Eight years have passed since the death of the first neonate Lucy Letby was convicted of murdering, while working as a neonatal nurse at the Countess of Chester Hospital (CoCH). In that time much has occurred in the world of medical science, and healthcare. In England and Wales, some 40% of all maternity units are deemed substandard, and perinatal deaths are on the increase for the first time in seven years. However, these findings were given no consideration in the trial of Lucy Letby.
Though, a benchmark event occurred in 2020, which had a marked shift on mortality rates in the UK, and the wider world, more generally. This particular event involved an increase in patient deaths in a hospital setting, and for specific at risk populations. It might be worthwhile to consider the circumstances of this event when attempting to reason about the increased mortality rate observed at CoCH in the period 2015-2016, but I will get to that later.
What is identified as an unusual occurrence in 2015-2016, quickly becomes something sinister if isolated from a comparison of other similar occurrences. Superficially, we might consider that deaths of neonates, in a hospital setting, are an alien concept. We might, as individuals who rarely attend neonatal units, assume that babies scarcely die in NICUs. This assumption might lead us to assume that where there is a cluster of deaths on a NNU this may point to something nefarious within the hospital itself.
The infants in the indictment had scores of overlapping features that instead of being used in a rational, scientific manner were actually used as cast iron proof that a human hand was acting to harm the infants on the unit. Taking those factors into account we find the following patterns:
A higher ratio of male:female deaths.
Morbidity disproportionately associated with the early neonatal period (1-7 days postnatal) versus late neonatal period (8 - 28 days postnatal), six early neonatal deaths : one late neonatal death.
Correspondingly high incidence of stillbirths clustering around the same period as the perinatal deaths.
Seasonal variation which repeats the following year, where the initial cluster appears in June 2015, another cluster appears the following year at the same time in June 2016.
A tendency to multiple pregnancies, e.g. twins, triplets.
Altered physiological observations occurring in the immediate perinatal period, which worsen with time
Reports from treating doctors that the infants exhibited signs of sepsis, but where no bacterial cultures were identified.
Physiological manifestations which yield few unnatural mimics, including internal bleeding, altered body temperature, increased jaundice, and skin manifestations.
Outside transfers associated with onset of patient symptom clusters which result in abnormal neurological events, (Child G, I, J).
Increased presentation of sudden illness, associated with night-time events.
This pattern is not unique to the neonatal unit at the Countess of Chester Hospital. Though, it is the failure by the expert witnesses to recognise the clinical patterns listed above and attempt to relate them to real world events. These patterns are all associated with sudden unexplained early neonatal deaths or collapse (SUEND). What separates these events from those oft reported cases in the media is that these deaths and collapses were occurring in a neonatal unit.
It is not a new phenomenon that individuals struggle to tolerate the concept that there exist all numbers of factors that bring about untimely death, not least in a population of premature babies. Though this case is unique in that the deaths occurred in a hospital, and from this point it is an assumption that deaths in hospitals are suspicious. I cannot challenge such an assumption, though I suspect more babies die in hospitals than they do in any other type of building. As such, what the expert witnesses are focusing on is not the details surrounding the deaths but rather the fact that a given number of deaths occurred in a defined period of time.
It should surprise no one of good sense, that such a mind set is not the foundation of a valid scientific investigation into the cause of death. Indeed, given that the incidence of death in neonates is higher within the first 7 days of birth for pre-term, very low birth weight (<1500g), male babies, born as multiples, then the fact this list describes a majority of the neonates who died at CoCH already places a great deal of uncertainty on the protocols adopted by the medical experts in their efforts to investigate the increased rate of SUEND, in the period 2015-2016.
When looking at what is essentially a single year event of increased mortality on a neonatal ward we must set out to find out whether such events have ever been described before. Thus, can we find an event in which 7 infants perish, in three distinct clusters, in the same hospital?
June-August 2015 - four infants, (3males, 1 female)
October 2015 -one infant, (1 female)
June 2016 - two infants, (2males)
Further can we identify an event where neonatal patients of a single hospital consisting of ten males : seven females are compromised and where the clinical notes consistently remark “possible sepsis?”
