This chapter is dedicated to professor Douglas Miller for his personal influence on my understanding of brain injury and his genius in originating the concept of targeted therapy, the implications of which most of us are only now coming to understand.
Introduction
An intracranial pressure (ICP) monitor is considered the gold standard intracranial monitor. In particular, the most recent guidelines for the management of severe traumatic brain injury (TBI) contain level 2 recommendations (moderate degree of clinical certainty) for the use of an ICP monitor in comatose TBI patients who have an abnormal admission computed tomography (CT) scan. Whether to use an ICP monitor in comatose patients who have a normal CT scan after TBI or in other disorders is less certain (level 3; clinical certainty not established), although neurosurgeons often will consider an ICP monitor in conditions as diverse as subarachnoid hemorrhage (SAH), meningitis, and liver failure. In addition, CT scan findings may not always predict whether ICP is increased. Consequently, there is variation across centers on how frequently an ICP monitor is used. There are very strong physiologic arguments for ICP-based management for many acute neurologic conditions that require admission to a neurocritical care unit (NCCU) and a well-described association between elevated ICP and mortality. Having the ability to continuously monitor ICP avoids “empirical” ICP treatment or blind prophylactic treatment. This is important because many treatments for ICP, although effective in lowering ICP, may have other deleterious side effects.
The Monro-Kellie Doctrine
The basic concept of ICP is couched in the Monro-Kellie doctrine. This doctrine is founded on the fixed volume of the skull, such that the interior pressure is a function of the “volumes” of the individual physiologic intracranial compartments—brain, blood, and cerebrospinal fluid (CSF)—plus mass lesions, when present ( Fig. 34.1 ). Following trauma, brain volume increases due to edema, either vasogenic or cytotoxic, the latter being predominant. The vascular compartment is primarily composed of venous blood, the volume of which can increase due to outflow obstruction. Such obstruction can be extracranial (e.g., increased intrathoracic pressure or compression of the jugular outflow system) or internal, from kinking of vessels draining into the sagittal sinus (Starling resistor), pressure-related venous compression, or thrombosis. The volume of the combined arteriovenous system can increase actively, due to increased flow recruitment from elevated metabolic activity, or passively, such as from hypercapnia or blood pressure elevation in the face of abnormal pressure autoregulation. CSF production is fixed, so volume increases within this compartment generally increase as a result of outflow obstruction or abnormal resorption.
Increases in the volume of one compartment (or the addition of a compartment such as a hematoma) may be buffered by compensatory decreases in the volume of other compartments. When the increase is sufficiently slow, these compensatory changes can maintain a stable ICP up to the point of their exhaustion ( Fig. 34.2 ). At that point, the ICP rises rapidly with any further increase in compartmental volume. If the rate of initial volume increase is rapid, the compensatory mechanisms are likely to fail earlier.
One consequence of these compensatory changes is that the brain compliance changes as its volume buffering capacity is exhausted. This means that for a given volume challenge, the resultant ICP elevation will increase as the compliance decreases. Therefore even when volume buffering keeps the ICP relatively stable, knowing the compliance of the brain is a very useful index of the status of that buffering system (vide infra).
The Monro-Kellie Doctrine and Intracranial Pressure Management
The Monro-Kellie doctrine helps explain ICP and is consistent with the framework that supports current therapeutic interventions, all of which aim to lower or prevent an increase in the volume of one or more of the intracranial compartments.
Mass Lesion
The simplest compartmental correction is for mass lesions. Removal of a mass (e.g., a hematoma) is the oldest approach to control intracranial hypertension and acts directly to improve the intracranial volume balance. This is one of the few areas in TBI or other acute neurologic disorders where a pathophysiologic entity can be dealt with directly and definitively and reinforces the value of a CT scan to avoid medical treatment of a surgical lesion.
The Cerebrospinal Fluid Compartment
The natural intracranial CSF buffering system attempts to shunt a larger volume into the spinal subarachnoid space (i.e., it is subject to outflow tract patency). If interventricular passages or egress from the fourth ventricle is compromised (e.g., a posterior fossa hematoma), the system will fail. An external ventricular drain can be inserted in such instances, or when caregivers want to facilitate CSF drainage by controlling it. This can be useful if there is a sufficient CSF volume and if the ventricular system does not collapse from drainage. An open ventriculostomy, however, renders ICP measurement unreliable, so these drains should be clamped before recording ICP.
