Dr Gordon Caldwell

Mother OPR didn't know either: "My hope and expectation was that Jim McCormack was going to deliver the babies as he had scanned me every week & reassured me that everything was going really well. Unfortunately, he had booked annual leave. It was very disappointing for me that he wasn't there. I don't know the doctor that did the C-section." p.5 thirlwall.public-inquiry.uk/…

"Sort of mentioning it" also recalls the extraordinary way air embolism evidence evolved in the Baby A case over a period of 25 months. This concerns the evolution of Dr Ravi Jayaram's recollection of Baby A's skin discolouration – and its subsequent resemblance to findings described in a medical paper published in 1989 – findings arguably with no direct medical relevance to venous air embolisms. But no medical expert was available at trial to carry that point on Letby's behalf. 8 June 2015: Baby A collapses and dies. Dr Jayaram's contemporaneous clinical notes record that the baby's legs appeared white and pale before cardiac arrest. There is no mention of distinctive migrating pink patches against blue-grey skin. 24 July 2015: Six weeks later, Jayaram provides a statement to the coroner investigating Baby A's unexplained death. Again, no mention of the distinctive pink-on-blue skin changes. By 30 June 2016: Jayaram has encountered Lee and Tanswell's 1989 paper, Pulmonary vascular air embolism in the newborn, which he circulates to colleagues on that date. June–September 2017: During the police investigation, Jayaram provides the distinctive description of flitting pink patches against blue-grey skin, appearing and disappearing – unlike anything he said he had previously encountered. So, more than two years after the event, the clinical observation acquired significant additional detail. Detail bearing a striking resemblance to rare skin findings described in the Lee and Tanswell paper. What happened under cross-examination? On 24 October 2022, Ben Myers KC explored the evolution of Jayaram's recollection in considerable detail. He began with Jayaram's contemporaneous clinical notes, which recorded Baby A's legs as white and pale before cardiac arrest. Jayaram initially explained that the observation had come from Dr Harkness because he had not yet been present. But Myers examined the timing more closely. Had Jayaram been present when CPR commenced? Yes. Had he subsequently seen the white, pale appearance himself? Again, yes. Of course, Harkness could have reported the pallor before Jayaram subsequently observed it himself. But it illustrates Myers's examination of the provenance of the clinical observations - what Jayaram had personally seen, what others had reported, and what he had recorded. Myers then turned to a more consequential omission. Jayaram had documented the comparatively ordinary finding of white, pale legs, along with numerous other clinical details, including events he had not personally witnessed. Yet he had omitted extraordinary pink-on-blue patches he subsequently described as unlike anything he had encountered before. Why document the comparatively ordinary pallor but omit the unprecedented discolouration? Jayaram explained that he had not appreciated its clinical significance at the time. Myers then asked why this unusual observation had also been omitted from his formal statement to the coroner six weeks later. Jayaram's explanation was that a coroner's statement had to be based on the contemporaneous clinical notes. Myers established that Jayaram's statement did, in fact, contain recollected details absent from those notes, including the withdrawal of an umbilical venous catheter. Jayaram conceded: "I accept that." So the supposed restriction to contemporaneous notes did not adequately explain the omission. Pressed further, Jayaram returned to his lack of appreciation of the skin changes' clinical significance. But Myers pointed out that an unprecedented physical observation during a fatal collapse was relevant to a coroner whether or not the attending doctor could explain its cause. Jayaram agreed, acknowledging its relevance in retrospect. And what did the other attending doctor remember? A further complication. Dr Harkness, who also attended Baby A, subsequently recalled extraordinary blue, purple, red and white skin patches. Yet neither his contemporaneous clinical notes nor his later statement to the coroner recorded that distinctive multicoloured appearance. The original records referred instead to pallor and poor perfusion. More importantly, Harkness's recollection differed from Jayaram's in a crucial respect. Where Jayaram described pink patches appearing, disappearing and flitting around the body, Harkness testified that the discolouration did not come and go. It appeared and remained, changing as Baby A's condition deteriorated. Thus, two doctors who had documented no distinctive multicoloured pattern at the time subsequently provided different accounts of its behaviour. This was no incidental detail: the alleged resemblance to Lee and Tanswell's description was being treated as diagnostically significant. Then Dr Jayaram's cross-examination