An Overview of Air Embolism
During the trial of Lucy Letby, the expert witness Dr Dewi Evans, testified that all the infants in the indictment died due to air embolism. Briefly, air embolism consists of an interruption in the circulatory system by bubbles of air (or other gas) that gain access to the vasculature. The air bubbles that enter the circulation enter the heart through the vena cava (superior or inferior), which are the veins that return deoxygenated blood to the heart. Once the air bubbles reach the heart, they become trapped in the right ventricle, due to blockage in the pulmonary artery. The air blocking the pulmonary artery means that the blood. cannot move through the heart to the lungs, where the CO₂ in the blood is exchanged for O₂. The mechanism of death in air embolism is 'pump failure' of the right side of the heart. This event is observed as a sudden collapse, and results in air filling the the vena cava and the right side of the heart, which causes the blood to become frothy, and further limits the heart from pumping blood. Air embolism happens rapidly, in a number of minutes.
Dr Evans based his hypotheses surrounding air embolism on x-ray images collected at different times following the death of the infants. There are no recent findings, identified in the scientific literature, that support the use of post-mortem x-rays to determine air embolism as a cause of death. Though, that is the primary evidence relied upon in the Lucy Letby trial. Amplifying this flawed approach, the histopathologist, Dr Marnerides testified that in the case of Child A that on his review of tissue samples taken from the neonate, that he observed globules of air in the veins in the brain tissue. The expert stated that this air 'most likely went there while this baby was alive."
The claim made by Dr Marnerides is quite exceptional, as in order for the veins in the brain to contain air bubbles the entry of the air embolism could not have occurred in a peripheral vein somewhere else in the body. This is because an air bubble that enters a vein will be returned to the heart first, and will not make its way to the brain. There is no way for the air bubbles to get to the brain via air embolism. This finding is described by Bajanowski et al., 1999, where it is stated that identification of air in the pial veins of the brain are a known artefact associated with death and cannot be attributed to air embolism.
Criminal Practice Directions and Limitations on Expert Evidence
In England and Wales, the legal authority governing the reliability of expert opinion is given by Criminal Practice Directions CPD V Evidence 19A Expert Evidence 19A.5. This Practice Direction strongly indicates that the opinion of the experts in the trial of Lucy Letby were inadmissible. The methods employed by the prosecution experts, in reaching their opinions, did not adhere to the most basic standards represented by the scientific field under consideration. The section relating to expert evidence states that:
“factors which the court may take into account in determining the reliability of expert opinion, and especially of expert scientific opinion, include: 'whether the expert's methods followed established practice in the field and, if they did not, whether the reason for the divergence has been properly explained.’”
The methods employed by the expert witnesses were not 'established practices' adopted by pathologists when called to make a determination of air embolism as a cause of death. Further, at no point did any expert provide a suitable rationale which would justify their departure from the 'established practices.'
The standard bearer for making a determination of air embolism as a cause of death is described in the seminal 1999 paper: "Proof of Fatal Air Embolism" by
Bajanowski et al., 1999. This paper sets forth that air embolism can only be described by a specific autopsy technique and key findings must be obtained (Figure 1).
Recently, air embolism has been reliably identified, post-mortem, using one of two techniques, Magnetic Resonance Imaging (MRI) and Computed Tomography (CT). Both techniques take multiple 2D images of the chest at different depths, and these images are then compiled into one 3D image. The presence and quantity of air can be determined by an algorithm that detects spaces containing air. This is an alternative method, but is still not wholly reliable, when compared to the autopsy methods. (Jackowski et al., 2004)
The Symptoms of Air Embolism were not Associated with the Sudden Collapses
Dr Evans' reliance on x-rays to ascribe air embolism as a cause of death should not have been tolerated in the trial. The presence of air bubbles observed in the x-rays and tissue samples from the neonates, post-mortem, could have occurred for myriad reasons. Moreover, other symptoms associated with air embolism were not present at the time of the collapses. One of the first signs of a venous air embolism is a sudden decrease in end tidal CO₂. End-tidal carbon dioxide (ETCO₂) refers to the concentration of carbon dioxide (CO₂) that is exhaled at the end of an outward breath. This reduction in CO₂ occurs due to the air bubble blocking the pulmonary artery, and restricted blood flow to the lung. The result of these restrictions is decreased CO₂ exchange, manifesting as a drop in ETCO₂ levels. The shift in ETCO₂ occurs in the moments prior to collapse. In the case of Child O, at 1:15 pm a doctor recorded that the neonate's heart rate was elevated, at 160-170 bpm, and that he had raised CO₂ and lowered O₂ arterial blood gases.
