Monitoring in Neurosurgical intensive-care Medicine 

K. Franz, R. Lorenz;  Neurosurgical Clinic; Johann Wolfgang Goethe University, Frankfurt am Main

Traumatic brain injury (TBI) accounts for a significant proportion of patients in intensive-care units.  In the USA some 500,000 people a year sustain TBI. 150,000 of these patients have severe TBI, which is fatal in 50,000–60,000 cases. The mean age is 15-24 years. TBI is the leading cause of death in men under 35 years of age and is present in about 50% of patients with multisystem trauma. 

A distinction is made between primary and secondary brain trauma. The only effective “treatment” for primary TBI is prevention. In secondary TBI, there is a reduction of intracellular ATP at the cellular level due to a hypermetabolic response of the brain to the trauma and to secondary ischemia. This leads to increased levels of monophosphate and adenosine, which, in turn, give rise to the hypoxanthine-mediated formation of free radicals and lipoperoxydation. The latter two factors then lead to further cell damage, since they permit the influx of Ca⁺⁺ and cause functional impairment of the Na⁺/K⁺ pump. This results in the intracellular accu­mulation of fluid (cytotoxic oedema), causing increased intracranial pressure (ICP). Other consequences include a reduction of cerebral perfusion pressure (CPP) and cerebral blood flow (CBF).

Lacking appreciable energy stores, brain tissue requires a constant supply of glucose and oxy­gen. Disruption of the energy supply leads to functional impairments within seconds and irre­versible structural damage within minutes. 

The supply of oxygen and glucose to the brain can be calculated from the product of the cerebral blood flow and the arteriovenous difference of oxygen and glucose concentrations. A constant supply of oxygen and glucose is ensured by the response of cerebral blood flow to changes in blood pressure and to PaCO₂ and PaO₂. Cerebral blood flow is calculated as the quotient of the cerebral perfusion pressure (CPP) and the cerebrovascular resistance. The CPP can be estimated by calculating the difference between the mean arterial pressure and the mean intracranial pressure: CPP = MAP - ICP.  Maintenance of adequate cerebral blood flow against the background of fluctuating perfusion pressure is ensured by cerebral autoregulation. This depends on changes of both the MAP and the ICP. 

Traumatic brain injuries as well as trauma of other organ systems lead to disruptions of these regulatory mechanisms. Brain injury causes impairment of cerebrovascular reactivity and changes of partial blood-gas pressures and autoregulation processes.

 

·       Increased (sometimes also reduced) cerebrovascular resistance

·       Intermittent increase of intracranial pressure

·       Episodes of hypoxaemia

·       Phases of arterial hypotension

 

In addition, brain-injury patients not only suffer from changes of physiologic variables (secondary damage) but also lack the possibility of adequate counterregulation. As a result, secondary ischemic cerebral lesions are gaining increasing importance. Secondary changes must be identified with the help of efficient monitoring, and a concept for preventing and correcting such regulatory dysfunctions must be established. 

An ideal monitor system should, at the very least, display the frequency, extent and duration of deviations from the norm and possibly also analyze possible causes, make suggestions with regard to therapeutic recompensation and reveal the safety and efficacy of a given therapeutic strategy. Thanks to advances in the development of software algorithms for diagnoses, we can expect the advent of computer-aided systems in intensive-care units in the future. 

I. Clinical monitoring 

State of consciousness

Conscious = Oriented patient with open eyes; sometimes somewhat sleepy; normal response to questions.

Somnolent = Sleepy, difficulty keeping eyes open; response to loud questions or gentle shaking; reduced spontaneous speech; concentration and attention practically absent

Stuporous = No response to shaking; vigorous physical stimulation is required to arouse the patient; possible presence of focal neurologic deficits

Comatose = No arousal in response to speech or physical measures; possibly movement of limbs in response to pain stimuli; four stages of coma 

Respiration

Cheyne-Stokes respiration, central neurogenic hyperventilation, ataxic respiration 

Pupils

Pupillary size and shape, comparison with the contralateral pupil

Response to light: direct, indirect, consensual; examine each pupil individually; note any signs of former ocular disease

Moderately dilated, nonresponsive pupils suggest damage to the midbrain.  Constricted, reac­tive pupils are indicative of pontine damage, but narcotics may also cause constriction of pupils.