These are the questions that any scientific investigator would ask surrounding this case.
In what circumstances do clusters of death and sudden illness occur in the neonatal population in a Level 2 neonatal unit?
Are there other instances that depict seasonal trends in neonatal illness?
Is there any statistical significance between the pattern of stillbirths and neonatal deaths?
Are there clusters of symptoms in the neonates? Do they show temporal or spatial clustering?
Where intentional infliction of harm is assumed, what factors will conclusively exclude this determination?
What clinical events relating to the neonates are suggestive of a unique presentation unrelated to the infant's general physiological state?
What is a reliable reference point upon which to determine whether a clinical event is unique?
Where there is a showing that such questions have not been asked and answered then the standing and reliability of the expert witness reports and testimony should at a minimum be brought into doubt. Given the severity of the consequences such a failing incurs, to both the families of the babies, and the person convicted of harming them, then where such minimal conditions have not been established, any claims that purport to find a cause of death or harm should be totally disregarded.
Mortality Rate in the NNU at CoCH
A review of the mortality rate at CoCH reveals an overlap between the time of increased stillbirths and perinatal deaths. In 2015, stillbirths and perinatal deaths clustered around the same time period. Between May 2015 and November 2015, there were ten stillbirths and six perinatal deaths. In the event that there was a viral infection in the community, one might expect to see a greater number of stillbirths. This would overlap with the possibility of in-utero viral transmission. It is noteworthy that many infants became ill after they were transitioning to breast milk. If the mother had the virus (symptoms will be minor, cold like symptoms) then viral particles would be present in her breast milk. In utero infections are much more severe and have a greater mortality rate. In addition, enterovirus infections are associated with an increased prevalence of premature birth and stillbirth.
Dr Evans failed to look at any infectious diseases in his investigation. However, an assessment of relevant infectious diseases prevalent in the UK during 2015-2016 should be assessed for clinical overlap with the clinical notes Dr Evans used to determine air embolism as a cause of death. The infants at CoCH are described as having symptoms that overlap with an assortment of infectious disease colonisations. Despite the clinical notes stating that there were no bacterial cultures observed, for the infants, the case was silent on viral infections. An infant's skin, respiratory tract, gastrointestinal tract, conjunctivae, umbilicus or bloodstream may become colonised with infectious pathogens. Infection in preterm infants diverts energy away from important neurodevelopmental processes in order to protect the host and fight infection, this in turn can perturb those homeostatic functions that are regulated by the nervous system.
Given the pronounced seasonal trend in stillbirths and neonatal deaths at CoCH in 2015, it is necessary to determine whether there could be any infectious disease pathogen which exhibits a seasonal trend in transmission, and where infection in the neonatal period could be particularly deleterious. Enteroviruses (EV) and Parechoviruses (HPeV) are part of the picornaviridae family. These small viruses exhibit a seasonal pattern of infection, from summer to autumn, which overlaps with the pattern of stillbirths and neonatal deaths at CoCH.
HpEV and EV infections are common in the neonatal period, accounting for a large portion of febrile illness during the summer months (Olijve et al., 2017). Unlike older children and adults, some neonates with EV and HpEV infection progress to multi-system disease, resulting in death. Multiple clinical syndromes of varying severity are associated with neonatal enterovirus infection: asymptomatic viral shedding, nonspecific febrile illness, aseptic meningitis, hepatic necrosis and coagulopathy and myocarditis (Zhang et al., 2021).
Between 2000 and 2011, England and Wales experienced 13,901 confirmed cases of enterovirus, with an average annual incidence rate of 2.9 per 100,000 people. Initially, there was a rise in cases from 2000 to 2001, but this was followed by a decrease, reaching the lowest point in 2006. However, from that year onward, there was a consistent annual increase until 2011 (Figure 2). Throughout these twelve years, the fluctuations were consistent across all age demographics, though, infants aged <3 months, accounted for almost one-quarter of reported cases (Kadambari et al., 2014). Importantly, the study confirms the seasonal trend in viral transmission for enteroviruses.
