The clinical record, as the Panel sets it out
The Panel records Baby B (its Baby 2) as a 31+2 week, 1.69 kg, female infant, the first of twins, delivered by emergency caesarean section for maternal hypertension. Her mother had anti-phospholipid syndrome and gestational cholestasis — neither charted in the medical records.
At birth she was pale and floppy with a heart rate below 60 per minute. She was resuscitated and intubated with a size 2.5 endotracheal tube after three attempts; her vocal cords were red, swollen and inflamed. A large air leak led to the tube being removed and continuous positive airway pressure applied instead. Her chest x-ray was consistent with hyaline membrane disease, and she remained on CPAP with an unstable respiratory status. Two days later she developed sudden apnoea, bradycardia, and purple discolouration of the skin with white patches. She was resuscitated, recovered, and was discharged home a month later.
The prosecution theory
As the Panel summarises the conviction: it was alleged that Baby B collapsed from injection of air into the intravenous line, causing air embolism resulting in collapse and patchy purple discolorations of the skin.
What the Panel actually concluded
The Panel’s conclusions on Baby 2 are two:
- Baby 2 collapsed from thrombotic emboli originating from a central intravenous catheter, potentially aggravated by maternal anti-phospholipid syndrome.
- There was no evidence to support air embolism.
The Panel’s reasoning:
- A kinked, impacted, non-heparinised long venous line. On the day of the collapse Baby B had a long venous line that was kinked and impacted, and which was not heparinised (not treated with anti-coagulants) — a state that predisposes to thrombosis. The Panel notes that thrombosis of long venous catheters occurs in 20–30% of infants, citing a Cochrane systematic review (2008, Art. No.: CD002772).
- Maternal anti-phospholipid syndrome raised the risk further. In this condition the immune system creates antibodies that attack the body’s own tissues and can trigger blood clots in arteries and veins. During pregnancy those antibodies can cross the placenta to the neonate and lead to thromboembolism.
- Her twin’s post-mortem showed a recent thrombotic event. The Panel records that the sibling (Baby A) suffered a similar collapse and died one day after birth, and that his post-mortem showed a recent non-occluding thrombus in the liver. Taken together, the Panel says, Baby B was at high risk for thrombosis.
- Her own blood results on the day pointed to thrombosis. On the day of the collapse Baby B had a D-dimer of 6219 (normal range 220–500 ng/ml) and a low platelet count of 103 — both, the Panel states, indicative of thrombosis. Because she survived, there is no post-mortem evidence of thrombosis either way.
- The skin signs are not diagnostic of air embolism. Patchy skin discolourations are caused by dilation and contraction of small blood vessels in the skin in response to hypoxia, which occurs in many conditions. The Panel cites Zhou and Lee (December 2024), who reported that non-specific localised patchy skin discolourations have not been reported in infants with venous air embolism, including intravenous injection of air, and that only Lee’s sign and Liebermeister’s sign are diagnostic of air embolism.
- The alleged act is difficult to reconcile with the timeline. The Panel observes that if air were deliberately infused through a central venous line to cause air embolism, the line would have to be reinfused with fluid to prevent detection — yet collapse from air embolism occurs instantaneously. The Panel doubts this could be achieved quickly enough before other staff on the unit responded to the collapse.
Why this case is anchor-dependent
The conviction on Baby B was argued to the jury in the context of the Baby A count: the pattern was one death followed by one attempt. The Panel’s reading removes both halves of that pattern. It concludes Baby A died from thrombosis (see our Baby A deep-dive) and that Baby B collapsed from thrombotic emboli from a central catheter. On that reading the two events are not a death and an attempt; they are two thrombotic events in twins sharing a maternal cause and a line-related mechanism.
The pattern-dependence problem
Baby B’s conviction was argued to the jury within a wider air-embolism pattern spanning several counts. If the air-embolism mechanism is not sustained on the underlying cases — and on Baby A and Baby B the Panel finds no evidence of it — the attempted-murder conviction on Baby B rests on a foundation that independent review does not support. This is the structural logic of the CCRC review.
What the jury did not hear on Baby B
- That her long venous line was kinked, impacted and not heparinised on the day of the collapse.
- That thrombosis of long venous catheters occurs in 20–30% of infants, on the Cochrane review the Panel cites.
- That her mother had anti-phospholipid syndrome, and that this was not charted in the medical records.
- That her D-dimer was 6219 (normal 220–500 ng/ml) and her platelets 103 — both indicative of thrombosis.
- Zhou and Lee’s December 2024 finding that non-specific patchy skin discolouration has not been reported in infants with venous air embolism.
- The Panel’s conclusion that she collapsed from thrombotic emboli from a central intravenous catheter, and that there was no evidence to support air embolism.
Sources
- International Expert Panel — Summary Report (February 2025), Panel Baby 2 (Trial child B).
- Zhou Q, Lee SK. Am J Perinatol, 27 December 2024. doi: 10.1055/a-2508-2733 — cited by the Panel on the skin signs.