A unilaterally dilated and unreactive pupil is a sign of compression of the third nerve (oculo­motor nerve). If herniation into the tentorial notch is present, this is a sign of midbrain dam­age. However, it may also be an accompanying sign of an aneurysm of the internal carotid artery. Other, later occurring signs of damage to the third nerve are ptosis and abduction of the eyeball.

Constriction of one pupil may be a sign of Horner’s syndrome: this would be accompanied by ptosis, enophthalmos and anhidrosis.

Contralateral hemiparesis and Horner’s syndrome are associated with a dysfunction of the sympathetic fibres around the internal carotid artery, which may be a sign of dissection of the internal carotid artery.

 

 

Corneal reflex

This is one of the brainstem reflexes which it is essential to test. 

Other neurologic findings

Spontaneous blinking in an unconscious patient originates from an intact pontine reticular formation.

Blinking as a response to bright light or sudden noise is indicative of at least partial preserva­tion of the visual and auditory pathways. Hemianopia: reflex closing of the eyes only in response to threatening movements from one side, brainstem reflexes, reaction to ocular stimuli, reaction to pain stimuli etc., spontaneous and induced eye movements. 

Oculocephalic reflex

(Cave: cervical spine trauma must first be ruled out.) If the reflex is intact, one speaks of a positive response, i.e. there is a deviation of the eyes to the left when the head is turned to the right and vice versa. In alert patients the oculocephalic reflex is always negative. 

If the oculocephalic reflex is negative and a drug influence has been ruled out, this indicates damage to the parapontine reticular formation near the core of the abducent nerve. 

II. Basic monitoring 

ECG 

Mean arterial blood pressure (invasive): continuous monitoring! 

Pressure in the radial artery correlates well with pressure in the internal carotid artery and is therefore suitable for deriving information on the cerebral perfusion pressure. In addition, access to the artery affords the possibility of drawing blood for determining PaO₂, PaCO₂ and pH. 

According to Jantzen the procedure is associated with the following complications:

·       Local infection rate (2.2–6.5%)

·       Thrombosis rate (10–70%), usually without permanent damage

From Rieke:

The radial artery is an end artery with collateral supply via the ulnar artery.

Adequate collateral supply should be ensured by means of Allen’s test before a puncture is made. Alternatives: brachial artery; dorsal artery of the foot; femoral artery;  with the latter, however, the risk of infection is increased and there is also a risk of AV fistulization. 

CPP = MAP – ICP 

Increased arteriolar resistance in patients with severe traumatic brain injury. The underlying mechanism is not fully understood. Factors include spasms and microvascular compression, e.g. due to oedema.  Normally a CPP of 40–50 mm Hg is adequate. Studies in patients with traumatic brain damage have shown that at a CPP of less than 70 mm Hg, there is evidence of cerebral flow-rate changes in the transcranial Doppler sonogram. 

Central venous pressure

 

The central venous pressure is important in the context of venous pressure monitoring, artifi­cial feeding, monitoring of volume replacement and evaluation of hydration status and cardiac performance. It provides information on volume status, cardiac function and venous compli­ance. 

In addition, a central venous line provides a good access for fluid, feeding and drugs. The normal value is 3–10 cm H₂O (1 cm H₂O = 1.36 mm Hg).  Low central venous pressure is indicative of hypovolemia with a high degree of certainty, whereas increased central venous pressure may be due to various factors: blood volume, heart failure, venoconstriction, vaso­pressor drugs, increased intrathoracic and intraperitoneal pressure, pulmonary embolism, pul­monary arterial hypertension, superior vena cava syndrome, chronic obstructive airway dis­ease, pericardial tamponade, constrictive pericarditis, cor pulmonale artefacts. The safest route of access is peripheral: 

·       Basilic vein: low risk of pneumothorax and haemothorax

·       Femoral vein: less suitable (high infection risk, less reliable central venous pressure (intra-abdominal pressure) 

If peripheral access is not possible, a central venous catheter should be inserted into the inter­nal jugular vein or the subclavian vein. With catheterization of the jugular vein the risk of pneumothorax is lower (<0.1%); however, it is suspected that venous drainage from cerebral veins may be affected. The risk of puncturing the internal carotid artery is 2–10%. The subclavian vein is usually preferred. The risk of pneumothorax associated with puncture of the subclavian vein is 1–2%, the risk of puncturing the artery 1%. In addition, catheterization of the subclavian vein causes the least hindrance for the patient. A pulmonary image is essential for checking the position of the catheter and ruling out pneumothorax. 

