LL Part 13: The Demise of Baby E

I don't know where to begin...

Following on from our previous post here.

I would like to begin by thanking my readers for their patience. It has taken longer to work through the available evidence regarding Baby E than it has taken for any of the previous neonates.

I recommend accessing the online version of this article, as there will be additional discussion with links to and screenshots from academic papers to explain and reference several of the key points I raise. Accordingly, this article is longer than the recommended substack email length and some content and conclusions may be dropped off of the email version.

July 29th, 2015

Babies E and F were identical twins born by caesarean section on July 29th, 2015.

Their mother believed they were in ‘good condition for the gestation they were in’ at birth. She says she does not remember any specific conversation about reduced insulin production for the twins, but remembers it being mentioned, and that it was 'normal' for premature babies.

Baby E was 1.3kg at birth and his APGAR scores were reported as 7/10 at 1 minute, and 9/10 at 5 minutes. These suggest his general condition was okay.

A retrospective nursing note written at 12:17am on July 30^th states that at 10:50pm Baby E had an incident of apnoea that necessitated use of the neopuff device and, that as a result, he was put on NCPAP with 30% oxygen.

July 30th, 2015

Baby E’s mother was able to cuddle Baby E as he wasn’t on CPAP during the day shift.

July 31st, 2015

At 6:45am Baby E is given his first infusion of Actrapid insulin.

August 1st, 2015

Nurse Taylor was the designated nurse for Baby E on Night Shift from 8:00pm on August 1^st to 8:00am on August 2nd.

Baby E had his first incident of bradycardia where his heart rate dropped by around half to 79bpm and his oxygen saturation was recorded as 84%. This incident lasted for around 45 seconds. No apnoea was identified in the clinical notes for this incident.

At around 2:00pm Baby E’s insulin infusion was lowered.

August 2nd, 2015

At around 2:10am Baby E’s insulin dosage was increased.

Nurse Taylor recorded that at 6:20am Baby E had a second low oxygen saturation (desat) event down to ’83-88%’ which lasted 2 minutes. Facial oxygen was required and Baby E appeared to have hiccups during this period.

Nurse Taylor’s notes during the night shift record that Baby E was breathing on his own, receiving 24% oxygen, that his ‘respirations were 60-70’ per minute and there was ‘minimal recession evident’. Recession when breathing is seen when the child’s rib cage or abdomen are seen to be ‘sucked in’ as a result of high negative pressure when inhaling, and is one of a number of otherwise not mentioned measures (including nasal flaring, tracheal tugging and so on) for whether the neonate is in respiratory distress. Nonetheless, his respiration rate may have been at the borderline for respiratory distress.

In contrast to this, Nurse Taylor recorded Baby E’s respiratory rate as “normal” and his overall condition as “stable”. While giving evidence in court she amended this, saying that Baby E’s observation charts merely showed “stable trends”.

During the day shift Baby E was recorded as having a third bradycardic and desaturation event that lasted approximately 30 seconds.

Nurse Taylor was again the designated nurse for Babies E and F during the night shift from 8:00pm on August 2^nd to 8:00am on August 3^rd.

Nurse Taylor recorded that at 11:50pm Baby E had a fourth bradycardic and desaturation event lasting 30 seconds that resolved with ‘gentle stimulation’.

August 3rd, 2015

At 1:00am as her night shift continued, Nurse Taylor recorded a fifth  bradycardic desaturation (slowed heart rate accompanied by low oxygen saturation). Baby E’s oxygen saturation dropped into the 70s and the event lasted 45 seconds. She also described Baby E’s stomach as ‘soft, not distended’, and that he had satisfactory blood gases, heart and respiratory rates and that IV fluids were currently being administered. She also noted that his bowels had not yet opened.

Nurse Taylor’s concluding notes recorded that Baby E had tolerated his feeds, and that his tummy remained soft. At the end of her night shift, Nurse Taylor handed over to an unnamed nurse for the day shift, who was the designated nurse for Babies E and F. While Baby E was considered to be stable and capable of breathing on his own, he was on and off oxygen for periods throughout the day shift.

A family communication note recorded at 10:42am by the unnamed day nurse said that Baby E’s mother had been on the unit from 9:00am, and that the mother had visited the unit twice during Nurse Taylor’s night shift to deliver EBM.

