Neonatal hypoxic-ischemic encephalopathy (HIE) is due to acute peripartum or intrapartum events causing cerebral ischemia and oxygen deprivation. In the subsequent 1–6 h, oxidative metabolism may be restored with partial recovery of cells or secondary energy failure may continue leading to an increase in cerebral inflammation and cell death. Untreated, cytotoxic edema, excitotoxicity, and secondary energy failure develop with further cell death, seizures, clinical deterioration, and death (1–3). Survivors with moderate and severe HIE often develop long-term neurological impairments (4).

Therapeutic hypothermia (TH) at a moderate core temperature of 33.5 ± 0.5°C, initiated within 6 h after birth and maintained for 72 h, has been shown to reduce cellular injuries and cell death. Meta-analysis of studies in high-income countries (HICs) reported improved survival, following TH in late-preterm and term neonates (LPTN, gestation of ≥35 weeks) with moderate and severe HIE (5–8). The standard equipment used is a commercially available servo-controlled device (SCD) consisting of either a cooling blanket or cap with a thermostatic mechanism that rapidly cools the neonate to a target rectal temperature of 33.5 ± 0.5°C and maintains it continuously for 72 h (9). At the end of 72 h, the neonate is rewarmed gradually to normothermia over 12 h, as too rapid rewarming can result in worsening of HIE and seizures (10). However, not all NICUs in low- and middle-income countries (LMICs) are equipped with SCD as it is very costly (19750–20000 Euros per set). In recent years, some NICUs in LMICs reported improved survival of moderate and severe HIE neonates following TH using low-cost cooled gel packs (11).

In Malaysia, HIE is common (12). Many Malaysian NICUs started to administer TH on HIE neonates in 2010. To ensure the appropriate use of TH, the Malaysian Ministry of Health (MOH) published a practice guideline (13), which recommended TH to be commenced within 6 h after birth for 72 h on stabilized neonates with moderate or severe HIE and Thompson score (14) ≥7 or seizures. Other criteria which must be met are the presence of at least two of the following: Apgar score <6 at 10 min or needing prolonged resuscitation, presence of acute perinatal events, cord pH of <7 or base deficit of ≥12 mmol/L, or arterial pH < 7.0 or base deficit >12 mmol/L within 60 min of birth. The methods/devices approved for TH were a servo-controlled cooling blanket (SCD-blanket), a servo-controlled cooling cap (SCD-cap), or passive cooling with/without the use of cool gel packs (P±CGPs). SCD was the preferred method, and P±CGP was to be used when SCD was not available. Passive cooling was described as exposing a neonate (without clothing and cap) in a cot to the NICU’s air-conditioned ambient temperature (average 22–24°C). If the rectal temperature of the neonate dropped below 33.5°C, the radiant warmer would be used to maintain the rectal temperature at 33–34°C or axillary temperature at 33.5–34.5°C. If the neonate’s temperature remained >35°C despite 60 min of passive cooling, cooled gel packs (CGPs) would be used to provide additional cooling. Before use, gel packs (about 4 Euro per pack, bought from local pharmacies) were first cooled with refrigeration without being frozen. They were placed in cotton bags on the neonate under the shoulders or upper back or across the chest.

The Malaysian National Neonatal Registry (MNNR) was established in 2005 and by 2019 had 44 members (comprising all major MOH hospitals, one university hospital, and three private hospitals). Participation in the MNNR was voluntary. Member NICUs prospectively submitted data of neonates using a standardized data form to the MNNR. Since 2012, all LPTNs admitted to these centers and diagnosed to have HIE have been included in the database. Since 2016, information on types of TH administered to each HIE LPTN was also included. This study aimed to investigate the types of therapeutic hypothermia (TH) used and risk factors associated with mortality in these high-risk neonates.

In this observational retrospective cohort study of HIE neonates in the MNNR, we found that all Malaysian NICUs provided some form of TH using either SCD or P±CGP, as only 55.4% of our NICUs had SCD. TH using SCD was received by 18.6% of mild, 51.6% of moderate, and 43.6% of severe HIE neonates. Multiple logistic regression analysis identified several factors to be significantly associated with mortality, including the severity of HIE, no TH, and TH using P±CGP.

Presently, the most reliable methods to diagnose and grade HIE are magnetic resonance imaging (MRI) and electroencephalogram (EEG). However, the majority of the Malaysian NICUs did not have timely access to MRI services during the study period, and only 22/44 NICUs had amplitude-integrated EEG (aEEG). The Sarnat score and Thompson scores were (and still are) used for diagnosis and management of HIE by the managing neonatologists in our NICUs. The findings in the present study suggested that both the Sarnat score and Thompson score could have misclassified some of the neonates, as we found some mild HIE neonates had high Thompson scores of ≥7 and died due to HIE (Table 4); and some moderate and severe HIE neonates who died due to HIE had low Thompson scores.

