Therapeutic Hypothermia for Management of Neonatal Asphyxia: What Nurses Need to Know

Asphyxial injury leads to progressive hypoxemia, hypercapnia, and metabolic acidosis associated with anaerobic metabolism.^1 (p11) In this situation, blood is redirected to provide increased blood flow to the brain, heart, and adrenal glands and reduced blood flow to the intestines, skin, and kidneys. Concurrent release of epinephrine helps maintain and increase blood pressure. If the asphyxia persists, this initial coping mechanism begins to fail. Impaired myocardial function causes a decrease in blood pressure and a subsequent decrease in cerebral blood flow.^3 (p400)

The brain has limited sources of stored energy and relies on adequate blood flow to extract the needed energy supplies for neuronal cells. With reduced blood flow, neuronal cells cannot extract enough glucose to convert to energy-storing adenosine triphosphate (ATP). This decrease in ATP stimulates a cascade of biochemical reactions that lead to early neuronal cell death via ischemia and necrosis (primary energy failure) or cell death via apoptosis (secondary energy failure).⁴  These reactions involve the destruction of cell membrane potentials as ATP levels decrease. Consequently, the control of movement of ions across the cell membrane is impaired. Accumulations of intracellular calcium, sodium, chloride, and water reach toxic levels, and the level of the excitatory neurotransmitter glutamate is elevated at the synaptic junction. Cell edema and lysis follow the influx of sodium and water.⁵  Elevated levels of glutamate and intracellular calcium cause further cell damage by breaking down cell proteins and hydrolyzing lipids in cell membranes, producing free radicals such as superoxide and nitric oxide.^5(p840),6  Superoxide in asphyxiated infants reacts with nitric oxide to form toxic reactive nitrogen species that cause mitochondrial damage and eventual cell death. In addition, cell damage signals inflammatory mediators that stimulate production of glutamate. Cell membranes damaged by the depletion of ATP stores cannot clear the accumulated glutamate, which reaches toxic levels, creating the environment for apoptosis.^1(p30),5 

Clinically, these reactions manifest themselves in altered brain function known as hypoxic-ischemic encephalopathy (HIE). Once an asphyxiated infant is resuscitated, although renewed blood flow restores the energy needs of the cells, it also triggers a secondary cascade of events that further contribute to apoptotic or delayed cell death.⁸  This secondary reaction can occur from 6 to 48 hours after the initial injury. Hyperemia in reperfused areas causes inundation by neutrophils and inflammatory markers, which block microvessels, causing further ischemia. In addition, oxygen, which becomes available with reperfusion, metabolizes to produce reactive oxygen species, which further exacerbate the biochemical pathways outlined earlier, leading to cell death.^{4(p515),8(p295)} Of note, encephalopathy or manifestation of brain dysfunction can be due to causes other than birth hypoxia, such as brain malformations, drugs, infections, and metabolic disease.^9 (p225) Thus, to correlate encephalopathy with asphyxia associated with birth, the American College of Obstetrics and Gynecology¹⁰  has developed a policy statement that defines the criteria for the diagnosis of birth asphyxia (Table 1).

Globally, according to the World Health Organization,¹¹  for 2006 (the most recent data available), almost 4 million children died in the neonatal period. Among these, 7 deaths per 1000 births were attributed to birth asphyxia; the majority of the deaths occurred in underdeveloped and developing countries. In contrast, for the whole of the United States for 2005 (most recent data available),¹²  the overall neonatal mortality rate was 4.54 deaths per 1000 live births. Death associated with birth asphyxia was 2.7% of the total infant mortality rate, or less than 1 death per 1000 live births.

In 1976, Sarnat and Sarnat¹³  proposed an HIE staging system that classifies the degree of encephalopathy (stages 1, 2, and 3) on the basis of clinical assessment. In 1983, Fenichel¹⁴  classified HIE on the basis of clinical signs and symptoms (mild, moderate, and severe). The 2 classification systems are quite similar, including definitions, and are often quoted interchangeably. To date, both systems are commonly used to define HIE (Table 2).

