14.4.4 Neurological monitoring

Perioperative monitoring of brain function is very important for children due to the high incidence of neurological complications: 6-25% [6], including 2.3% acute complications [17]. While these tend to be embolic in adults, many neurological sequelae are ischaemic in children, due largely to the high incidence of episodes of low blood flow or circulatory arrest in deep hypothermia [1].

Unfortunately, no clinical data are available establishing the superiority of one monitoring technique over the others, or proving its impact on children's neurological outcomes [11]. None of these techniques may be considered to be indispensable. According to the standard classification used in international guidelines, all neurological monitoring systems are class IIB (the procedure may be considered) or III (the procedure is not beneficial). The level of evidence is category B (observational studies) or C (registries, expert consensus) [11]. Only two measures appear to offer benefits in terms of neurological sequelae (class IIA: the procedure is reasonable).
 

For a more detailed description of neurological monitoring techniques, see Chapter 7 (Brain Function), Chapter 18 (Aortic Arch Surgery, Monitoring) and Chapter 19 (Neurological Monitoring).

Electroencephalographic techniques

It is rarely possible to use an EEG with children in the operating theatre during surgery as it adds clutter and is difficult to interpret. It may be replaced by a brain function monitor such as a CSA (Compressed Spectral Array), which displays a spectral analysis of waves (Fourier transform) and only requires four electrodes placed on the mastoid processes and in the centre of the anterior orbital margin. EEG may be useful for determining the degree of cerebral cooling prior to a circulatory arrest, whereby perfusion must be discontinued 5-10 minutes after achieving electric silence [12]. In neonates, it reveals electrical epileptic seizures in 30% of cases, although these are not linked to neurological outcomes. In contrast, a 36 to 48-hour delay in the recovery of normal electrical activity is linked to neurological sequelae and death [8].

Simplified technology such as bispectral index (BIS™) is capable of confirming electric silence prior to circulatory arrest and may offer a means of monitoring brain functions during unstable periods, since it falls if blood flow is low or in cases of severe hypotension [9], although it already drops 1-2 units per C° during hypothermia [16]. Since it analyses the EEG based on changes related to anaesthesia rather than ischaemia, the BIS is inappropriate for perioperative neurological monitoring and offers no guarantees in terms of the degree of cerebral protection [1,5].

Transcranial Doppler (TCD)

It is possible to record the velocity of the arterial blood flow to the brain by Doppler measurement in the temporal fossa or anterior fontanelle. Although emboli (HITS: high-intensity transient signals) and blood flow variations can be detected easily, the flow velocity actually reflects the total blood flow only if the vessel diameter does not change, if the viscosity stays the same (which is not the case with haemodilution), and if the sensor remains completely stable (which is unpredictable in children). TCD shows that cerebral blood flow increases when using pH-stat regulation [26], that autoregulation is maintained at normothermia, but lost at deep hypothermia [10], and that the minimum CPB flow for maintaining cerebral blood flow is 20-30 mL/kg/min [30]. Since postoperative neurological status is not linked to the rate of perioperative emboli in children [20] and the system does not work during circulatory arrest, TCD does not provide a sufficient guarantee of brain survival in critical moments [15].

Jugular oxygen saturation (SjO₂)

Although this is the gold standard technique for measuring cerebral blood oxygenation in children, it is invasive and difficult (retrograde jugular cannulation). The normal value is 55-75%. The critical value is approximately 50%. It rises in cases of hyperaemia, hypercapnia, arteriovenous fistula or hypothermia. It drops due to systemic factors (arterial desaturation, hypocapnia, acute anaemia, hypotension) or cerebral factors (intracranial hypertension, hyperthermia, convulsions, vasospasm). Values < 40% are linked to ischaemic brain damage and neurological sequelae [22]. The catheter also provides the most effective means of measuring brain temperature. However, SjO₂ is a general cerebral perfusion indicator and regional ischaemia may go unnoticed [15]. It may be useful for confirming reduced metabolic demand prior to circulatory arrest.

Near-infrared spectroscopy (NIRS)

Near infrared spectroscopy (NIRS) is used for local measurements of haemoglobin oxygen saturation (ScO₂) (see Figure 19.6). It is a simple and child-appropriate system that indicates cerebral oxygenation. Its value coinsides with the SjO₂ value. The normal value is approximately 70% in children with normal blood oxygen levels, falling to 40-60% in cyanotic subjects [13]. ScO₂ values are lower in alpha-stat than pH-stat regulation due to hypothermic vasoconstriction [19]. ScO₂ rises with hyperoxia and hypothermia (due to reduced metabolism), but also in the state of brain death [27]. Values falling to 30-35% indicate severe damage, although neurological recovery is still possible if it lasts less than 10 minutes [29]. The most appropriate haemodynamic is one that normalises ScO₂. In hypothermia, ScO₂ helps regulate the minimum subclavian or carotid continual perfusion flow to meet the brain’s requirements (10-15 mL/kg/min) [25]. During circulatory arrest, brain tolerance to ischaemia can be assessed based on variations in ScO₂. At 20°C, desaturation is approximately 1%/min, while at 36°C it is 20%/min. The drop is > 70% of the baseline value and the nadir is reached in 15 minutes [14]. Since autoregulation is suppressed for several hours by hypothermia, the rewarming period presents a high level of risk since cerebral O₂ consumption becomes dependent on O₂ supply and therefore also on arterial pressure. Perioperative monitoring of cerebral oxygen saturation based on ScO₂ alerts the anaesthetist to any risk of brain damage should the level suddenly fall.

Two issues are raised by the use of NIRS. The first is defining the threshold below which neurological deficiencies are certain. This has still not been satisfactorily identified. Based on experience of carotid surgery under regional anaesthesia, it is estimated that the minimum tolerated ScO₂ level is generally 35-40%. However, these values relate to adults with normal blood oxygen levels [18]. If a correlation exists between ScO₂ reduction and neurological sequelae, the following reference values may be useful, although they have yet to be validated [1].
 

In cyanotic children, the values are probably lower. The speed of ScO₂ variation is just as valid as the value reached – the faster the fall and the more prolonged the period < 50%, the more serious the situation [7]. In infants, postoperative cognitive impairment tends to be exacerbated if ScO₂ falls below 40% or if it drops more than a third in relation to its baseline value [21,23].

The second issue relates to the steps to be taken if ScO₂ falls dangerously low. O₂ supply to the brain must be increased [24].
 

However, there is insufficient evidence in the available literature to suggest that maintaining ScO₂ at its normal value for the child (50-70% depending on SaO₂) or correcting it if it deviates from this value has a guaranteed impact on neurological outcomes [3].

Brain temperature

Monitoring of retrograde jugular temperature has demonstrated a hyperthermic period (mean value of 39.6°C) occurring up to 6 hours after a deep hypothermic episode during CPB, probably linked to the release of inflammatory mediators [2]. This rise in brain temperature is directly related to postoperative neurological deficiencies (see Deep Hypothermia and Circulatory Arrest). Brain temperature is difficult to measure without a catheter in the jugular bulb. Normal practice is to use a tympanic probe or a nasal probe in contact with the ethmoidal sinuses (posterosuperior wall of the pharynx).

 

© BETTEX D, BOEGLI Y, CHASSOT PG, June 2008, last update February 2020
 

 
References