It is important to note that in the neonate raised CO₂ is directly related to tachycardia (increased heart rate). The newborn heart does not respond significantly to decreasing blood oxygen levels. However, shortly after birth, it becomes highly reactive to increasing levels of blood carbon dioxide. This heightened response to CO₂ serves to increase the heart rate, enhance blood pressure, improve blood flow, and increase oxygen delivery, helping to prevent a drop in circulation during instances of oxygen deprivation (Cohen et al., 2012). It appears that the cause of the tachycardia described during the trial, was in part due to increased CO₂ levels in the infant. The cause of this increase is unlikely related to an event such as air embolism. This is because air embolism is a rapid event which occurs in the order of minutes. For Child O, he did not experience a cardiac collapse until 4:15 pm, which was three hours after a doctor recorded the increased CO₂ in his blood.
It is entirely likely that the failure to evaluate other major causes of collapse in neonates was the primary failing in these cases. Despite that a core element of good medical practice is the exclusion of differential diagnoses the experts failed to explain how they ruled out other major causes of death, which might appear superficially similar to air embolism. Separate findings reveal that septic shock is a major cause of death in preterm neonates. In particular, early-onset sepsis with symptoms occurring within the first 72 hours after birth can be particularly deleterious to life. All but one of the infants that died exhibited initial symptoms within the first 72 hours of birth, and all were suspected of having sepsis but did not test positive for bacterial cultures. It is not known whether the clinicians at CoCH tested the infants for viral infections, though up to 42% of all cases of sepsis are thought to be due to viruses (Lin et al., 2018). Septic shock is included on the differential diagnosis for air embolism:
Acute myocardial infarction
Heart failure
Tension pneumothorax
Septic shock
Ischemic stroke
Air Embolism and its Association with CPR in Preterm Neonates
Aside from these other diagnoses the cause of the air in the vasculature could be as benign as occurring simply as an artefact relating to CPR. This claim is supported by a recent publication demonstrating that there is a high prevalence of air embolism in neonates born prematurely (< 32 weeks gestation), when they undergo cardiopulmonary resuscitation (Halbertsma et al., 2015). Similarly to the cases at CoCH, CPR was also associated with increased mortality, as 80% of the neonates in the study died due to the original cardiovascular collapse.
In their study, the researchers reviewed cases where CPR was administered both at delivery of the neonates and as patients on the NICU. It is thought that premature neonates are at greater risk from air embolism owing to their lung tissue being fragile and prone to air leakage during chest compressions. Air embolisms were detected on post mortem CT/MRI in n = 8 (89%) patients, and intracardiac air embolisms were found in all of these patients. Contrary to the claims made by Dr Marnerides air was observed in the arteries of several organ systems, including the brain, but no air bubbles were identified in the veins of the brain.
The researchers also demonstrated that the air embolism described in these neonates was specific to CPR and was not the cause of their death. Autopsies were performed on four of the neonates, along with a postmortem CT/MRI. In only one of these patients were the findings of a CT air embolism confirmed on autopsy. This suggests that the appearance of air embolism is an artefact of CPR. Given that postmortem investigations are usually limited to autopsy, the disappearance of intravascular air as observed in these cases may explain why such a phenomenon has not been reported elsewhere. If air embolism were the primary cause of the cardiac event necessitating CPR, it would be expected that the autopsy would support the evidence of air observed in the CT imaging. These findings strongly suggest that the appearance of air in the bodies of the infants was related to CPR, not air embolism as described by Dr Evans.
The Case Law surrounding the Expert's assertion that cause of death was due Air Embolism
The primary issue surrounding the expert witnesses' claims about air embolism is that while the experts could demonstrate that there was air in the vasculature through x-rays, this does not mean it is proof of air embolism. There could be multiple reasons for air appearing in the vasculature on an x-ray. These reasons could be unrelated to an air embolism, such as procedural interventions, post-mortem changes, or even x-ray artefacts. Jumping to the conclusion that the presence of air in the vasculature on an X-ray image is definitive proof of an air embolism without considering or eliminating other possible reasons is a significant overreach.
The conduct of the experts in identifying air embolism as a cause of death, from x-rays can be compared to R v Reed and Reed [2009] EWCA Crim 2698. In the Reed case, the Reed brothers were accused of murdering Peter Hoe based on DNA evidence and circumstantial clues. The DNA evidence was derived from low copy number (LCN) profiling on alleged knife handle fragments. While there were initial challenges to the LCN process's validity, the main concern centered on the expert’s interpretation of the DNA evidence, specifically about how it might have transferred.