took another turn. When challenged about his failure to report the distinctive flitting patches to the coroner, Jayaram invoked hospital management's discouragement of the consultants from raising concerns about a possible connection between Letby's presence and unexplained collapses. He expressed regret that he had not been more courageous at the subsequent inquest. Myers drew the crucial distinction: reporting an unusual clinical observation to a coroner is not the same as accusing a colleague of criminal conduct. And management pressure in 2016 could hardly explain why the distinctive skin changes had been omitted from Jayaram's clinical notes in June 2015 and his statement to the coroner the following month. Dr Jayaram ultimately accepted that nothing had prevented him from reporting the physical signs. Thus, under questioning, several explanations for the missing observation were examined: its perceived lack of clinical significance, the supposed constraints of a coroner's statement, and, in discussing the later inquest, institutional resistance to raising suspicions about Letby. None satisfactorily explained why such an out-of-the-ordinary, highly memorable physical observation had not been written up in the original records. Where had the "pale and white" come from if these otherwise highly memorable flitting patches had been there from the beginning? Finally, Myers confronted Jayaram with the resemblance between his later description and the Lee–Tanswell paper, which he had read before giving his detailed account to the police. Could the published description have influenced his recollection? Jayaram acknowledged the danger of confirmation bias but denied that it had affected him. His final answer was unequivocal: "No. Absolutely not. Categorically." That categorical assurance is worth weighing against the documentary chronology. It means a sincere witness can be mistaken about the accuracy or independence of a memory, particularly after exposure to subsequent information. What the cross-examination exposed was substantial uncertainty about the provenance of a clinical observation that subsequently acquired considerable diagnostic significance. Then we come to the paper itself. What did Lee and Tanswell actually describe? Lee and Tanswell's 1989 paper reviewed neonatal pulmonary vascular air embolism, predominantly associated with severe respiratory disease and complications of high-pressure mechanical ventilation. This was the era before surfactant replacement became routine, when very premature infants frequently required substantially higher ventilation pressures. The authors reported pulmonary air-leak syndromes in 94% of cases for which the relevant information was available. They also reported a mean peak inspired oxygen concentration of 90% among 27 infants with recorded measurements. This was the concentration delivered during ventilation, not a measurement of the composition of the gas within the circulation. Cutaneous signs were reported in only 11% of cases. But the precise descriptions deserve closer attention. Lee and Tanswell reported two distinguishable observations: Blanching and migrating areas of pallor, noted in several cases. Bright pink vessels against a generally cyanosed background, observed in one of their own cases. What migrated in the first observation was pallor. What appeared pink in the second was blood vessels. Jayaram's later description involved migrating pink blotches, approximately 1-2 cm across, against blue-grey skin. The resemblance is interesting. But - it is not the precise clinical description reported in the paper. Rather, it appears to combine elements of two separate observations. Clarity matters when the apparent precision of the match is offered as corroboration in a murder case. Was Jayaram recalling a clinical observation independently, or had his later understanding of the medical literature influenced the form in which he remembered it? The cross-examination established the chronology and the discrepancies. It could not definitively answer that question. What about the underlying physiology? There is a further, potentially more fundamental problem. Lee and Tanswell's reported cases arose predominantly under physiological circumstances substantially different from the deliberate intravenous air injection alleged by the prosecution. Three differences matter. 1. Clinical context Predominantly ventilator-associated pulmonary injury and barotrauma, rather than deliberately introduced intravenous air. 2. Gas composition Highly oxygen-enriched ventilation in much of the series, versus atmospheric air containing approximately 78% nitrogen. Oxygen and nitrogen have different solubilities and behaviour in the circulation, although the composition of the original emboli was not directly measured. 