COMET: Cooling in Mild Encephalopathy Trial

Research question

Does whole-body hypothermia to 33.5oC, initiated within 6h of birth and continued for 72h, improve cognitive development at two years of age after mild hypoxic ischemic encephalopathy (HIE) compared with targeted normothermia at 36.50C?

Background

In the UK, around 800 babies (0.8 per 1000 livebirths) are admitted to neonatal units with mild HIE. These babies have lower cognitive scores at 2 years, and lower IQ during school age compared with healthy peers and 38% require special educational support. Whole-body hypothermia, an evidence based intensive care therapy for babies with moderate or severe HIE, is increasingly used for babies with mild HIE in the NHS without an adequate evaluation of the safety and efficacy.

Observational reports suggest that hypothermia increases several adverse outcomes in these babies including need for invasive ventilation, opioid use, disseminated intravascular coagulation, hepatic

dysfunction, cardiac dysfunction, thrombocytopenia, coagulopathy, metabolic acidosis and increases intensive care stay.

Aims

The goal of this randomised control trial is to evaluate the safety, efficacy, and cost-effectiveness of whole-body hypothermia as a therapy for babies with mild HIE.

Methods:

Multi-centre open label two-arm randomised controlled trial with an internal pilot and masked outcome assessments recruiting babies born at or after 36 weeks from 60 NHS hospitals over a 2 ½ year period.

Babies with evidence of intrapartum asphyxia AND mild encephalopathy on neurological examination AND normal amplitude integrated EEG between 1 and 6 hours will be recruited, following parental consent, and randomised to whole-body hypothermia (33.50C) or targeted normothermia (36.50C) within 6h of birth. The primary outcome is the mean Cognitive Scale Composite score from 

the Bayley-III examination at 24 (+2) months, performed by a central team of three examiners masked to the allocation. Short term outcomes and adverse events will include mortality, duration of intensive care and hospital stay, duration of ventilatory and inotropic support, bloodstream positive infection, thrombocytopenia and

coagulopathy requiring blood products, seizures, cerebral and pulmonary bleeding, opioid use, and breastfeeding at hospital discharge. Decision-analytic modelling will be used to estimate long term cost-effectiveness across the whole life span.

A sample size of 382 infants in total (191 in each group) was calculated to detect a clinically important minimum difference of 5 points (0.3 SD), at a 0.05 significance level and 90% power as the Bayley III

Composite score has a mean of 100 and SD of 15. This increases to 426, after allowing for a conservative 10% drop-out rate.