More accurate determination of the volume status can be obtained by catheterization of the pulmonary artery using a Swan-Ganz catheter. This is necessary if the patient has severe cardiac or pulmonary disease. The pulmonary artery pressure (normal value 10–20 mm Hg) reflects right-ventricular contractility, left-ventricular shunt and pulmonary vascular resistance. The mean pulmonary capillary wedge pressure (PCWP) reflects left-ventricular end-systolic pressure (normal value 5–12 mm Hg). Cardiac output and systemic vascular resistance can also be estimated. 

Pulse oximetry 

Changes in oxygen saturation can be measured by photoelectric means using sensors for the ear and finger and can be reproduced in the form of a peripheral pulse wave. This provides immediate information on arterial oxygen saturation. The procedure presupposes good peripheral vascularization. The value should be checked occasionally against the results of arterial blood-gas analyses. 

Temperature 

Body temperature should be monitored continuously.  Temperature increases, which may be due to central dysregulation or infection, are harmful with respect to the development of increased ICP, cerebral oedema and increased cerebral metabolism. The difference between body temperature and peripheral temperature provides information on general peripheral blood flow. 

PaCO₂

 

Partial CO₂ pressure has a significant influence on cerebral blood flow. CO₂ diffuses across the blood-brain barrier and exerts a direct influence on pH changes in the extracellular space and on cerebral perfusion. A PaCO₂ rise from 20 to 80 is attended by a fourfold increase in cerebral blood flow. A hyperventilation-induced reduction of PCO₂ to levels of 20 or less leads to cerebral hypoxia (due to massive vasoconstriction) and increased cerebral lactacidosis. This is further exacerbated by the Bohr effect, i.e. a left shift of the Hb dissociation curve with reduced oxygen release in tissue. 

III. Laboratory tests 

Monitoring of laboratory parameters is part of the surveillance of patients with traumatic brain injury. As in every intensive-care unit, blood gases, blood picture, blood sugar, sodium, osmometry and CSF values (pH and lactate, neuron-specific enolase) are regularly determined.

IV. ICP monitoring 

Intracranial pressure is one of the most important parameters to be monitored in a patient with traumatic brain injury. It is dealt with in detail in another paper. 

V. Electrophysiologic monitoring 

EEG:

Spontaneous activity at the cortical level can be measured by EEG. This activity is assessed by determining the basic rhythm following classification of the waves on the basis of their fre­quency as alpha, beta, theta and delta waves. Frequency slowing arises as a result of reduced cerebral perfusion and sedation. A delta rhythm, for example, may occur during normal sleep, during a coma (irrespective of its cause) or as a result of sedatives.  Because of these different etiologies, EEG has no predictive value for the outcome of traumatic brain injury. Established indications are the recording of isoelectric EEG as a confirmation of clinical findings without further observation time and the visualization of seizure potentials. In addition, EEG is essential for monitoring barbiturate-induced burst suppression patterns.

EEG becomes more significant and the evaluation more reliable when frequent recordings are performed. In cases where alpha waves, a sleep-wake rhythm or EEG reactivity to external stimuli recur, the prognosis is favourable. Exceptions are the locked-in syndrome, in which an alpha rhythm may be present without indicating a favourable outcome. 

Evoked potentials: AEPs (auditory  evoked   potentials), SSEPs (somatosensory evoked potentials), VEPs (visual evoked potentials), MEPs (motor evoked potentials) 

Early potentials, i.e. Potentials that occur in the first 40 ms after stimulation, are suitable parameters in nonassessable cases of severe traumatic brain injury, since they are practically unaffected by drugs. Evoked potentials represent the physiologic response to the stimulation of specific neural pathways. To be informational, the curves must be reproducible after multi­ple tests. Evaluation criteria include the presence of responses to the stimulation as well as the latency and amplitude of the potentials. In intensive-care units AEPs, SSEPs and VEPs are primarily measured. 

