Although we may not be privy to each and every element presented during proceedings, as scientists we are able to identify the scientific standards and research relating specifically to the scientific elements included in the case. Additionally, we can identify the proper methodology that should be employed when conducting a scientific investigation based upon clinical medical notes. A significant part of this type of scientific consultation involves extensive reviews of patient medical records, along with clinical reports of infants and children with rare diseases. This work necessitates the adoption of rigorous scientific investigative principles and protocols, along with an understanding of problematic variables such as observer bias, in order to yield a report that can be acted upon by the clinical team. Unlike a forensic investigation, such approaches have serious consequences to the health of specific patients if the conclusions made are erroneous or are scientifically unsubstantiated. Even where such an investigation involves patients with the same underlying conditions, great care is taken to recognise the significant influence of non-genetic factors which may alter outcomes. 

The claims made by the experts in this case were reviewed and the references upon which they relied to make such claims were sourced, analysed and cross referenced with the body of literature on the related topics. Based upon these factors, it is quite apparent that the quality of the scientific evidence is deficient and scientifically unfounded. For example, Dr Evans’ primary assertion is that some of the infants were harmed by air embolism, where air was deliberately injected into a vein, or through the stomach. Peculiarly, Dr Evans relies on a research paper from 1989 dealing with gas embolisms, which occur through the use of high ventilation pressures in preterm neonates. The cause of death due to gas embolism from high pressure ventilation differs substantially from the cause of death due to air embolism. It is never properly explained how the article Dr Evans references could ever have been used as a basis for describing air embolism using ambient air with 21% oxygen introduced through various tubes. The paper Dr Evans references describes air embolism caused by high pressure 100% oxygen being delivered to the lung with such force it caused an air leak in the lung. When demonstrating the symptoms of a given physiological state, it is necessary to demonstrate this state in multiple peer reviewed journal articles. There is no evidence supporting the finding of air embolism in any of the infants, as the expert witnesses rely on a journal article depicting gas embolism and not air embolism.

Not a single expert appears able or willing to present any similar case involving a comparable setting, patient population and set of symptoms as scientific certainty. Given that in the case of every infant, the medical records report ‘possible sepsis’ and the symptoms described point to an infection of some nature, it is shocking that no infectious disease expert has been called upon to review these cases. In the absence of such an expert, there should be a proper explanation detailing why no investigation surrounding infectious diseases took place. This evaluation would require at minimum an assessment of the possible infectious disease pathogens that pose a risk to neonates during the perinatal period, where infection is associated with sudden unexplained neonatal collapse/death. This evaluation should be accompanied by an assessment of the factors relating to sepsis for each child and why an infectious disease was ruled out.

The indictment centres around claims from co-workers of the defendant, which assert the infants were stable prior to collapse. Where such statements are made and are based on clinical insights they must be defined, and the definition must be in present day use, and be widely applicable across the patient population being studied. In a medical setting, the absence of physiological phenomena equates to the term "stable". Thus, considering that a diagnostic test establishes parameters which define clinical significance, the use of terminology such as "stable" should be reflected by the absence of clinical issues in the patient’s medical notes. One does not assess patients on the basis of whether they are "stable". Clinical diagnosis is reached based on whether the patient exhibits signs of clinical instability. Despite this clear discrepancy, all the medical professionals who testified for the prosecution repeatedly referred to neonates being “stable”, “very stable” or “stable enough.” Indeed, it was stated that a child who was born at 23 +6 weeks gestation was “very stable” prior to experiencing a bout of sudden projectile vomiting. This child had apparently not undergone proper MRI brain scanning and was receiving a combination of Gaviscon, breast milk and formula via nasogastric tube. The condition of the infant does not reflect that of clinical stability where the child is unable to feed independently and where there is no relevant assessment of development of white matter brain injury, which is present in up to 50% of very low birth weight infants (Romero-Guzman & Lopez-Munoz, 2003; Agut et al., 2020).

In the present case, those infants who died did so after a period of excessive cardiopulmonary resuscitation, that exceeded the ILCOR guidelines. There is no discussion surrounding findings that preterm babies (born at <37 completed weeks of pregnancy) have a higher mortality, morbidity and risk of impaired motor and cognitive development in childhood than babies born at term (Heino et al., 2016). This is despite the fact that seventeen out of 18 infants described in the indictment were born before 37 completed weeks of gestation. Moreover, in England and Wales, the incidence of preterm birth is higher in multiples, where ~50% of all multiples are born at a gestational age of less than 37 weeks. Taking the infants included on the indictment, we find that eight of the eighteen infants were multiples, of which four infants were very preterm (<32 weeks gestation) and the other four were moderately preterm (32 – 34 weeks gestation). Among the single births, three infants were both extremely pre-term (<28 weeks) and extremely low birth weight (<1000 g). In England and Wales the incidence of perinatal death for very preterm infants in 2015 stood at ~ 8%.  