The Court of Appeal ruled that while most of the DNA evidence was admissible, the expert went beyond her expertise in certain conclusions she drew. In such cases, a jury's decision is not solely based on expert testimony but integrates various types of evidence. The expert's role is to aid the jury in understanding scientific evidence, but they should not integrate it with other evidence themselves.
The Court of Appeal also advised that experts are required to focus on how scientific evidence supports competing hypotheses without commenting on their probabilities, especially when combining with other evidence. In the Reed case, the expert overstepped these boundaries. This was further bolstered by reference to the Turner case (see R v. turner [1975] QB 834), which set the precedent that experts should not interpret evidence that jurors can decipher themselves. In Turner, the court found that the jury was asked to assess the motivation behind Turner's alleged crime using expert analysis of his psychological state. While the expert's analysis was considered relevant, it risked complicating the jury's assessment and potentially being given undue weight because of the expert's qualifications. The Court of Appeal concluded that experts should be wary of overcomplicating matters with jargon or giving opinions that may mislead the jury due to their professional stature.
When evaluating Reed-type cases in the context of expert testimony and expert interpretation of evidence evaluation in court, they can be summarised as follows:
Reed-type cases: These are cases where the validity of an expert's inference at the activity-level is questioned.
Activity-level vs. Source-level proposition:
Source-level proposition: Refers to the basic evidence that an expert presents. For instance, in a forensic case, the source-level might be about the match between a fingerprint found at a crime scene and the fingerprint of a suspect.
Activity-level proposition: Is an inference drawn from the source-level. Using the fingerprint example, an activity-level proposition might be that because the fingerprints match, the suspect was present at the crime scene.
Validity:
The source-level evidence is typically not disputed in these cases. This means that both parties generally accept that the method used to gather the evidence was correct and reliable.
The contention arises over the validity of the activity-level inference. The "evaluative validity" is about whether the inference drawn from the source-level evidence is valid and reliable.
Issues Raised by Reed-type cases:
Reliability of evaluative inferences: It questions if the conclusions drawn by the expert from the evidence are dependable.
Province of the jury: There is a concern about whether the conclusions drawn by experts are intruding into the decision-making space that should be reserved for the jury. Essentially, are experts making conclusions that should be left for the jury to decide?
Probative value: The expert's claims raise the question of when an expert's inference has enough value to be relied upon in conjunction with other evidence to prove a fact. "Probative value" refers to the ability of a piece of evidence to prove something that's important in a trial.
A Basic Takeaway, in the event nothing else made sense
In simpler terms, "Reed-type cases" involve situations where the basic evidence provided by an expert is not disputed, but the conclusions they draw from that evidence are. This raises questions about the reliability of those conclusions, whether experts are overstepping their bounds, and how much weight should be given to their conclusions when deciding a case. In the trial of Lucy Letby, multiple expert witnesses freely made reference to x-rays depicting air bubbles in the vasculature and other body parts, and without acknowledging that such findings can be generally explained, repeatedly asserted that air embolism was the most probable cause of the air observed in the x-rays. Not only is such a finding not possible to determine from an x-ray alone, but it defies the 'established practices' required to determine air embolism as a cause of death.
References
Bajanowski T, West A, Brinkmann B. Proof of fatal air embolism. Int J Legal Med. 1998;111(4):208-11
Cohen G, Katz-Salamon M, Malcolm G. A key circulatory defence against asphyxia in infancy--the heart of the matter! J Physiol. 2012 Dec 1;590(23):6157-65.
Halbertsma F, Mohns T, Bok L, Niemarkt H, Kramer B,
Prevalence of systemic air-embolism after prolonged cardiopulmonary resuscitation in newborns: A pilot study. Resuscitation, 2015;93:96-101
Jackowski C, Thali M, Sonnenschein M, Aghayev E, Yen K, Dirnhofer R, Vock P. Visualization and quantification of air embolism structure by processing postmortem MSCT data. J Forensic Sci. 2004 Nov;49(6):1339-42
Jeng MJ, Lee YS, Tsao PC, Soong WJ. Neonatal air leak syndrome and the role of high-frequency ventilation in its prevention. J Chin Med Assoc. 2012;75:551–9.
Lin GL, McGinley JP, Drysdale SB, Pollard AJ. Epidemiology and Immune Pathogenesis of Viral Sepsis. Front Immunol. 2018 Sep 27;9:2147