the matters vitiating the plea must be demonstrated (e.g. that the plea was equivocal, unintended or affected by drugs etc.; there was a ruling leaving no arguable defence; pressure or threats narrowed the ambit of freedom of choice; misleading advice was provided or a defence was overlooked). For the second category, it must be shown that there was a legal obstacle to the defendant being tried for the offence or there was a fundamental breach of the accused's right under article 6 (whether he or she was guilty or not), and for the third category, it needs to be established that the appellant did not commit the offence. If that standard is not met, we would not expect an appeal against conviction following a guilty plea to succeed.”

(emphasis in original)

[58]It is clear that Category 3 cases are exceptional. As this Court stated in R v BRP

[2023] EWCA Crim 40

at [56], cases can only fall within Category 3 where “a clear injustice has been done”. Tredget is authority for the proposition that, for a case to fall within in Category 3, it is insufficient to demonstrate that an appellant was probably not guilty of the offence to which he pleaded guilty; or, indeed, that he may not have committed the offence. The bar is considerably higher, recognising the status of the guilty plea and the public confession of guilt. An appellant must establish on appeal that he did not commit the offence.

[59]We turn to the earlier decision of this Court in R v Young

[2016] EWCA Crim 1321

;

[2017] 1 Cr. App. R 2

(“Young”) referred to in detail by Ms Lewis in her written Grounds,. The facts of Young were similar to the present case although at police interview the appellant admitted shaking the baby violently for up to a minute. Herargument before this Court was that his acquittal for manslaughter in 2014 was inconsistent with his guilty plea to the offence of inflicting grievous harm in 1999. As here, the guilty plea had been admitted in evidence under s. 74 of PACE.

[60]The apparent factual similarities between Young and the present case end as soon as it is recognised that in Young there was evidence of other fractures and bruising to the child which amounted to grievous bodily harm. Although that evidence was excluded by the trial judge in 2014, the appellant pleaded guilty to an offence which included the commission of these fractures. It followed that there was no inconsistency between the jury’s not guilty verdict (necessarily related solely to the brain injury) and the appellant’s earlier guilty plea.

[61]Thus, as this Court in Young proceeded to explain:

“36. This appeal does not turn on reflecting the evidence from the manslaughter trial back into the circumstances in which the appellant faced his original trial. The evidence is and was different. Thus, Mr Waterman's premise that the appellant was wrong to plead guilty because the medical evidence proved that the ultimately fatal brain injuries were caused by CVST and/or infection and not by his admitted shake does not identify the correct starting point: the question is whether the medical evidence was sufficient properly to justify advice that the appellant had caused serious bodily injury to Michael in circumstances when it was not suggested that anyone else had injured him. Suffice to say that, in the light of the evidence available for the first trial, there is no basis for contending that a defence to a charge of inflicting grievous bodily harm would quite probably have succeeded. There is thus no basis for treating the plea of guilty as a nullity or the conviction as unsafe.”

(emphasis added)

[62]The factual dissimilarities between Young and the present case mean that this authority does not advance the Applicant’s case materially (by way of distinction or otherwise) and Ms Lewis was right not to dwell on it.

The Application/Appeal

[63]Ms Lewis submitted that the Applicant’s conviction by his guilty plea is unsafe because he has now established that he did not commit the offence. As we have set out above, her primary formulation of Ground 4 is that the consequence of the jury’s acquittal in the manslaughter trial is that his earlier guilty plea on the s. 20 charge is unsafe. It was said that the jury’s not guilty verdict amounted to a conclusive and binding finding, made on the balance of probabilities, that the Applicant did not cause Molly grievous bodily harm. Thus, “the jury were satisfied on the balance of probabilities that the plea was equivocal and that he did not cause Molly grievous bodily harm.”

[64]Ms Lewis’ alternative formulation was predicated on an acceptance that each of Grounds 1 to 3 could not succeed when taken in isolation, but, when considered as a package, they amount to a series of exceptional or troubling circumstances which compel the conclusion that the Applicant’s conviction based on his guilty plea is unsafe. Ms Lewis made it clear that she was not contending for a new (fourth) category under Tredget principles. Her submission was that the existing three categories are sufficiently flexible to permit a consideration of a combination of unusual features in the overall interests of justice.