In three cases involving infants who died, it was reported that they were removed from ventilation support in the immediate period before death. There appears to be no clinical explanation for the removal of ventilation. Instead, there are reports within the notes of a deterioration in the hours prior to removal. There is no criticism of any of the medical treatment rendered to these infants despite the RCPCH report from that same time detailing that the consultants only held two ward rounds per week, the ward had insufficient senior cover, and there was a reluctance to seek tertiary level advice and escalate concerns in a timely manner. Rather, both the expert witnesses, Dr Evans and Dr Bohin, represent the conduct of the medical professionals responsible, for the patient’s care, as a reflection of the clinical standard. In many cases, the approaches and actions of the clinical team can only be described as inappropriate and demonstrative of an abject failure to apply the principles of diagnostic work-ups. Several infants are noted to require a lumber puncture, presumably to ascertain risk for meningitis. This does not ever occur because the infants die before the procedure is performed (see, Child C and Child D). In a separate incident, both the consultant, Dr Brearey, and the registrar, Dr Ventress appear unable to determine the cause of desaturations in an infant while on ventilation. Both clinicians determine that the cause of the desaturation must be equipment malfunction. Once the ventilator is changed and the same phenomena recurs, they appear to take no diagnostic steps other than to remove the child from the ventilator and reintubate her five minutes later. It is not clear whether the infant is being intubated prior to analgesia or muscle relaxant infusion, though that is apparently the case.

There appear to be significant inconsistencies with regard to ventilation strategies and approaches. Despite the fact that morphine 100-mcg/kg bolus followed by 10 mcg/kg/h continuous infusion for 7 days or less (median duration of exposure in the treatment group was 77 h) showed no detrimental long-term neurological effects (Simons et al., 2003), there were reported claims of infants fighting the ventilator and this was used to justify their removal from ventilation. There was little investigation as to why this phenomena was apparent in very preterm infants, and in one representative case, Dr Gibbs removed an infant from the ventilator and she died ninety minutes later (Child I).  Removal from the ventilator was not discussed as a factor in assessing the cause of death.

There is no discussion surrounding the predicted outcomes for the population of neonates in the indictment. For example, the experts fail to reconcile the finding that up to 50% of extremely preterm infants fail extubation, even when extubation criteria are met, such as the level of ventilator support and respiratory function parameters. In a study of 3343 extremely low birth weight (<1000 g) neonates who received mechanical intervention, 2867 (85.8%) survived to discharge. Mortality was associated with exposure to a greater number of mechanical ventilation courses (Jensen et al., 2015). In this case, three of the infants are classified as extremely low birth weight, however there is little discussion of the unique physiological circumstances confronting such infants, and whether mechanical ventilation represented a particular risk factor in terms of their clinical prognosis.

Much weight is given to the coincidental element that Ms Letby was present for some of the events characterised as ‘sudden collapse.’ Records not presented to the court, nor referenced by the experts, reveal a pattern of stillbirths and perinatal deaths that for a five year period follow the exact same trend. Indeed, there were a significant number of stillbirths that occurred alongside the perinatal deaths in the period between June 2015 – November 2015. When establishing a cause of biological phenomena, it is not permitted to manipulate or massage the data. Thus, the greatest insight into the “sudden collapse” of the infants in this case required a thorough review and comparison of the symptoms, and circumstances of all the infants on the ward during that same period. It does not appear that any such review took place.

The absence of expert evidence could be viewed as demonstrative of the unique nature of the allegations made in this matter. The problem with this idea, though, is that the experts have failed to report on actual evidence that directly opposes the claims they have made, and in some cases this shortcoming reveals a failure in either expertise or honesty. For example, Professor Hindmarsh testifies that there are no studies detailing the adsorptive properties of insulin, which is the tendency for insulin to adhere to most any surface. This is untrue. There is a significant body of research relating to this phenomenon, and upon review it shows that such adsorptive properties of insulin to the venous lines can result in a decrease of insulin delivery by as much as 70%. There is further evidence that insulin forms bonds to the bag delivering the ternary parenteral nutrient (TPN) solution and/or dextrose saline solution, and this can reduce the delivery of insulin by as much as 60%. The increased binding to plastic surfaces is compounded by research which reveals that insulin added to TPN and/or dextrose solutions is unstable and results in a decreased bioavailability of insulin by ~40%. This finding persists over at least 24 hours. Professor Hindmarsh fails to make any of this important and relevant empirical evidence available to the jury, and fails to give it any consideration in his testimony.