Possible outcomes for survivors of birth asphyxia vary widely, from a normal outcome to death, with a wide range of disabilities in between. Cerebral palsy is a major outcome of HIE, but other outcomes have also been reported (Table 3).

In a prognostic model for term infants with HIE,¹⁶  3 independent predictors were highly significant (P<.05) for death or severe neurodevelopment disability (a developmental quotient of <70) in survivors: (1) administration of chest compressions for longer than 1 minute (P=.004); (2) base deficit, according to blood gas analysis (done within the first 4 hours of delivery), of more than 16 mmol/L (P=.005); and (3) onset of respirations 30 minutes or more after delivery (P=.006). When infants with HIE had these 3 characteristics, the probability of severe adverse outcomes was 93%. In contrast, if none of the 3 predictors was present, the probability of severe adverse outcome was 46%. Recently, amplitude-integrated electroencephalography (aEEG) has become available as a bedside tool for monitoring brain activity. Unlike traditional EEG machines, aEEG machines are small and portable. They record cerebroelectrical background activity in neonates via 2 electrodes placed on either side of the head and 1 electrode placed mid forehead that acts as ground. The signals received by the electrodes are compressed and displayed as background activity and microvoltage on the monitor.¹⁷  Interpretation of the findings is based on pattern recognition of the bandwidth (called amplitude) produced by the compressed signals and the voltage measurement of the upper and lower limits of this bandwidth (Table 4).^17 (p80) Findings on aEEGs are predictive of later neurodevelopmental outcome. In a meta-analysis¹⁸  of 8 studies, the sensitivity of aEEG for correct prediction of poor developmental outcome, defined as the presence of cerebral palsy and a developmental quotient of 85 or less on the Griffiths developmental scale,¹⁹  varied from 73% to 100%. The specificity of aEEG for correct prediction of no neurodevelopmental deficit also ranged from 73% to 100%. The aEEG patterns associated with poor outcomes were burst suppression, continuous low voltage, flat tracing, and single or repetitive epileptiform activity tracings. Continuous and discontinuous aEEG tracings were associated with no neurodevelopmental deficit. In further studies,^20,21  60% of infants in whom a repeat aEEG showed recovery from burst suppression, flat tracings, and low voltage within 24 hours after an asphyxia event had mild disability or normal neurodevelopmental outcomes on follow-up.

Treatment of HIE has centered on dampening or blocking the biochemical pathways that lead to neuronal death. Allopurinol has been used to prevent formation of free radicals. However, Chaudhari and McGuire²²  were unable to determine if allopurinol has a beneficial effect after HIE and called for larger trials of allopurinol therapy.

Another treatment, which has been studied widely, initially in animals^23,24  and then in neonates,^25–30  is therapeutic hypothermia. The reduction of body temperature by 3ºC to 5ºC less than normal body temperature during reperfusion before secondary energy failure occurs, can reduce cerebral injury. How this reduction is achieved is not altogether clear. Most likely the cooling reduces the hyperemia during reperfusion and decreases neuronal metabolic rate, therefore reducing the need for ATP, interrupting the excitotoxic cascade of glutamate and the inflammatory response, thus blocking neuronal cell death by apoptosis.^{6(p295),8(p64)} In studies on therapeutic hypothermia, the decrease in body temperature was achieved by head cooling only, with a special cap with circulating cold water placed on the head of the neonate, or by systemic whole-body cooling with either a thermostatically controlled cooling blanket or hot/cold gel packs placed under or around the neonate.