Of additional importance is the fact that EVs and HPeVs are able to persist in wastewater due to viral shedding in faeces. There were a number of reports that wastewater was present in the NNU at CoCH. The prosecution appeared to dismiss this fact as playing a role in the collapse of the infants. However, in 2015 and 2016, the most lethal enteroviruses were extracted from wastewater in England. Both Echovirus11 (Echo11) and Coxsackievirus B1-5 (CVB 1-5) were of a high concentration in wastewater samples, suggesting that this pathogen was prevalent in the community (Majumdar et al., 2018). Both Echo11 and CVB 1-5 are associated with severe infection in the neonatal population, and most recently they have been associated with outbreaks of enterovirus infections in neonatal units in Wales, France, Italy and China (Anis et al., 2023; Grapin et al., 2023; Lu et al., 2020).
Viral transmission and Infectivity
The transmission of enteroviruses/parechoviruses can occur from mother to foetus and this is thought to be the primary origin of any infection in a neonatal unit, however infection can occur through acquired infection (Figure 3). Transmission from mother to foetus is called vertical transmission, whereas transmission from one infant to another is known as horizontal transmission. Vertical transmission of enterovirus can occur through various routes. The most common route is through the placenta, where the virus crosses the maternal bloodstream into the foetal bloodstream. Enteroviruses can also be transmitted during delivery, when the foetus comes into contact with infected birth canal secretions or faecal matter.
Once the virus is transmitted to the foetus, it can infect various organs and tissues, potentially leading to a range of clinical manifestations. The severity of the infection can vary widely, ranging from asymptomatic or mild cases to severe illnesses with significant complications. Premature infants are at particular risk from severe infection. In some cases, vertical enterovirus transmission may result in congenital infections, where the foetus is infected in utero. Congenital enterovirus infections can lead to conditions such as myocarditis (inflammation of the heart muscle), hepatitis (inflammation of the liver), encephalitis (inflammation of the brain), or even stillbirth.
In many instances, a positive viral infection is not obtained at the time of illness. This is confirmed by a retrospective analysis of microbiological and clinical data from a tertiary children’s hospital in the South of England over a 17-month period, in 2012-2013. The study identified 30 cases of enterovirus infection, in their patient population, which was likely the cause of the symptoms the patients exhibited. The symptoms associated with infection included sepsis (n = 9), myocarditis (n = 8), meningitis (n = 8) and encephalitis (n = 5). Cases with sepsis or myocarditis were significantly younger than those with central nervous system disease (median age 21 and 15 days vs.79 days) (de Graaf et al., 2016).
In one representative case, a 9-day-old infant, with enterovirus sepsis, had an out-of-hospital arrest. On arrival in the emergency department he was observed to have focal seizures with secondary generalisation, along with circulatory compromise and required endotracheal intubation and ventilation. The infant died on the paediatric intensive care unit, six days after being admitted. The post-mortem examination did not identify a definite cause of death, however, a non-typable enterovirus was detected on a throat swab (de Graaf et al., 2016).
This case is relevant as the infant was a newborn who experienced a sudden cardiac arrest followed by seizure activity and the autopsy failed to reveal a specific cause of death. The similarities between cases involving enterovirus infection in neonates and the symptoms observed in infants on the neonatal unit at CoCH are striking. These similarities cannot be dismissed as speculative when the entire approach adopted by Dr Evans goes beyond speculation and is totally unprovable. If there are preserved tissue samples from the infants, even where they have been fixed, these samples could be used to test for the presence of enterovirus and other viral infections.