Gestational diabetes is associated with hyperinsulinemia The production of maternal antibodies to insulin in response to maternal insulin treatment can result in insulin readily crossing the placenta. It was identified that 24% of umbilical vein cord blood contained non-human insulin, demonstrating the transfer of exogenous insulin from the mother to her infant in utero (Lindsay et al., 2004). A later study showed that more than half of the infants of mothers with insulin-dependent diabetes have maternal insulin autoantibodies (mIAAs) at birth, and the close correlation between the mIAA levels in the newborn infant and those in the maternal circulation verifies the claim that the IAAs in cord blood represent transplacentally transferred antibodies. These mIAAs could form complexes with the infant’s endogenous insulin, thereby prolonging the half-life of insulin. Pregnancy, in general, induces non-immunoglobulin transfer of maternal insulin into foetal circulation, and this tendency is increased at the time of delivery in both diabetic and non-diabetic mothers (Ronkainen et al., 2008).

Despite the fact that Professor Hindmarsh is an Emeritus Professor, based at University College London Hospital (UCLH), there is nothing to suggest that he has any involvement with hypo/hyperglycaemia pathways in preterm neonates at either UCLH or Great Ormond Street Hospital. Nor does he possess any advanced or specialist knowledge in the assessment of assays used to test C-peptide and insulin. It is provided that “The expert must be able to provide impartial, unbiased, objective evidence on the matters within their field of expertise. This is reinforced by Rule 19.2 of the Criminal Procedure Rules, which states that an expert has an overriding duty to give opinion evidence which is objective and unbiased.” Where a witness refers to a single blood test result and suggests that there is only one way the result can occur, which is through poisoning, this must be more than the expert’s simple opinion, especially in light of the significant body of evidence and empirical data showing that such a claim by a witness is strongly associated with wrongful convictions (Marks, 1999). Professor Hindmarsh stated that a blood test with an insulin concentration of 4657 (units not given pmol/L or mU/L) and very low c-peptide could only occur due to exogenous administration. Remarkably, he then leaps to a conclusion that this insulin must have been administered via dextrose/TPN solutions. This is a stunning claim to make, not least when a concentration of insulin of 4657 pmol/L or 4657 mU/L, would kill two grown men. Yet the infant, who was both very low birth weight and very preterm, recovered without any sequelae in just a few hours.

Professor Hindmarsh fails to provide examples of such discordant insulin/c-peptide levels in clinical case studies, despite the fact that such evidence does exist. Villaume et al (1982) detailed a case of spontaneous hypoglycaemia in which very high plasma insulin (18000 pmol/L) but low normal plasma C-peptide levels occurred. This case could not be attributed to exogenous insulin administration as it otherwise would have been. The mechanism in Villaume's case is unknown: the authors postulated that reduced removal of (endogenous) insulin by liver and peripheral tissues but normal removal of C-peptide produced the observed results. Such a finding likely occurred in the case of Child F, where he presented with increasing signs of jaundice and increasing levels or creatinine and urea, which results in shifts in the glomerular filtration rate in the kidney.  Where both liver and renal function is disturbed, the concentration of insulin in the blood would be increased owing to a reduction of insulin metabolism and breakdown occurring via hepatic and renal routes.

With regard to Child L, the fact that studies demonstrate 50% of infants born to mothers with gestational diabetes have mIAAs in their circulation at birth (Ronkainen et al., 2008) suggests that such a scenario must be considered when hypoglycaemia is identified at birth, as was the case for Child L. The presence of mIAAs in the neonate’s circulation would explain the diminished response to dextrose infusion. The availability of insulin bound to mIAAs would be dependent on the affinity of the antibody for insulin. Moderate to high affinity mIAA binding would result in a pool of insulin being released slowly over time, resulting in persistently low blood sugar. Once all the insulin is released from the antibody complex, the antibodies will be degraded and hypoglycaemia is unlikely to recur. This phenomenon has been observed in several autoimmune conditions.