AEPs permit functional monitoring of the brainstem. Latency delays or even the absence of waves III–V always indicate an immanent deficit of brainstem function.  Although AEPs are not established prognostic parameters, they nevertheless have a limited predictive value. As with EEG, frequent recordings enhance the reliability of the information obtained. An absence of waves III–V indicates a poor outcome.

 

However, false-optimistic predictions occur, i.e. a poor outcome despite good AEPs. This is explained by the failure of AEPs to reflect cortical lesions, which greatly affect outcome.

In a study by Kroiß and Stöhr all patients who exhibited loss of wave V died. However, pre­traumatic auditory disturbances, bilateral petrous bone fractures and long-term treatment with aminoglycosides must be ruled out. Recovery of the loss of waves I–V during the clinical course was never observed and must be interpreted as a sign of a poor prognosis in connec­tion with the development of preexisting secondary brainstem failure. 

Measurement of SSEPs is the only electrophysiologic method for evaluating brainstem (partial) and cortical function. Measured SSEPs attest to the restitutive capacity of the brain in cases of increased intracranial pressure. If SSEPs disappear in the course of monitoring a patient with traumatic brain injury, this constitutes a warning that should be promptly acted upon. Regarding the prognosis in patients with traumatic brain injury, the literature reports an accuracy of 80%. However, in the initial phase only a prediction regarding a poor outcome (i.e. death or apallic outcome) or a good outcome (from severe neurologic deficit to restitutio ad integrum) can be made. SSEPs following medianus stimulation are usually evaluated by deter­mining the central conduction time (difference between N 13 and N 20 nuchal potentials and the first negative cortical excursion). The longer the central conduction time is delayed, the more unfavorable is the prognosis. However, no conclusion can be drawn on the basis of the central conduction time alone. 

Repeated recordings enhance the reliability here as well. The more quickly and completely an initial pathologic SSEP finding approaches normalization in follow-up examinations, the bet­ter is the prognosis. 

False-optimistic prognoses occur when the damage leading to a poor outcome does not affect the sensory pathway examined. 

VEPs play only a minor role in the monitoring of patients with traumatic brain injury. In comatose patients they can only be studied with light flashes. The responses show consider­able interindividual variation. Moreover, VEPs are affected by sedatives. Consequently, their predictive value is less than that of SSEPs and AEPs. 

According to the current literature, motor evoked potentials (MEPs) do not appear to have a place in the monitoring of intensive-care patients. However, experience with their evaluation is still rather limited.  Some authors have found that, in comparison to SEPs, MEPs have practically no predictive value. 

VI. Transcranial Doppler sonography (TCD) 

This is a noninvasive method that can be used continuously and can yield very useful information. With a pulsed 2 MHz ultrasound transducer, signals are obtained through the thin squamous part of the temporal bone from which information on the systolic, diastolic and time-averaged flow rates can be derived. The difference between the systolic and diastolic flow rates divided by the mean flow rate gives a value that relates to the cerebrovascular resistance. A reduction of the cerebral perfusion pressure (due to increased intracranial pressure or arterial hypotension) leads to a fall in the flow rate, especially the diastolic flow rate. Increased flow rates may suggest hyperaemia or increased perfusion. After traumatic brain injury the values in the middle cerebral artery, which is usually continuously monitored by ultrasound, are initially often low (around 65 cm/s). If they persist at these low levels during the further clinical course, this is indicative of an unfavourable prognosis. 

The flow rates measured by TCD are an index of cerebral blood flow. Disturbances of cere­bral blood flow play an important role in the pathophysiology of traumatic brain injury. The blood flow rates also depend on other factors, especially the PCO₂ but also on the patient’s age (maximum in the first decade of life) and haematocrit. 

Flow rate in the extracranial and basal cerebral arteries decreases in parallel with the cerebral perfusion in hypocapnia and increases in hypercapnia. The diameter of the basal cerebral arteries is practically unaffected by either hypocapnia or hypercapnia. 