3. Circulatory mechanism Gas entering damaged pulmonary vessels and reaching the systemic circulation through the pulmonary veins and left heart, versus injected venous air initially entering the right heart and pulmonary circulation. The prosecution did offer an anatomically possible route across the third distinction: air passing through a patent foramen ovale into the arterial circulation. But the existence of a possible route does not establish that it was traversed or that air introduced in this way would produce the same characteristic skin findings. How could clinical observations arising predominantly from one physiological mechanism be given diagnostic significance for another without first establishing that they were transferable? Lee and Tanswell acknowledged the possibility of air entering through intravascular lines. What their paper did not provide was direct empirical validation of its distinctive cutaneous findings as diagnostic indicators of intravenous air injection. And here there is a useful subsequent cross-check. In their 2025 review of 117 neonatal vascular air-embolism cases, Zhou and Lee identified ten involving accidental intravenous air injection. None reported localised, transient skin discolouration of the kind at issue, although two exhibited generalised discolouration. Ten cases are far too few to establish that such a sign is impossible following venous injection. The published cases were also not a controlled diagnostic study. But the findings provide little empirical support for assuming that the distinctive skin appearances from predominantly ventilator-associated literature could reliably identify intravenous air embolism. There is another potentially relevant clinical detail in Baby A's case. Positive-pressure ventilation using a Neopuff was administered during the emergency, and Dr Harkness's later account placed unusual skin changes after ventilation had begun. That does not establish barotrauma, still less that resuscitation caused the original collapse. The oxygen concentration and airway pressures being delivered at the relevant moment also remain unverified in the evidence examined here. But it does raise another question: if the characteristic skin appearances were observed during resuscitation, what excluded the collapse itself or its treatment as contributing to those appearances? This is particularly relevant when the scientific paper being invoked concerned predominantly ventilator-associated pulmonary injury. Where does that leave the evidence? Myers clearly challenged the chronology of Jayaram's recollection and its resemblance to the published description. His cross-examination of Professor Sally Kinsey subsequently exposed additional limitations in the physiological explanations advanced. Kinsey acknowledged that she was not an expert in air embolism. Her account of how bubbles might produce unusual skin appearances drew upon general circulatory physiology and, in part, literature concerning decompression sickness in deep-sea divers. That research concerned yet another physiological mechanism, whose limitations Myers also explored. So the criticism is not that nobody raised these questions at trial. It is that I can find no satisfactory demonstration in the published trial evidence that the apparent resemblance between Jayaram's later description and Lee and Tanswell's findings had the diagnostic significance attributed to it. The prosecution also relied upon other clinical and circumstantial evidence. But the existence of other evidence cannot independently validate a disputed medical sign. We are left with two fundamental questions. First, how much independent corroborative weight can be placed on a distinctive clinical observation that was absent from the contemporaneous notes and coroner's statement but emerged in detail after the witness had encountered a closely resembling published description? Second, even accepting the later recollection as entirely accurate, what established that the observed appearance was diagnostically significant for deliberate intravenous air injection, given the substantial differences between that mechanism and the predominantly ventilator-associated cases described in the original paper? The prosecution needed more than a resemblance between a retrospective recollection and an authoritative scientific publication. It needed to establish that the resemblance was reliable evidence of the particular mechanism alleged. The paper was not necessarily wrong. The recollection was not necessarily false. But neither proposition established that the paper could reliably diagnose what the prosecution claimed had happened to Baby A. And that, surely, is where the scientific scrutiny should have begun. #LucyLetby Principal sources * Lee SK and Tanswell AK (1989), Pulmonary vascular air embolism in the newborn, Archives of Disease in Childhood, 64:507–510. * R v Letby, Day 12, 20 October 2022, evidence of Dr Harkness. * R v Letby, Day 14, 24 October 2022, cross-examination of Dr Ravi Jayaram, particularly pp. 105–119. * R v Letby, Day 40, 29 November 2022, evidence of Professor Sally Kinsey, particularly pp. 53–69. * Thirlwall Inquiry evidence concerning the circulation of Lee and Tanswell's paper on 30 June 2016. * Zhou Q and Lee SK (2025), Vascular Air Embolism in Neonates: A Literature Review, American Journal of Perinatology, 42:1819–1824. * R v Letby [2024] EWCA Crim 748, Court of Appeal judgment.