There is no way to effectively understand what the evidentiary standard was in bringing this case to trial. Based upon the lack of direct scientific evidence, it is apparent that it was not brought based on the findings of a forensic investigation. Still, we believe that in the face of the multitude of experts insisting on the outcomes arising from deliberate harm, it would be an act of recklessness to fail to call for a trial on the issues. The expert witnesses appearing before the court did so in full knowledge that the scientific standard they applied to this case falls far below that which is required of any person who asserts that their expertise is grounded in experimentally-derived biological science. It is particularly problematic that the experts overrode the presumption that the infants died of natural causes, given that 6 of the infants underwent an autopsy where the cause of death was determined to be natural. In its simplest form, before one can even begin to conduct a murder investigation and hear evidence derived from that investigation, there is a requirement to provide an explanation for the findings made at autopsy. This marks the initiating step in deciding as to the claim asserted by the medical doctors, when they reported to Cheshire Constabulary that Ms. Letby is England’s most prolific female child murderer. In order to even conduct such an assessment, one would at the very least need to re-examine the bodies.

Remarkably, the consultant, Dr Gibbs claimed that he suspected Ms Letby of murder in 2015. He not only failed to properly lodge this complaint, but he also failed to ensure the proper collection of serum and blood samples after the death of the patients. Additionally, it appears that Dr Gibbs failed to notify the coroner that he had firm suspicions that the infants were the victims of murder. It is also the case that the hospital failed to lodge the deaths with the Child Death Overview Panel, which would have conducted an independent investigation at that time. Despite claiming he held concerns beginning in June 2015, Dr Gibbs failed to take the necessary steps to ensure the proper preservation of the body after death, permitting the heating, bathing and holding of the infants for hours after the loss of all vital signs of life. The treatment of the bodies after death, combined with the lack of blood and serum samples collected at the time of death compounds the inherent difficulty in determining the cause of death. These shortcomings mean that there is no appropriate refutation of the autopsy findings, and nor can there be, given that the bodies were not exhumed for re-examination.

When confronted with a probable rare disease, there is a certain approach one must take to determining whether there is one primary condition which leads to diverse, partially overlapping outcomes, or whether there are wholly distinct entities at play. It appears the medical experts failed to determine either of these possibilities. There is no proper analysis of the overlapping symptoms, the time of appearance, nor any effort to characterise reported physiological phenomena. This is despite the emerging theme surrounding the instability of the infants concerning the fact that the vast majority were very premature infants (<32 weeks gestation) or extremely premature (<28 weeks gestation) and that such infants have inherently unstable autonomic nervous system reflexes.

In every case what is being described is the interplay between the inherent instability of the infants and some other force. That other force may have been a viral infection or some other birth complications. There is circumstantial evidence, based upon the symptoms described and onset of the those symptoms, that strongly implicates a viral infection as playing the primary function in the death and destabilisation of the infants. We are not seeking for any person to simply take us at our word. Rather, Science on Trial was formed to enable individuals who are interested in the science behind this case to gain an understanding of the manner in which the British criminal justice system wilfully and deliberately disregards basic tenets of the scientific method. This is achieved through the invitation of unqualified medical doctors, with limited experience in scientific investigations to conduct complex forensic examinations with little more than radiographs, and essentially no toxicology, blood testing or genetic data.

There is great concern that the scientific claims made by the experts are not grounded in logical scientific reasoning and that the individuals who carried out the investigation were not simply deficient in their approaches, but further, that they went against basic scientific investigative standards. The consideration of such claims poses a real threat to the integrity of both the role science plays in society, and its role in the delivery of justice within the legal system.

It has taken many hours to put together this material and I could produce reams more. What is presented here is a fraction of the information this case brings up for a scientist. It is of great concern that by reducing the deaths of the infants concerned into a spectacle of confusing and dubious claims, we are effectively undermining the functioning and integrity of our legal system. We cannot as a group of people, however committed, make a determination of guilt on the back of scientific claims that are totally baseless. I fear this is precisely what has happened in this case. In an ideal world, only the highest quality evidence would reach the court. Nobody outside the realm of science can reasonably be called upon to judge the scientific basis of claims made by individuals asserting they are experts, but whose claims are purely speculative and unsupported by peer-reviewed scientific evidence. It falls on us, as scientists, to maintain the integrity of our discipline by speaking faithfully to that which our fellow scientists present in peer reviewed studies. I hope this information will be received as a good faith effort to ensure truthfulness in science.