Aim Therapeutic hypothermia (TH), shown in developed countries to improve outcome in infants with hypoxic-ischaemic encephalopathy (HIE), was introduced into standard care at Tygerberg Children's Hospital in 2008. We aimed to describe the management and characteristics of infants treated with TH at this tertiary centre as well as the logistical challenges encountered.

Methods Infants admitted for TH between 2008 and 2011 were included. They fulfilled TOBY study entry criteria and were cooled using a whole-body cooling system. A retrospective analysis of the cooling process and clinical findings was made using data collected during treatment.

Results 100 infants with mild (32%), moderate (45%) and severe (23%) HIE were treated over 3 years. Mean time to admission was 4.87 (±1.63) hours, median time from delivery to target temperature was 7.5 h (range 2.5–15.5 h). Mean temperature on admission was 35.5°C (±1.5°C). Overall, rectal temperature was within target temperature for 82.8% of the time. Complications noted were clinically suspected/proven infection (45%), abnormal coagulation tests (48%), thrombocytopenia (34%), need for inotropic support (17%), hypoglycaemia (4%) and hyperglycaemia (10%). Rate of follow-up at 1 year among survivors was 57%. Infants not attending 1-year follow-up were more likely to have HIV-infected mothers, but there were no other demographic or clinical differences when compared with those who attended follow-up.

Conclusions Cooling is feasible in a resource-limited setting, within a strict protocol. With close monitoring, the known and common complications occur as frequently as in less resource-limited settings. Surrogate markers of later outcome need to be explored where follow-up is problematic.