A nurses note recorded either at 10:44am (police analyst’s evidence) or 10:50am (nurse’s evidence) records that Baby E was still breathing on his own in ‘25% ambient oxygen’ with no signs of respiratory distress. That he was pink and well perfused and ‘handles well’. Caffeine had also been administered with his intravenous fluids via an umbilical longline, as prescribed. Caffeine is often administered to premature neonates both to counteract the effect of any medications they are taking that might retard breathing (e.g. opioid pain medication, muscle relaxants), and to prevent them from stopping breathing, especially when sleeping (apnoea).

Baby E was on a ‘cautious feeding regime’. A minimal 1ml of partially digested milk was aspirated via the nasogastric tube, which was considered to be a sign he was digesting the milk he had been given.

A doctor visited and at 11:45am recorded that Baby E had suspected sepsis, hyperglycaemia, that he no longer needed to be under the florescent lights for jaundice, had satisfactory blood gases and that aspirates were ‘ok’. He noted that Baby E was tolerating feeds with EBM well but that he had not actually examined the neonate because he had been ‘having cuddles with the mother’. He charted a plan to return and examine Baby E later and for increased feeds.

At 2:10pm Dr Thomas noted that she had come to examine Baby E as he had been having skin-to-skin contact with the mother earlier in the day. Her note records that he had good tone and movements and had been handling well throughout the day. It was at this point Baby E was prescribed a small dose of insulin, which was administered via his infusion line.

Baby E’s blood gases were taken at 2:38pm, finding that his pH was 7.293 and therefore, very slightly acidic.

At 3:00pm a prescription for a new insulin infusion for Baby E at 0.5units/kg/hr is made, then crossed out, then rewritten with what was described as ‘a correct dose’ of 0.2units/kg/hr. The unnamed nurse started the new insulin infusion at this time.

At 5:00pm the mother commenced having skin-to-skin contact with Baby E, which continued until around 6:00pm.

A retrospective nursing note made at 5:24pm but recording events from 2:30 - 3:00pm records that Baby E was breathing on his own, that his blood gas readings were ‘satisfactory’, and that his feeds were increased. The primary issue she focuses on was his blood glucose reading which she admitted was actually identified from the blood gas readings (thereby contradicting her clinical notes that said the blood gasses were ‘satisfactory’). Baby E’s blood glucose reading was 18.5mmol, which is significantly elevated and well outside the normal, healthy range (my maternity textbooks and this review say 1.5 - 6.0mmol in the first few days of life). Baby E’s blood glucose readings had been consistently above 12.8 in the preceding two-and-a-half days (12.8, 18.4, 13.5, 12.9), which the unnamed nurse sought to dismissively brush aside during testimony as being ‘at the higher end of normal.’ By contrast, most literature defines neonatal hyperglycaemia as anything above 8.3mmol. In truth, Baby E had been constantly hyperglycaemic during his short and unfortunate stay in CoCH’s neonatal unit.

The nurse communicated Baby E’s latest blood glucose result to one of the doctors who ordered that Baby E be re-commenced on an insulin infusion at the rate of 0.2units/kg/hour. It appears that the prescription at 3:00pm was a telephone order and that the Nurse may have initially recorded it incorrectly (at 0.5units rather than 0.2units). For a tiny neonate like Baby E this difference (250%) in dosage could have been significant and resulted in unintended adverse consequences. Finally, she records that Baby E’s blood culture test returned a negative result, that is, it was absent of bacteria.

The issue here is that to entirely dismiss the potential for or seriousness of infection based on this one negative test is both dangerous and clinically flawed. Especially when other clinicians have already suspected sepsis exists for this child. A recent study from 2020 found that most paediatric patients (86%) with severe sepsis or septic shock will return a negative blood culture test. While they report overall mortality to be higher in those that returned a positive result (43% versus 20% of those with a negative result), this may have been found in their study for a variety of reasons including a significantly higher incidence of comorbidities in the positive group. There were also fewer outward signs of infection and less incidence of identifiable organ dysfunction in the negative test group. Clinically, we might describe this as a cryptic or covert infection - because there is little outward indication of the impact and effects of what is actually a severe sepsis infection.

Until suddenly, the child ‘crashes’ and... there it is.

It is possible, nay probable, that the elevated blood glucose was itself an outward sign of bacterial infection.