In a review of 8 studies with a total patient population of 638 infants, Jacobs et al³¹  analyzed (1) the effects of therapeutic hypothermia on mortality, long-term neurodevelopmental disability, cerebral palsy, neuromotor delay, and developmental delay (as indicated by scores on the Bayley Scale of Infant Development³² ) and (2) clinically important side effects. The 8 studies were published during the period 1998 to 2005. The infants were greater than 35 weeks of gestation and weighed more than 2000 g. All the studies had well-defined criteria for birth asphyxia. In 4 studies, hypothermia was achieved by using both head cooling and whole-body cooling; in the other 4 studies, whole-body cooling alone was used. Hypothermia was started by the time the infants were 6 hours old, and target temperature varied from 32.5ºC to 36.5ºC. Duration of hypothermia was from 48 to 72 hours. The rate of rewarming varied from 0.5ºC per hour to 0.5ºC every second hour, and the procedure required 8 to 12 hours. Composite outcomes of the studies indicated that therapeutic hypothermia was beneficial for term infants with HIE. Compared with infants who received standard care, those treated with hypothermia had a significant reduction in death or major neurodevelopmental disability (P<.001; 95% confidence interval [CI], 0.65–0.89). Mortality rates were 24% in the hypothermia group and 33% in infants who had standard care (P=.01; 95% CI, 0.58–0.94). Neurodevelopmental disability in survivors was 28% in the hypothermia group and 42% in the standard care group (P=.01; 95% CI, 0.51–0.92). Cerebral palsy rates were 25% in the hypothermia group and 34% in the standard care group; this difference was not significant (P=.09; 95% CI, 0.52–1.05). The most important adverse outcomes were sinus bradycardia, hypotension requiring inotropic support, and increases in thrombocytopenia.

Despite evidence on the efficacy of hypothermia in mitigating the adverse outcomes of HIE, questions about hypothermia still remained,³³  including the method of achieving hypothermia, the duration of sustained hypothermia, the optimal hypothermia temperature, the optimal age of starting hypothermia, methods of rewarming, and how to determine best which infants will benefit most.

In a more recent meta-analysis that included studies completed after those reviewed by Jacobs et al,³¹  Shah³⁴  analyzed 13 trials from the period 1998 to 2009 with a combined population of 1440 infants. In 6 studies, selective head cooling along with whole-body cooling was used to achieve hypothermia; in the other 7 studies, hypothermia was achieved by whole-body cooling alone. The target temperature varied from 33ºC to 36.5ºC, and core temperature was measured via the rectum, esophagus, and nasopharynx. Gestational ages varied from more than 35 weeks to 37 weeks or more.

Compared with the normothermia group, the hypothermia group had a significant reduction in the combined risk of mortality or moderate to severe neurodevelopmental disability (relative risk, 0.74; 95% CI, 0.65–0.83). However, more importantly, Shah addressed some of the outstanding questions about therapeutic hypothermia. The meta-analysis indicated that selective head cooling (relative risk, 0.69; 95% CI, 0.56–0.84) and whole-body cooling (relative risk, 0.77; 95% CI, 0.66–0.90) were equally effective in reducing mortality or moderate to severe disability and neurodevelopmental disability among survivors. Whereas whole-body cooling was effective (P<.05) in reducing the mental developmental index to less than 70, the psychomotor developmental index to less than 70, and the occurrence of cerebral palsy, selective head cooling was not. Temperatures less than 34ºC reduced the combined mortality and neurodevelopmental disability rate, the mental developmental index to less than 70, the psychomotor developmental index to less than 70, and the occurrence of cerebral palsy, whereas temperatures greater than 34ºC did not. The most important side effects were cardiac arrhythmias and thrombocytopenia. Tables 5 and 6 provide a comparison of outcomes and side effects reported in these 2 meta-analyses.

After careful analysis, review of evidence, and discussion with experts in the field, whole-body therapeutic hypothermia (WBTH) was adopted as a standard of care in our neonatal intensive care unit. Table 7 gives the inclusion and exclusion criteria. The goal is to achieve a rectal temperature of 33ºC to 34ºC, and the protocol (Table 8) is to be started within 6 hours after birth. The therapeutic hypothermia is maintained for 72 hours, and then the infant is gradually warmed to normal body temperature (36.8ºC–37ºC). Because the emphasis of therapeutic hypothermia is reduction of core temperature, we opted to use cool packs to achieve WBTH. We also established an interprofessional “cool crew” to review each case, so that deficiencies can be identified and rectified. Even though WBTH is a standard of care, each infant’s parents must be informed of the infant’s diagnosis, prognosis, expectant management, and the role of WBTH. Nurses in our unit are an integral part of the resuscitation team. Along with the neonatologist, they play a valuable role in the initial identification of infants in need of WBTH and later at the bedside, by providing the physical and emotional support to the infant and the infant’s family.

