Introduction The precision of temperature control achieved in clinical practice during therapeutic hypothermia in neonates has not been described.

Methods The hourly rectal temperature recordings from 17 infants treated with servo controlled and an equal number treated with manually adjusted cooling equipment were examined. The target rectal temperature for all infants is 33.5°C for 72 h.

Results During 6 to 72 h after start of cooling, the mean (95% CI, variance) of the averaged rectal temperatures was 33.6°C (95% CI 33.4°C to 33.8°C, 0.1°C) in the manually adjusted group and 33.4°C (95% CI 33.3°C to 33.5°C, 0.04°C) in the servo controlled group (means, p=0.08; equality of variance, p=0.03). The variance was also significantly different between infant groups during 1 to 5 h after start of cooling, p=0.01, but not during rewarming.

Conclusion The rectal temperature can be maintained close to the target temperature with either manually adjusted or servo controlled equipment, but there is less temperature variability with the servo controlled system in use in the UK.