It is well-known in the clinical literature that hyperglycaemia (elevated blood glucose) occurs even in early, and therefore cryptic or covert, severe sepsis and septic shock.

Further, we know that the changes that occur in the body during sepsis do not just affect glucose, but the entire glucose/insulin regulatory system.

We know that insulin has another important role separate to that of glucose mediation. Even though insulin resistance increases during systemic infection, it has an important role that sees it mediating the effect of inflammation during sepsis.

While sepsis through such mechanisms as the lipopolysaccharide (LPS), an endotoxin released by the bacteria, promotes insulin resistance, it can also elicit the immune response that stimulates the body to release stress response hormones and reactants such as cortisol, insulin and insulin-like growth factor I. In my own crude way I would describe this as: bacterial sepsis causes inflammation and insulin resistance, the increased insulin resistance causes hyperglycaemia, hyperglycaemia and inflammation and the presence of LPS can all rise to levels that stimulate the body to release cortisol, insulin and insulin-like growth factor that, because of the insulin resistance, do not get ‘used up’ by the body. Thus, if we take the totality of evidence to its logical conclusion, you could possibly see a neonate who has: (a) elevated blood glucose; (b) elevated insulin resistance; and (c) elevated occult insulin levels - all at the same time. The one thing the sources I have highlighted above agree on (and they are the tip of a very large iceberg of medical literature), is that there is a lot we do not know about these processes and a lot of questions ripe for future clinical research. If nothing else, this contradicts just how self assured each of the doctors present at CoCH at the time events occurred, and who gave evidence during the trial, are of the idea that fulminant (severe and sudden) crashes in neonates could not have occurred due to some other process like infection, and that these things can all only be explained by Lucy Letby harming the neonates. It would seem that even though there is reticence in some neonatal and paediatric clinicians to report sudden neonatal death, many other doctors appear to disagree with those involved in the Lucy Letby case (e.g.: here, here, here, here, here, here and here).

Sudden Unexpected Early Neonatal Deaths (SUEND) were once considered rare. This is primarily because until incubators, humidicribs and the specialty of neonatology existed, the death of a premature baby was almost a certainty and most did not live very long. However, because around 50% of early premature neonates tend to live long enough to be discharged, recording of these deaths shows that they occur now with such frequency that SUEND is regularly discussed in the medical literature - with something like 580 papers published that specifically use the term when reviewing cases of unexpected neonatal death, and almost 200 published in just the last ten years. Complications during pregnancy such as birth asphyxia (11%), congenital defects (33%), prematurity (42%) and infections (5%) account for the largest proportions of all such early neonatal death. An Australian study found that early skin-to-skin positioning, being a primigravid (first time) mother and the use of peripartum opiates (analgesia during the labour and birth) were also significant risk factors for a sudden unexpected neonatal death within 7 days after birth.

At 6:00pm Baby E’s mother says she changed his nappy and wiped him around the eyes and neck. At 7:00pm she left the neonatal ward and returned to the post-natal ward. At 7:30pm Dr Thomas noted that Baby E’s C-Reactive Protein (CRP - a protein that your liver makes in response to inflammation in the body) was now 1.0, a minor elevation based on his previous levels which, while not conclusive on it’s own, can be indicative of a sepsis infection. Given that suspected sepsis was noted earlier in the day (at 11:45am) based on other symptoms, the minor elevation in CRP would likely, or should, have been considered confirmatory. However, Dr Thomas’ note goes on to say that antibiotics were ‘likely to stop at 36hrs’ as Baby E was ‘improving’.

When describing the accumulation of her clinical notes and observations during this day shift under direct examination during the trial, the unnamed nurse stated that Baby E ‘was doing well on that shift, apart from the high blood sugars’ that she opined might have been nothing more significant than a ‘stress response’ Under cross examination she yet again amended her position by acknowledging that there were significant risk factors for Baby E, including his ‘prematurity’, his ‘being a twin birth’, and that these things taken together mean he was ‘vulnerable to health complications’.

The mother claims to have returned to the neonatal ward at around 8:55pm to deliver expressed breast milk (EBM). Defence counsel Ben Myers KC countered that he believed she actually returned to the unit with the EBM at 10:00pm, which the mother denied.  

