2 The procedure
2.1.1 Hypoxic perinatal brain injury is caused by a decrease in the amount of oxygen supplied to an infant's brain close to the time of birth. Surviving infants may develop hypoxic-ischaemic encephalopathy and other organ damage, which can lead to severe, lifelong disability or death.
2.1.2 Hypoxic perinatal brain injury is characterised by fetal distress and is associated with acidosis. Diagnosis includes clinical examination, paired umbilical arterial and venous blood gas analysis and amplitude-integrated electroencephalography.
2.1.3 Hypoxic perinatal brain injury is traditionally treated with supportive care only, since no specific pharmacological agents or interventions have been shown to prevent the neuronal damage that perinatal hypoxia causes.
2.1.4 This procedure is usually carried out on infants with a gestational age of 36 weeks or more and this is reflected in the published literature.
2.2.1 In therapeutic hypothermia the infant is cooled to between 33°C and 35°C, with the aim of preventing further neuronal loss in the days following the hypoxic injury.
2.2.2 Hypothermia is usually induced by cooling the whole body with a blanket or mattress (or sometimes by cooling the head only with a purpose-made cap). Intracorporeal temperature is continuously monitored, using a rectal or nasopharyngeal thermometer, as a proxy for brain temperature.
2.2.3 Treatment is started as soon as possible after diagnosis, usually within 6 hours of birth, and continued for approximately 72 hours. The infant is then slowly warmed to normal body temperature.
Sections 2.3 and 2.4 describe efficacy and safety outcomes from the published literature that the Committee considered as part of the evidence about this procedure. For more detailed information on the evidence, see the overview.
2.3.1 A systematic review of 10 randomised controlled trials (RCTs) (1320 infants in total) reported a lower risk of death in cooled infants (whole body or head) in the first 18 months of life than in infants treated by standard care (relative risk [RR] 0.78; 95% confidence interval [CI] 0.66 to 0.93) (follow-up not stated). In 3 of these RCTs with 18-month follow-up (767 infants in total) the combined risk of death and severe disability was significantly lower in cooled infants compared with infants treated by standard care (RR 0.81;95% CI 0.71 to 0.93) and cooling increased survival with normal neurological function compared with standard care (RR 1.53; 95% CI 1.22 to 1.93) at 18-month follow-up.
2.3.2 An RCT of 325 infants treated by whole body cooling or standard care reported survival without neurological abnormality in 44% (71/163) and 28% (45/162) of infants respectively at 18-month follow-up (RR 1.57; 95% CI 1.16 to 2.12). Among the surviving infants, there was a lower rate of cerebral palsy in those treated by cooling (28% [33/120] vs 41% [48/117]; RR 0.67; 95% CI 0.47 to 0.96).
2.3.3 An RCT of 234 infants treated by head cooling or standard care reported death or severe neurodevelopmental disability at 18-month follow-up in 55% (59/108) and 66% (73/110) of infants respectively (odds ratio 0.61; 95% CI 0.34 to 1.09). In the systematic review, the 3 RCTs with a total of 767 infants reported that the rates of severe disability and cerebral palsy in surviving infants were significantly lower in the cooled infants than infants treated by standard care at 18-month follow-up (RR 0.71; 95% CI 0.56 to 0.91 and RR 0.69; 95% CI 0.54 to 0.89, respectively).
2.3.4 The Specialist Advisers listed key efficacy outcomes as improvement in survival without neurological impairment, reduction in severe disability, improvement in Motor and Psychomotor Development Index scores and reduction in cerebral palsy.
2.4.1 An RCT of 208 infants reported a higher incidence of hypocalcaemia in cooled infants than in those treated by standard care (27% [28/102] and 19% [20/106] respectively; p values not reported).
2.4.2 In the RCT of 234 infants, 1 cooled infant (who died of other causes) had skin breakdown and local haemorrhage under the cooling cap. A case report described fat necrosis in an infant treated by whole body cooling using ice packs applied to the skin. At 9 months the infant had asymptomatic firm nodules with no calcification present. In another case report, an infant treated by whole body cooling using a water-filled mattress developed sclerema on the area of the back that was in contact with the cooling mattress at 3-week follow-up; this resolved without scarring after 3 months.
2.4.3 The Specialist Advisers considered theoretical or anecdotal adverse events to include metabolic disturbances, blood hyperviscosity syndrome, increased infections, and seizures during rewarming if it is carried out too quickly.
2.5.1 The Committee noted the uncertainties and difficulties in selecting neonates for this procedure. Specifically, they noted the lack of evidence for using the procedure in neonates with less severe hypoxic brain injury, and the difficulties in deciding not to use the procedure for neonates whose degree of brain injury or comorbidities are too severe to expect survival without severe neurological deficit.