Many of the HIE neonates were treated with P±CGP in our NICUs. Although this resourceful method was commendable, multiple logistic regression analysis in our study showed that, after controlling for various potential confounders, its use was associated with a significantly higher (>1.5 times) risk of mortality than SCD. The major problems of using P±CGP include difficulty in maintaining a constant core temperature and labor-intensiveness. A neonate on P±CGP needs a longer time to achieve the target core temperature (about 1 h). Each gel packs in-use needs to be changed every 3–4 h to maintain the target temperature. Some NICUs used axillary temperature as a surrogate for monitoring core temperature which could further compromise the effectiveness of P±CGP, as the axillary temperature does not reflect core temperature accurately.

According to current management guidelines (8, 13), TH was indicated only for neonates with moderate and severe HIE. In this cohort, there were 18.6% of mild HIE neonates treated with TH using SCD. The benefits of TH in mild HIE are still controversial and side effects have been reported (17, 18). Yet, at the same time, there were 353 (19.6%) moderate or severe HIE neonates receiving no TH. It is uncertain whether this was due to their failure to meet the criteria for TH, or the ignorance of some staff on the benefits of TH. The significant increase in TH during the fourth year (2019) suggested that perhaps more NICU staff were aware of these benefits recently. However, because of the inadequate number of SCD, this increasing trend of TH was reflected mainly by increased use of P±CGP.

During the study period, there were only five large NICUs; they had more sets of SCD per NICU than many smaller NICUs and treated more neonates with SCD and had lower mortality rates. Their better outcome could be attributed to be having overall better equipment and expertise than many smaller NICUs. These findings suggest that the regionalization of TH in Malaysia may lead to better outcomes for HIE neonates in this country. However, for TH to be effective, it must be administered within the first 6 h of life (5, 6, 8). Like many LMICs, traffic jams and/or inadequate neonatal transportation services in Malaysia are common barriers to ensure high-risk outborns reaching regional NICUs for TH within 6 h of birth. A second option to improve outcomes is to equip all Malaysian NICUs with an adequate number of sets of SCD to provide effective TH equitably and timely to any neonate who needs it.

Currently, all NICUs provided care to all grades of HIE. Our findings suggest that to improve outcomes of HIE neonates in Malaysia, equipping NICUs with adequate sets of SCD, supportive equipment, and trained personnel, and improving access to diagnostic facilities for HIE should be considered.

Our real-world findings concur with those reported from HICs that TH was associated with better survival in both moderate and severe HIE neonates (8). Our results contrasted with those of a multicenter randomized controlled study (the HELIX study) carried out in several NICUs in three LMICs, which reported significantly higher mortality associated with TH than no TH (19). Besides the difference in the design of their study, other possible reasons for the differences in the outcome of TH between the present study and the HELIX study include the relatively smaller sample size, different diagnostic criteria used, and a much higher proportion (80%) of outborn neonates recruited in the HELIX study.

Many studies from HICs reported better long-term neurodevelopmental outcomes and cost-effectiveness in survivors of moderate HIE following TH (20, 21). In Malaysia, there is a need to carry out similar studies to confirm the benefits of TH on these high-risk neonates, including comparing the outcome of HIE neonates treated with P±CGP vs. SCD, like studies reported in other LMICs (11, 22).

The strengths of this study include its large sample size, inclusion of all HIE neonates from all the major NICUs in a middle-income country, systematic and prospective collection of patients using a standard format, and comparison of mortality of HIE receiving different types of TH in a real-world situation. The major limitations of this study are not a randomized controlled study; no information included in the MNNR database on each neonate’s Apgar score at 10 min, mechanical ventilation or resuscitation at 10 min, cord pH, and clinical features of HIE for us to verify the diagnostic accuracy of HIE in this study; no information in the database on the timing of Sarnat examination, seizures prevalence during the first 6 h of life, actual age, and indications for commencement of TH, age of initiation and duration of TH, adverse effects of TH, brain MRI findings, and neurological findings at discharge.

Both SCD and P±CGP were used for TH in Malaysian NICUs. Moderate/severe HIE, TH using P±CGP, and receiving no TH were among the risk factors associated with mortality.

The original contributions presented in this study are included in the article/supplementary material, further inquiries can be directed to the corresponding author/s.

The studies involving human participants were reviewed and approved by the National Medical Research Registry of the Ministry of Health of Malaysia. Written informed consent from the participants’ legal guardian/next of kin was not required to participate in this study in accordance with the National Legislation and the Institutional Requirements.

NB: conceptualization, data analysis, and manuscript preparation. NB, SN, and SC: data extraction and review and editing of the manuscript. All authors contributed to the article and approved the submitted version.

The authors declare that the research was conducted in the absence of any commercial or financial relationships that could be construed as a potential conflict of interest.

All claims expressed in this article are solely those of the authors and do not necessarily represent those of their affiliated organizations, or those of the publisher, the editors and the reviewers. Any product that may be evaluated in this article, or claim that may be made by its manufacturer, is not guaranteed or endorsed by the publisher.