On entering the neonatal unit the mother described hearing Baby E crying loudly. She said that when she looked in his incubator there was blood around Baby E’s mouth. She said that when she entered the room Lucy Letby had been at the computer workstation and not anywhere near Child E. The mother unsuccessfully attempted a ‘containment technique’ that she said she had been taught to calm Baby E. She also described the occult blood she saw as being ‘a little above the lips, but mostly below the mouth.’

The mother asked Lucy Letby why Baby E was bleeding, and the response was that the feeding tube was ‘irritating’ the back of the throat and this would have caused it. After this brief conversation the mother returned to the post-natal ward, and at 9:11pm the mother called the father. She did not return to the neonatal ward until later in the evening, after 10:00pm, and during the resuscitation attempt on Baby E, at which time she sat in the corridor. A midwife from the post-natal unit called the father and suggested he return to the hospital at 10:52pm.

Dr Harkness was called to the neonatal ward at 10:00pm because Baby E had blood in his vomit. This is the first time in the trial testimony we have been told that Baby E had been vomiting, which itself is another sign of sepsis and impending NEC. In direct testimony he described the blood as ‘miniscule flecks’ yet his contemporaneously recorded clinical notes from the night contradict this, wherein he recorded that Baby E had a ‘large, very slightly bile-stained aspirate’ followed by a ‘sudden large vomit of fresh blood and 14ml aspirate’. In court, Dr Harkness sought immediately to diminish this blood by strongly asserting that it was not clear from his contemporaneous clinical note how much of the 14ml of vomit was actually blood. Bizarrely, he followed that up with the claim that the fresh blood was all he witnessed, as that was all he was called to see - that he did not see the vomit, only the fresh blood as a product of it. These utterances were later shown to be in conflict with his police statement that mentioned a ‘dirty’ and ‘mucus-y’ aspirate but seemingly failed to mention the blood. During testimony, Dr Harkness deflected away from discussion of Baby E’s blood by moving on to say that the records showed Baby E’s blood pressure, CRP, heart rate and saturations (on oxygen) were all good.

Dr Harkness made an additional note at 11:00pm noting another 13ml of blood-stained fluid, the contents of Baby E’s stomach, had been expelled via the nasogastric tube. He also recorded another desaturation event to 70%. He administered 100% oxygen to Baby E and his saturation remained at 60-70%. His treatment plan was to replace all fluid losses, and an order for an x-ray to check Baby E’s lungs and abdomen in an attempt to identify why Baby E was clearly deteriorating. Between 11:28-11:30pm he prescribed several medications for Baby E.

Midwife Brookes on the post-natal ward recorded in her nursing notes that at 11:30pm she received a call from the neonatal unit asking for Baby E’s mother to come down because he had had a bleed, was ‘doing poorly’ and required intubation. Midwife Brookes came to the neonatal unit with the mother and stayed with her outside the nursery room in the corridor for 10 minutes. The mother was allowed in to see Baby E once medical staff had stabilised Baby E.

The crash call occurred at 11:40pm when Baby E’s oxygen saturation dropped to between 60-70%. Dr Harkness was present with both Lucy Letby and another nurse when he observed the ‘sudden deterioration’ that occurred, and he testified that the nurses were away from the incubator and involved in gathering drugs that he had requested for Baby E. This means Lucy Letby was not there to have done anything immediatly untoward to have caused the crash. Dr Harkness’ retrospective notes record that he observed bradycardia at 80-90 beats per minute, oxygen desaturation to 60%, poor perfusion, colour change over abdomen and purple patches (purpura).

On trainee GP Dr Wood’s arrival to the neonatal unit, Registrar Harkness ordered a bolus of morphine for Baby E. Dr Wood drew up this injection. Dr Wood described that resuscitation efforts had already commenced before he arrived, and that Baby E’s saturations rose to 80% after the morphine had been administered. Shift Leader Nurse Oakley acknowledged assisting during the resuscitation attempt but stated that she could not recall what, in fact, she actually did.

At 11:45pm Dr Harkness intubated Baby E, and put him on a ventilator with 100% oxygen. Dr Harkness described that he and Dr Wood were standing at the end of the incubator discussing the medication plans while observing Baby E, and watched him crash again while on the ventilator. At this point resuscitation efforts began. Baby E’s heart rate did not recover until 1:01am, and that during resuscitation copious amounts of blood had been coming out of Baby E’s mouth and nose and that this suggested his blood pressure was very low.