[65]Ms Lewis accepted that the evidence bearing on the circumstances surrounding the Applicant’s guilty plea tendered on 1 June 2006 are unclear, due in the main to the paucity of the Applicant’s recollection and the obvious effects of the passage of time, and the lack of available documentary evidence. However, she drew our attention to the Applicant’s denials of guilt at police interview, to social services, and at the manslaughter trial. Ms Lewis argued that the evidence in 2006 amounted to a “pure triad” case, and that there was no other evidence which unequivocally established his admission of guilt by his plea. Further, his guilty plea was entered on the basis of advice that the medical evidence was stacked against him, and in order to get a shorter sentence. As Ms Lewis put it, in the present case “there was no clear admission of guilt, additional and separate, to the plea entered”, and on that basis the case of Young could be distinguished.

[66]As for the factors relied on under the umbrella of Ground 3, Ms Lewis relied on developments in medical science since 2006. In 2009 there had been a meeting of the Royal College of Pathologists due to disagreements within the field on the appropriate interpretation of various aspects of post mortem findings in so-called “Shaken Baby Syndrome”. A consensus statement was produced which was to the effect that the presence of the “triad” amounts to a prima facie suspicion that injuries are due to mechanical trauma including vigorous shaking; all individual elements fall within a differential diagnosis and other possible non-traumatic causes must be considered and excluded; in the current state of knowledge the presence of the “triad” should not be regarded as absolute proof of mechanical trauma; and certain findings would suggest a need for greater caution. Further, in 2016 a literature review undertaken by the Swedish Agency for Health Technology and Assessment of Social Services concluded that there was insufficient scientific evidence on the diagnostic accuracy of the “triad” in identifying traumatic shaking, and there was limited scientific evidence that the “triad” and its components can be associated with traumatic shaking.

[67]Further, Ms Lewis relied on the CPS’s current charging guidelines following the decision of this Court in R v Henderson, Butler and Odeyiran

[2010] EWCA Crim 1269

;

[2010] 2 Cr App R 24

. Prosecutors are directed that cases involving the “triad” must be approached with caution, and that medical knowledge is never complete and comprehensive. In particular:

“Where the prosecution is able, by advancing an array of experts, to identify a non-accidental injury and the defence can identify no alternative cause, it is tempting to conclude that the prosecution has proved its case. Such a temptation must be resisted. In this, as in so many fields of medicine, the evidence may be insufficient to exclude, beyond reasonable doubt, an unknown cause.”

[68]Ms Oakley on behalf of the Respondent submitted that it has not been established that the Applicant’s confession of guilt in 2006 was other than freely made, and that the jury’s verdict in the manslaughter trial does not prove to the requisite standard that the Applicant did not inflict grievous bodily harm on Molly in 2005. She drew attention to the various paragraphs in the pre-sentence report to which we have already referred. In relation to Ms Lewis’ alternative formulation, Ms Oakley submitted that one way or another the Applicant had to bring his case within Tredget Category 1 (in relation to Grounds 1 and 2) and Category 3 (in relation to Ground 3) in order to succeed, and that he could not.

[69]We are grateful for the quality, clarity and economy of both counsel’s written and oral arguments.

Analysis

[70]We begin with Ms Lewis’ primary submission on Ground 4, namely that the Applicant’s acquittal in the manslaughter trial establishes without more that the 2006 conviction is unsafe.

[71]It is important to bear in mind what the jury’s acquittal in the manslaughter trial did (and did not) establish: it established only that the jury concluded that the Applicant probably did not commit manslaughter (and so logically also the s. 20 offence).

[72]However, for a case to fall into Category 3 so as to render unsafe a conviction based on an (otherwise unimpugned) guilty plea, it must be established that the defendant did not commit the offence. It is not enough to establish that the defendant probably did not commit the offence. Against the background of a valid guilty plea, the relevant premise at this stage, more is needed. As set out above, examples of cases where the necessary threshold is met are those where DNA evidence proves that the defendant must be innocent; or where it can be shown that the defendant could not have been present at the time of the offending.