Infectious Diseases in the Hospital Setting
Between May and August 2016, 26 cases of parechovirus infections were reported in neonates at Leicester NHS hospital (Tang et al., 2016). None of these cases were fatal, although the fact that they were detected in time may have improved infant outcomes. Symptoms can occur as early as day one of life, with severe symptoms generally appearing within the first two weeks of life. Frequent findings include fever or hypothermia, irritability, lethargy, anorexia, rash, jaundice, respiratory symptoms, apnoea, hepatomegaly, abdominal distension, emesis, diarrhoea and decreased perfusion. (Tang et al., 2016)
It is not clear whether the parechovirus infections observed in Leicester were due to an outbreak in the community or the hospital. In a separate study, a descriptive historical cohort analysis aimed to understand how varying definitions of healthcare-associated respiratory viral infections (HARVIs) could affect their reported incidence in a paediatric hospital setting. The study covered five years, between July 2013 and June 2018, and focused on patients aged between 0 and 21 years who were admitted to a 490-bed paediatric hospital in northern Texas, ranging from primary to quaternary care (Most et al., 2023).
HARVIs were identified through microbiological confirmation, the emergence of new symptoms during hospitalisation, and an exposure time exceeding the virus-specific minimum incubation period. Cases manifesting after the maximum incubation period were labeled as "definite," while all others were deemed "possible." These criteria were juxtaposed against alternative definitions based on different timings of onset and symptom requirements. The study also gathered data on patient demographics, diagnoses, and the severity of illnesses.
Out of the total cases reviewed, 498 HARVIs were pinpointed, with an incidence rate of 0.98 per 1,000 patient days (comprising 0.63 "definite" and 0.35 "possible" cases). The viruses most frequently detected were rhinovirus or enterovirus (58%) and respiratory syncytial virus (10%). Typically, the detection of HARVI occurred 10.5 days post-admission (with an interquartile range of 5–30 days). However, when definitions were altered, the HARVI incidence fluctuated, ranging from 0.96 to 2.00 per 1,000 patient days.
The study concluded that HARVIs continue to be a prevalent concern in paediatric hospitals, with the incidence significantly swayed by the precise definition applied. Given the widespread and potentially severe outbreaks of respiratory viruses, there's a pressing need for uniform definitions to enable consistent comparisons both within and between hospitals.
The incidence of HARVIs was more common in the Neonatal Intensive Care Unit (NICU), a statistic that was statistically significant (P < .001). These patients also generally presented with more complex clinical conditions, as evidenced by higher Chronic Condition Complexity (CCC) scores, with a median of 1 (interquartile range, or IQR, of 1–2) as opposed to a median of 1 (IQR of 0–2, P < .001). Additionally, they exhibited higher Pediatric Logistic Organ Dysfunction Version 2 (PELOD2) scores, indicating a more severe degree of illness (Most et al., 2023).
Finally, during the period running from June 2015 to June 2016, there were reports that enteroviruses were prevalent in the UK, and infants between the ages of 0-90 days were more likely to be infected with these viruses (Bubba et al., 2020). This study suggests that enterovirus infections were prevalent in the UK in 2015 - 2016, and it is probable that more hospitals across the UK encountered these viral pathogens and they went undetected due to the failure to recognise the importance of viral infections in at risk populations, such as neonates (Figure 4).
Back to the Beginning
It is now that I return to a statement I made in my second paragraph. I referred to an event that occurred in the interval between 2015 and 2023, where a sudden, unexpected increase in patient deaths occurred in a hospital setting, and where those most critically impacted were the most vulnerable of clinical populations. The event I am referring to is the coronavirus outbreak, or what we all know as COVID-19. For most individuals news of COVID-19 reached a height of frantic worry in the spring of 2020. For many of us we focused on the risk to our elders and the immune compromised in our lives, though many overlooked that the population most at risk of increased mortality were those who were already bound to intensive in-patient care.
The 2020 COVID-19 outbreak took people by surprise and to date, there are many individuals who doubt that an unseen virus could truly be responsible for a sudden increase in deaths in patients who were medically compromised. When turning to the COVID-19 outbreak what we observe is that specific clusters of individuals were disproportionately affected by the viral infection. Moreover, the symptoms of those affected were highly inconsistent. In prior years, outbreaks of infectious disease that became endemic to hospitals were nearly always bacterial or fungal. What is left unspoken is that bacterial and fungal infections make themselves very well-known, owing to the underlying biological properties of these invading pathogens. Viruses, are distinct from other pathogens in that their small size means they are able to hi-jack the cell regulatory machinery, such that they can enter the host cell. This process allows for viruses to evade immune responses, which permits high levels of viral replication and systemic infection before detection by the immune system.