The unnamed female paediatrician was the consultant on call during the night. She testified that in agreement with the coroner she spoke to, she recorded necrotising enterocolitis (NEC), a sudden and severe progression of severe sepsis or septic shock, as Baby E’s cause of death. She testified in court that the fact that Baby E’s observations were ‘very stable’ up until the sudden crash did not normally fit with a diagnosis of NEC. As we have seen already, ‘stable’ may not be a correct description of the clinical picture for Baby E, nor is her characterisation how severe sepsis itself is described in the medical literature. The doctor’s statements regarding stability up until the sudden crash and the absence of signs of air on the abdominal x-ray taken almost two hours before his death may have misled the jury. First, NEC itself is characterised by sudden onset and a wide variability of signs and symptoms that can persist for one or more days, or rapidly progress from onset to death in as little as a few hours. It would have been easy for the jury to misunderstand her comment and believe that NEC does not normally appear with sudden onset. Second, while pneumoperitoneum (abdominal air) is one of the most common signs of NEC in the premature neonate, it is by no means an absolute or ever-present sign. This is especially true where, as the evidence on this point appears to lead, only a single top-down x-ray is taken. The proper radiographic examination is to take two projections (vertical and horizontal). On these points the unnamed doctor’s statements on the stand may have been misleading and open to misinterpretation by the jury.

August 4th, 2015

Dr Woods notes record the efforts to resuscitate Baby E from 12:37am. During the resuscitation effort five doses of adrenaline were administered, seemingly with little effect. At 12:50am a blood transfusion was finally ordered for Baby E to replace the more than one-quarter of his blood volume that had now been lost. Chest compressions were briefly ceased from 1:01am because Baby E’s heart rate rose. Ventilations continued but his heart rate fell again at 1:15am and CPR had to be recommenced. At 1:23am the decision was made to cease CPR and staff began to clean Baby E. A minute later at 1:24am ventilation was withdrawn and the now deceased Baby E was released to his parents.

A retrospective nursing note written at 1:30am referred simply to ‘abdominal discolouration’. Dr Harkness’ October 2018 police statement refers to discolouration on the abdomen. Neither describes  the appearing and disappearing purple blotches described in court by Dr Harkness.

A later pathology report showed Baby E also had a gastrointestinal bleed. That report also found that Baby E’s Prothrombin Time (PT) was 19.5 (normal range 12.5-15) and Activated Partial Thromboplastin Time (aPTT) was 53.6 (normal range 26-35). These prolonged values suggest at the very least that Baby E would be prone to bleeding (which was clinically observed prior to and during resuscitation), had possibly received anticoagulant therapy (e.g. heparin), or was potentially at the early stages of liver disfunction. While Dr Harkness sought to dismiss these as high but not shocking during his testimony, such prolonged PT or aPTT are a common factor in patients with sepsis, disseminated intravascular coagulopathy (DIC), and NEC.

Discussion and Conclusions

These were first-time parents and sensational headlines in the mainstream and social media have been made from the mothers anxiety-laden testimony in court. Headlines sufficient to create doubt in some people’s minds such that there are some who believe Baby E’s death, and the subsequent events of his sibling Baby F, to be the events that Lucy Letby might have been inextricably linked to.

Unfortunately, Baby E did not receive an autopsy because an unnamed female paediatrician told the parents there was ‘little point’ as it ‘wouldn’t tell them much’. When giving testimony on October 16^th, that paediatrician actually turned and apologised to the parents of Baby E for giving them this patently inappropriate advice.

Under cross-examination the mother admitted that in response to Baby E’s crying, staff appeared not to be alerted to or concerned sufficient to come into the room at any time. This, Mr Myers said, suggested that the ‘horrendous screams’ the mother testified to might not have been anywhere near as bad as she had made out. The mother denied this and held firm to her impression that the noises made by Baby E had been ‘horrendous’. She diffidently admitted that the bleeding she had observed was seemingly not as bad as it had during her direct testimony, as it was ‘not fresh’, not ‘coming out’ and not ‘going on the bedding’. Evidence regarding when the father was told about the blood was ambiguous, as there seemed to be disagreement between the testimony led by the wife and that of the father, and the position put forward by the prosecution and defence counsel.