[73]We add that we do not consider it to be arguable that the jury’s acquittal in the manslaughter trial renders the Applicant’s guilty plea in some way “equivocal” for the purpose of his primary case. That is an issue for us to consider in the context of his alternative submission, and Ground 1.

[74]We turn then to that alternative submission, addressing first each of Grounds 1 to 3 separately, before standing back to consider the position overall.

[75]As for Ground 1, there is no question but that it was the Applicant’s decision, made of his own free will, to plead guilty. He frankly accepted this in cross-examination in the manslaughter trial.

[76]Ms Lewis accepts that the burden then is on the Applicant to establish that the advice that he was given in 2006 was incorrect and that it deprived him of a defence that would probably have succeeded. However, she fairly conceded that it is impossible for there to be any fair or reliable assessment of what advice the Applicant was given. Given the passage of time, no contemporaneous papers now exist; the Applicant’s memory is imperfect; the barrister who advised him in connection with his plea has died; and the different barrister who represented him at the sentencing hearing (the occurrence of which the Applicant does not recall) has no memory of this case.

[77]In any event, even on the Applicant’s account without more, there is nothing to indicate that he was given incorrect legal advice. That issue must be judged with reference to the circumstances and evidence as they presented at the time. The expert evidence in 2006 was, as he was advised, “stacked against him”, and at the time (and arguably even now) there was no alternative explanation for Molly’s injuries consistent with his innocence. Plainly, it was not incorrect to advise the Applicant that a guilty plea would result in a lesser sentence - indeed it was the duty of his representatives so to advise him - and it is (at the least) doubtful whether the Applicant could have been advised that he would receive five to six years’ imprisonment following conviction at trial (when the maximum sentence was five years’ imprisonment) and only nine to twelve months’ imprisonment following a plea. Further, it is difficult to see how the Applicant was deprived by any incorrect legal advice of a defence which would probably have succeeded, as matters stood in 2006.

[78]As for Ground 2, there is no need, either in line with Tredget or general principle, for there to be other evidence unequivocally establishing the Applicant’s acceptance of guilt at the time in order for a conviction based on a guilty plea to stand. It is the Applicant’s public confession of guilt that is the focus. In any event, the matters relied on are at best neutral. Thus, there are passages in the pre-sentence report which point clearly to an acceptance of guilt by the Applicant.

[79]In short, neither Grounds 1 or 2, either alone or together, arguably undermine the safety of the 2006 conviction.

[80]As for Ground 3, it is true that medical science has evolved since 2006 and that the evidential picture is now more nuanced. However, these subsequent developments do not demonstrate that the Crown’s case in 2006 (that this was a case of non-accidental shaking) was wrong. The changing evidential landscape since 2006 might be capable of raising new doubts as to the mechanism of injury, but that is far from being enough to unsettle the conviction. Ms Lewis realistically accepted that a jury properly directed could have convicted the Applicant of manslaughter in 2024. The development of scientific understanding since 2006 falls well short of establishing that the Applicant did not commit the s. 20 offence.

[81]For these reasons, none of Grounds 1 to 3 considered individually afford any proper ground for concluding that the 2006 conviction is unsafe.

[82]We have then stood back and considered them cumulatively in the context of the Applicant’s acquittal in the manslaughter trial. We have been persuaded that it is arguable that, taking account of the Applicant’s acquittal in the manslaughter trial, alongside all of the circumstances identified under Grounds 1 to 3, the 2006 conviction is unsafe.

[83]However, we have reached the clear conclusion that it is not.

[84]The Applicant’s guilty plea in 2006 was entered into of his own free will. It has not been established that he was given any incorrect legal advice at the time, or that he was thereby deprived of a defence that would probably have succeeded. Neither developments in medical science since 2006, nor his acquittal in the manslaughter trial, establish that he did not commit the s. 20 offence. No clear injustice has been done.

[85]It follows that there are no good grounds for finding that the Applicant’s conviction is unsafe.

Disposal

[86]Having considered the Applicant’s alternative case under Ground 4 to be arguable, we grant the necessary extension of time, leave to vary and leave to appeal. It is in the interests of justice to do so. However, for the reasons set out above, the appeal must be dismissed.