Many individuals have stated that premature babies are already on the margins of survival, and this could explain the number of deaths in 2015. However, that does not explain the loss of life with regard to these infants, nor does it explain the higher incidence of stillbirth. Recent studies demonstrate that neonates are increasingly at risk of infectious disease in the first week of life. A retrospective study in a NICU in the Netherlands, over a 29-year period, found that the incidence of late onset sepsis increased from 7.1% in the years 1988-1992 to 17.4% in 1998-2002 and 13.9% in 2003-2006.(van den Hoogen et al., 2010).
As Covid-19 demonstrated, around the world, infectious diseases play an
ever increasing concern in the mortality and morbidity of at risk populations. Disturbingly, in the UK, the NHS has an almost non-existent viral testing regimen, leaving vulnerable populations open to viral infections that go overlooked, and where health measures cannot be appropriately implemented. In a prospective multi-centre study based on data from a UK neonatal infection surveillance network, it was reported that preterm neonates and those in NNUs were more likely to be diagnosed with early onset (9/1000 neonatal admissions) and late onset sepsis (29/1000 neonatal admissions), however, the study failed to evaluate the role of viral pathogens in this patient population (Vergnano et al., 2011).
The failure of UK hospitals to effectively test for viral pathogens, which we know to be deleterious to the neonatal population, likely results in untold number of deaths of neonates due to infectious disease. The experts in the trial of Lucy Letby failed to properly investigate the seasonal trends observed in stillbirths and early neonatal deaths, at CoCH, which may have its origins in an undetected viral infection. In light of the treating doctors at CoCH, and the expert witnesses, testifying that they had never observed the symptoms described in the infants, in their professional careers, it appears that they are not the best placed to be attempting to identify causes of death. Ultimately, their lack of experience in identifying possible infectious diseases is not sufficient cause to assume that the babies were deliberately harmed.
In his opening statement to the court, the prosecuting barrister, Nicholas Johnson, KC, referred to Lucy Letby as a malevolent force, responsible for the collapse and death of some 18 infants. The symptoms observed in these babies may have been caused by a malevolent force, but it was truly a force that went unseen. That force was likely a combination of an undetected viral infection, which infects only a limited population of very young infants, and impaired brain development, owing to prematurity. The result of these two phenomena was a heightened risk of sudden death. This is a much more logical explanation than the specious and confused claims presented by Dr Evans and the seven NHS doctors who acted as expert witnesses to bolster his claims.
References
Anis H, Basha Shaik A, Tiwari A, Alemayehu A, Nazir A, Atulinda L, Wojtara M, Uwishema O. Outbreak of severe myocarditis in England: Havoc by a harmless virus. Health Sci Rep. 2023 Aug 31;6(9):e1541. https://doi.org/10.1002/hsr2.1541.
Bubba L, Broberg EK, Jasir A, Simmonds P, Harvala H; Enterovirus study collaborators. Circulation of non-polio enteroviruses in 24 EU and EEA countries between 2015 and 2017: a retrospective surveillance study. Lancet Infect Dis. 2020 Mar;20(3):350-361. https://doi.org/10.1016/S1473-3099(19)30566-3.