Flow rate changes as a function of the peripheral arteriolar resistance, which is elevated in hypocapnia due to vasoconstriction and reduced in hypercapnia due to vasodilation. 

TCD can also help in assessing the intactness of cerebral autoregulation. Cerebral autoregula­tion is understood to be the system by which the cerebral arterioles that regulate resistance are able to maintain constant cerebral blood flow despite arterial blood-pressure fluctuations. Cerebral blood flow is normally  55 ml/100 g/min. In the presence of intact autoregulation it remains constant in the mean arterial blood-pressure range of 50 to 150 mm Hg. During con­tinuous monitoring of cerebral blood flow in the middle cerebral artery, if the blood pressure falls within this fluctuation range in patients with intact cerebral autoregulation, the blood-flow rate decreases for approximately 5 s and then returns to the previous level long before the blood-pressure fluctuation has disappeared. In patients in whom autoregulation is not intact, the cerebral blood-flow rate passively tracks the blood-pressure fluctuations. 

VII. Cerebral oxygenation 

Three methods are currently available for continuous monitoring of cerebral oxygenation: 

Continuous measurement of oxygen saturation in the bulb of the jugular vein using a retrograde fibre-optic catheter. The cerebrovenous oxygen saturation is measured spectroscopically. The normal value is 70%. Values below 50% are associated with prolongation of the arteriovenous latency difference (indicator of reduced perfusion). Values below 50% indi­cate a so-called desaturation episode. The more frequently such episodes—which may be caused by intracranial hypertension, arterial hypotension, forced hyperventilation or arterial hypoxia—occur, the poorer is the patient’s prognosis. 

This method is relatively sensitive to artifacts.  It is appropriate only in ventilated and sedated patients, since any movement gives rise to artifacts. Equipment and personnel requirements are considerable. Hence, few centres use this technique. 

Direct intraparenchymal monitoring of partial oxygen pressure in tissue is likewise an invasive method. Moreover, it is a local, not global, technique, so that it has certain limitations. Partial oxygen pressure in tissue and oxygen saturation in the jugular vein exhibit parallel behaviour. Unlike jugular vein monitoring, however, it is a stable, artifact-free technique pro­viding reliable values in 90—95% of cases. 

Near-infrared spectroscopy is a noninvasive method for measuring relative, i.e. not abso­lute, values of regional tissue oxygen saturation.  A near-infrared beam is directed through scalp, skull cap and brain tissue at the height of two attached optodes. An algorithm is then used to calculate the regional oxygen saturation from the light detected after passage. At pre­sent no adequate and established clinical evaluation exists. Problems arise in particular from diurnal differences in scattering properties, movements of the patient and detachment of the electrodes due to transpiration. 

Conclusions 

Early detection and prevention of secondary brain damage in comatose patients following traumatic brain injury can only be achieved through further improvement of neuromonitoring in the form of multimodal techniques.  Monitoring of multiple parameters by means of, inter alia, cardiovascular, clinicochemical, Doppler sonographic and electrophysiologic methods offers better possibilities for investigating cerebral function and therapeutic strategies. The ideal solution, and one that is already evident in advanced developments with suitable hardware and software, is the simultaneous, if possible on-line, recording and analysis of many different parameters in order to identify interactions and utilize them for the treatment of our patients. 

____________________________

 

Bock WJ, Lumenta C, Brock M; Klinger M (Eds.): Advances in Neurosurgery, Vol. 19, 187-276, Springer Verlag 1991

Frowein RA, Brock M -, Klinger M. (Eds.): Advances in Neurosurgery, Vol. 17, 17-122, -Springer Verlag 1989

Hacke W (Ed.): Neurocritical care. Springer Verlag 1994

Moskopp D; Wassmann H: Zerebro­protektive Maßnahmen bei Energiekrisen des Gehirns. Biermann Verlag, 1994

Narayan RK, Wilberger JE, Povlishock JT(Eds): Neurotrauma.  McGraw-Hill 1995

Stöhr M, Dichgans J; Buettner UW, Hess CW, Altenmüller E: Evozierte Potentiale

Springer Verlag 1996

Von Büdingen HJ, von Reutern G: Ultraschalldiagnostik der Hirnversorgenden Arterien

Georg Thieme Verlag, 1993