It was evident from cross examination that Ben Myers KC sought to challenge the mother’s obviously confused timeline of events. The midwife’s retrospective notes and the medical notes for Baby E’s ‘crash’ are timed at 11:30pm and 11:40pm respectively and call into question the mother’s timings for ‘crash’-contemporaneous events such as her attendance to deliver EBM and the call to the father to return, which she testified happened at 10:00pm and 10:52pm. Such confusion about the timing of events, especially for anxious and fearful parents, is not unusual and entirely understandable. However, it can cause confusion and cloud how others like the jury and mainstream media may remember testimony and perceive the ordering of events.

There were other small time-related inconsistencies in the information presented to the jury in court. For example, the police analyst described a nurses note with particular content as having been recorded at 10:44am on August 3rd. Yet, when the nurse who made that note was giving direct testimony, the same note with the same content was described as a ‘top-to-toe’ note recorded at 10:50am. While this might not seem like a significant issue the computer system would have recorded the note at a single precise time, and in the interests of accuracy that single precise time should have been consistently presented in court. The unnamed nurse also had to correct her written and oral evidence several times, including how in her contemporaneous and retrospective clinical notes she had miscategorised the blood gas readings as satisfactory when in fact the blood glucose reading was seriously elevated, and that Baby E’s variable observations were stable.

Dr Harkness’ police statement that failed to mention the existence of blood, and his testimony that sought to diminish its appearance and importance is troubling. Especially when he appeared to potentially become argumentative - admonishing defence counsel with the response that answering questions about how the blood loss that he acknowledged to be as much as 25% of Baby E’s total blood volume may have affected Baby E was ‘part of a separate conversation with expert witnesses’. Also troubling is the juxtaposition between his begrudging acknowledgement of ‘change’ as Baby E was clearly deteriorating very rapidly, and his following that with the strong assertion that Baby E was ‘still stable’. His assertion that there was ‘still time’ even as he stood with Dr Wood and watched Baby E crash and require resuscitation, on the face of it also seems absurd. The final incongruity is that there were no contemporaneous or retrospective clinical notes by him of the purpura (purple blotches) that he now asserted seeing on Baby E. His testimony that red blotches on one baby and purple blotches on another were ‘similar enough’ seems, on the face of it, both unscientific and unspecific. Given that there were things he now couldn’t remember that were described in the clinical notes, that he could remember something like these spots that was not in them seemed incredible (meaning not credible).

Other significant issues include that on the night shift of August 3^rd for the entire hospital (maternity, delivery, paediatrics, A&E and neonatal units) there was only one paediatric senior house officer (SHO), Dr Woods - who was a GP trainee and admits being inexperienced with neonates, and one paediatric registrar, Dr Harkness.

Bringing together all of the different elements discussed during the above case examination, Babies E and F were born to an anxious first time mother who by virtue of her caesarean section would have received some form of anaesthetic and opioid analgesia at the very least just prior to and during the surgical procedure. The available evidence also suggests she had very early skin-to-skin contact with Baby E in the first hours of life. Baby E had suspected sepsis and a range of confirmatory evidence to support a proper diagnosis of that sepsis, yet his condition continued to be described as ‘well’, ‘stable’, and ‘unconcerning’. This evidence also suggests that rather than being ‘at the higher end of normal’, a stress response or diabetes, sepsis was the likely cause of his repeat hyperglycaemic episodes (and prolonged PT and aPTT) and that the extremely low-dose insulin he was given may have been insufficient to resolve both the primary hyperglycaemia and the secondary inflammation caused by the sepsis itself. Baby E was not, as the doctors led the police and jury to believe, an unexpected death that can only be explained by the hand of Lucy Letby. Rather, his demise was entirely explainable when we bring together all of the information and seek the simplest (Occam’s razor) explanation. That explanation is bacterial sepsis or septic shock caused by a systemic bacterial infection and resulting in sudden onset NEC.

Note also that Lucy Letby’s minimal involvement comes right at the end of what was already several days of clearly declining health condition, and that even Dr Harkness admits he stayed on the unit for most of the period and was actually at the end of the incubator when Baby E spontaneously ‘crashed’.

(Every effort has been made to ensure the accuracy of this analysis. Errors and Omissions Excepted)

The next post in our series on the Lucy Letby trial can be found here.