de Graaf H, Pelosi E, Cooper A, Pappachan J, Sykes K, MacIntosh I, Gbesemete D, Clark TW, Patel SV, Faust SN, Tebruegge M. Severe Enterovirus Infections in Hospitalized Children in the South of England: Clinical Phenotypes and Causative Genotypes. Pediatr Infect Dis J. 2016 Jul;35(7):723-7. https://doi.org/10.1097/INF.0000000000001093
Grapin, M., Mirand, A., Pinquier, D., Basset, A., Bendavid, M., Bisseux, M., Jeannoël, M., Kireche, B., Kossorotoff, M., L’Honneur, A.-S., Robin, L., Ville, Y., Renolleau, S., Lemee, V., Jarreau, P.-H., Desguerre, I., Lacaille, F., Leruez-Ville, M., Guillaume, C., … Aubart, M. (2023). Severe and fatal neonatal infections linked to a new variant of echovirus 11, France, July 2022 to April 2023. Eurosurveillance, 28(22). https://doi.org/10.2807/1560-7917.es.2023.28.22.2300253
Kadambari S, Bukasa A, Okike IO, Pebody R, Brown D, Gallimore C, Xerry J, Sharland M, Ladhani SN. Enterovirus infections in England and Wales, 2000-2011: the impact of increased molecular diagnostics. Clin Microbiol Infect. 2014 Dec;20(12):1289-96. https://doi.org/10.1111/1469-0691.12753.
Lu J, Kang M, Zeng H, Zhong Y, Fang L, Zheng X, Liu L, Yi L, Lin H, Peng J, Li C, Zhang Y, Sun L, Luo S, Xiao J, Munnink BBO, Koopmans MPG, Wu J, Zhang Y, Zhang Y, Song T, Li H, Zheng H. Tracking echovirus eleven outbreaks in Guangdong, China: a metatranscriptomic, phylogenetic, and epidemiological study. Virus Evol. 2020 May 10;6(1):veaa029. https://doi.org/10.1093/ve/veaa029.
Majumdar M, Sharif S, Klapsa D, Wilton T, Alam MM, Fernandez-Garcia MD, Rehman L, Mujtaba G, McAllister G, Harvala H, Templeton K, Mee ET, Asghar H, Ndiaye K, Minor PD, Martin J. Environmental Surveillance Reveals Complex Enterovirus Circulation Patterns in Human Populations. Open Forum Infect Dis. 2018 Oct 1;5(10):ofy250. https://doi.org/10.1093/ofid/ofy250.
Most ZM, Jackson P, Sebert M, Perl TM. Contrasting definitions and incidence of healthcare-associated respiratory viral infections in a pediatric hospital. Infect Control Hosp Epidemiol. 2023 Jan;44(1):55-61. https://doi.org/10.1017/ice.2022.33.
Olijve, L., Jennings, L., & Walls, T. (2018). Human Parechovirus: an Increasingly Recognized Cause of Sepsis-Like Illness in Young Infants. Clinical Microbiology Reviews, 31(1). https://doi.org/10.1128/cmr.00047-17
Tang JW, Holmes CW, Elsanousi FA, Patel A, Adam F, Speight R, Shenoy S, Bronnert D, Stiefel G, Sundaram P, Pande S, Sridhar A, Kairamkonda V, Bandi S. Cluster of human parechovirus infections as the predominant cause of sepsis in neonates and infants, Leicester, United Kingdom, 8 May to 2 August 2016. Euro Surveill. 2016 Aug 25;21(34):30326. https://doi.org/10.2807/1560-7917.ES.2016.21.34.30326
van den Hoogen A, Gerarads L, Verboon-Maciolek M, Fleer A, Krediet T.G. Long-term trends in epidemiology of neonatal sepsis and antibiotic susceptibility of causative agents. Neonatology. 2010; 97: 22-28
Vergnano S, Menson E, Kennea N, Embleton N, Russell AB, Watts T, Robinson MJ, Collinson A, Heath PT. Neonatal infections in England: the NeonIN surveillance network. Arch Dis Child Fetal Neonatal Ed. 2011 Jan;96(1):F9-F14. https://doi.org/10.1136/adc.2009.178798.
Zhang, M., Wang, H., Tang, J., He, Y., Xiong, T., Li, W., Qu, Y., & Mu, D. (2021a). Clinical characteristics of severe neonatal enterovirus infection: a systematic review. Bmc Pediatrics, 21(1).