Our WBTH protocol does not require specialized equipment. Although the use of simple alternatives, such as bottles filled with cool water,³⁶  has been recognized as effective in other settings, achieving and maintaining hypothermia with cool packs required the nurses’ constant attention and vigilance to ensure that the temperature remained within the prescribed range. In our experience, the temperature decreased to lower than the prescribed parameters, requiring the warmer to be turned on briefly or use of a hat and blankets to avoid severe hypothermia (<32ºC). Further, during WBTH, the infant can feel cold to the touch and look dusky because of decreased blood flow to the extremities. These changes are considered normal so long as oxygen saturation is in the required zone (94%–98%).

Tremors observed in twin A after the start of anticonvulsive therapy could have been due to either shivering or ongoing seizure activity. Expertise with the use of the cerebral function monitor could have helped the nurses and the medical teams differentiate between shivering and ongoing seizure activity. If a cerebral function monitor is not available, an EEG can be helpful for diagnosing seizures and beginning appropriate anticonvulsive therapy.

In our experience, pain scores in a sedated infant are not helpful in determining if hypothermia is painful, and this area needs further exploration. Nurses must be aware of subtle clues that can indicate pain and must advocate for appropriate interventions.

Parents can feel helpless when they see their infant shivering or shaking and will need reassurances and support. In our experience, use of a separate quiet room is welcomed by parents and nurses, because it provides an opportunity for privacy and parental bonding and allows for more environmental control of ambient temperature. However, use of a separate quiet room may be a challenge in units that have an open floor plan. Maintaining therapeutic relationships with an infant’s family, keeping the family members informed and involved in the infant’s care plan, encouraging breast pumping, and judicially referring the parents to support services are important aspects of the nursing care plan for WBTH and are much appreciated by the infants’ parents.

Therapeutic hypothermia has now become a standard of care in many other units beside ours. As mentioned previously, some questions still need to be answered.^{33(pp943, 946),37}  Do gestational age and the time hypothermia is started have an effect on outcomes? Most studies on therapeutic hypothermia have been done with term infants, and treatment has been started within 6 hours of birth. In our case study, because the infant was 1 day short of the gestational age specified in the WBTH protocol, the cooling process did not begin until 11 hours after her birth.

Also not known is the optimal rate of cooling and rewarming. As noted, we had some difficulty in keeping rectal temperature at 33ºC to 34ºC.

Who benefits most from therapeutic hypothermia and what is the best way of answering that question? Although a cerebral function monitor was recently introduced in the unit, staff members were not familiar with its use, and so it was not used in the treatment of twin A. Subsequently, we had difficulty assessing actual seizure activity vs shivering vs pain response. Because twin A was born during the night shift, we were also unable to get an EEG before we began the cooling process because in-house EEG technicians are not available during the night shift. These deviations from the standard cooling protocol raise questions about the ultimate effect of these variables on outcome. With the help of the cool crew, these deviations and deficiencies were identified, and resources were allocated for the education and availability of appropriate personnel to avoid these problems in the future. Thankfully, birth asphyxia and HIE are rare conditions. However, this rarity also prevents nurses and physicians from becoming experts in the management and application of WBTH.

Does the availability of this WBTH influence a clinician’s decision to offer redirection of care toward palliation or withdrawal of medical treatment? A meta-analysis^34 (p7) detected no difference in the risk of withdrawal of life support between infants who underwent therapeutic hypothermia and infants who were normothermic. Even so, this point could have played a role in our case study, because palliation or withdrawal of medical treatment was not discussed with the parents. Each new neonatal therapy is associated with the risk of pushing the established limits without firm answers to all outstanding questions. Because of the disastrous effects of birth asphyxia, the limitations of treatment options, and the chance of improving the odds with WBTH, our experience suggests that parents would urge clinicians to take that risk.