Therapeutic hypothermia (TH) reduces mortality and neurodevelopmental morbidity in asphyxiated neonates and is now the standard of care for asphyxiated neonates in high-income countries [1]. As servo-controlled equipment are expensive, many innovative low-cost methods have been developed for cooling and are being used in low- and middle-income countries (LMICs). A recent systematic review of TH in LMICs concluded that TH was not associated with a statistically significant reduction in neonatal mortality or neurodevelopmental morbidity [2]. One among the reasons stated was the probable inefficiency of the low technology cooling devices. The review emphasized the need for further studies on safety and efficacy of TH in LMICs using low-cost device before it could be offered in routine clinical practice [2].

This was a multi-centric, uncontrolled clinical trial, done in 11 centers in India from November 2014 to December 2015. We used Miracradle - Neonate Cooler (Pluss Advanced Technologies) to provide TH. The manufacturer of the device and Christian Medical College (CMC) Vellore, the development partner, conducted a workshop at CMC, Vellore in August 2014. In this workshop, neonatologists from all the 11 centers were trained on providing TH using PCM-based device, and various clinical aspects of TH were discussed. The study centers were level-3 neonatal intensive care units (NICUs) from both public and private sector who were already using cool gel packs for TH in neonates and/or expressed interest in using the PCM-based device for the purpose. None of the centers were using servo-controlled equipment for TH. Thereafter, the cooling device was supplied to all the centers along with relevant user instructions.

We developed a registry to record the temperature, complications and other relevant data of the cooled infants. The document was supplied in hard as well as soft copy to all the centers and the data was collected prospectively over the study period. All the centers got approval from their institutional ethical board to use the equipment and collect the data for publication.

Babies were excluded from the study if they had chromosomal disorder or major congenital anomaly or inability to start cooling before 6 hours of age.

Continuous monitoring of vital parameters was done. The study infants had a central venous line and an arterial line whenever feasible. Blood counts, prothrombin time (PT) and activated partial thromboplastin time (aPTT), serum creatinine, serum electrolytes, liver enzymes and C-reactive protein (CRP) were monitored during cooling.

Sedation was not routinely given during cooling. Sedation/analgesia was given according to the protocol of the unit. All the other treatment modalities including ventilation, medications (antibiotics, anticonvulsants and inotropes) and use of blood products were done as per the protocol of individual units.

The study enrolled 103 neonates [JIPMER, Puducherry (25), CMC Vellore (23), JMMC, Thrissur (15), SJMC, Bangalore (n=10), The Cradle, Gurgaon (10), KMC, Manipal (9), SRMC, Chennai (3), Fernandez Hospital Hyderabad (3), Cloudnine Hospital Gurgaon (2), Neo Clinic Jaipur (2) and Shaiva Critical Care Ahmedabad (1)]. The mean (SD) gestational age and birth weight was 38.5 (1.5) and 2925 (458), respectively. Three (2.9%) infants had mild, 81 (78.6%) had moderate and 19 (18.4%) infants had severe encephalopathy; 58 (56.3%) babies had seizures.

TABLE I	Temperature and Cooling Profile of Enrolled Neonates (N=103) 

Table II shows the adverse events in the cooled neonates. The most frequent events were related to hypotension, coagulopathy and sepsis. Seven (6.8%) babies died before discharge; 4 babies due to severe encephalopathy, 2 due to severe persistent pulmonary hypertension (PPHN) and 1 baby due to sepsis.

TABLE II 	Adverse Events During Cooling Period in Enrolled Neonates  (N=103)

The study has some limitations. There was no control group. The nursing interventions were not objectively measured. We did not have follow-up data as we looked only at the feasibility and short-term outcome.

Most of the complications reported in cooled neonates are likely to be due to the asphyxia per se rather than cooling. In the recent Cochrane review [1], only thrombocytopenia and sinus bradycardia were significantly higher in babies being cooled, compared to non-cooled neonates. The complications in our study were much less as compared to the TOBY trial [7]. This could be due to fewer babies with severe encephalopathy in our study (only 10% as compared to the 60% in the TOBY trial). Three babies had subcutaneous fat necrosis in our study, which was similar to the 2.8% reported in the Swiss Cooling Registry [11]. There was a high rate of discontinuation of cooling before 72 hours in our study due to reasons such as hemodynamic instability/ refractory shock, PPHN and bleeding. This was probably because some of the centers were relatively inexperienced in TH and had a low threshold to withdraw cooling.

Though our study shows that TH can be practiced safely, adequate supportive care, continuous vitals monitoring and monitoring of laboratory parameters are imperative during cooling. All cooled neonates should preferably have central venous and arterial lines. TH cannot be recommended in level 2 NICUs, as cooling without adequate supportive care and back up of invasive ventilation may be unsafe.

We conclude that therapeutic hypothermia of neonates with HIE using PCM-based cooling device is feasible and safe when practiced in level 3 NICUs in India. PCM-based device seems to be comparable to standard servo-controlled equipment in maintaining the target temperature.

Contributors: NT: concept, design, data collection, data analysis, manuscript writing and review; AT, VB, MV, SR, SW, LL, UB, SM, JM, AD, PYN and SN: design, data collection, data analysis, manuscript writing and manuscript review; All authors approved the final version.

Funding: The Miracradle-Neonate Cooler was supplied by Pluss Advaced technologies free-of-cost to all the participating centers. The workshop where the participants were trained (including travel and accommodation) in using the Miracradle was funded by Christian Medical College Vellore, Adit foundation and Indian Overseas Bank (funds from the latter two raised by Pluss Advaned technologies, the manufacturer of Miracradle).

Competing interests: NT was involved in the concept and design of the Miracradle. None stated for other authors.