Cerebral Blood Volume
Management of the cerebral blood volume (CBV) can address the venous or arterial compartments. Optimized drainage of intracranial venous blood by (1) elevating the head of the bed, (2) avoiding compression of the venous blood in the neck (e.g., by a tight cervical orthosis or a tracheostomy tie), or (3) reducing intrathoracic pressures through careful ventilator management can help prevent elevation of the venous compartment or interference with its volume buffering capacity. CBV can be decreased through vasoconstriction. Commonly this is achieved through hyperventilation (HV) to induce hypocapnea. Because HV influences the resistance vessels that are on the inflow side, it also decreases cerebral blood flow (CBF) and so may cause ischemia. Prophylactic HV therefore has a limited if any role to play in ICP management because of this potential toxicity. If used, for example, when hyperemia contributes to increased ICP, HV should be used carefully and an additional monitor to look for ischemia should be considered if it is used aggressively. HV-induced vasoconstriction is a relatively transient effect, and is thought to last less than 24 hours, with a risk of rebound vasodilation on discontinuation. Normobaric hyperoxia or hyperbaric oxygen therapy also may induce vasoconstriction and so alter ICP with fewer deleterious effects on cerebral metabolism than HV. However, the role of hyperbaric oxygen is still to be elucidated. Finally, mannitol may induce transient vasoconstriction through altered blood viscosity, but the effects on CBF and volume and hence on ICP depend on autoregulation.
The Brain
The natural volume buffering capacity of brain tissue is limited and has a very long time constant. In general, the brain volume increases after injury, due primarily to edema. Both cytotoxic and vasogenic mechanisms occur in post-traumatic edema. However, several lines of evidence support a cellular origin that involves aquaporin-mediated transmembrane water passage. The role of vasogenic mechanisms remains unclear. This is important because current osmotic agents (mannitol, hypertonic saline) work through the extracellular compartment. The only medical treatment to decrease extracellular edema in the brain compartment is hyperosmolar therapy (osmotherapy). Alternatively, increased brain volume can be reduced through the surgical removal of tissue, either contused or pulped parenchyma or normal, “silent” tissue. The surgical removal of contusions remains controversial and has largely been abandoned.
Decompressive Craniectomy
The final Monro-Kellie–related approach is to invalidate the premise that underlies the doctrine through decompressive craniectomy (DC); this renders the intracranial space no longer a constant. A generous skull opening and a large durotomy or nonconstraining duraplasty compensates for increased volumes of one or more intracranial compartments by increasing the volume of containment and adding a compliance buffer (because the bone-free area is not rigid). DC effectively reduces increased ICP in conditions such as stroke, TBI, and SAH. However, the exact timing of DC and its effect on outcome particularly in TBI remains debated.
Implications of Monro-Kellie–Driven Therapy
The Monro-Kellie doctrine is a useful framework to organize and understand medical and surgical approaches to the injured brain. However, it is focused on the clinical management of ICP, and the invoked therapeutic modalities are phenomenologic and based on their perceived ability to lower ICP. None of these treatments, however, specifically addresses the pathophysiology of TBI and other acute brain injuries. Osmotherapy probably works to decrease the water volume of the least expanded compartment. Reduction of brain blood volume is phenomenologic because primary venous or arterial hypervolemia usually is not an underlying pathologic event in adults, and induced vasoconstriction may decrease oxygen delivery and be harmful. CSF drainage can address a specific pathophysiologic abnormality if outflow obstruction occurs but in the absence of hydrocephalus remains phenomenologic. Management such as sedation, analgesia, neuromuscular blockade, and second-tier interventions such a high-dose barbiturates and DC also are focused on ICP and not the underlying pathology. Therapeutic hypothermia, when delivered early and not as a treatment for intracranial hypertension, was an attempt to address some of the pathophysiologic abnormalities initiated at the time of trauma, but unfortunately this treatment does not improve outcome from adult or pediatric TBI. By contrast, induced hypothermia is useful in cardiac arrest ; however, in this condition it is not used as specific ICP treatment.
The importance of this skepticism is to remind caregivers that because treatment modalities are not targeted at specific molecular or biochemical processes that must be reversed, they need to avoid harming the patient in the choice of or persistence with individual treatments. Obviously, ICP and cerebral perfusion can be manipulated in many ways. When a given approach appears ineffective or begins to exert systemic toxicity, the therapeutic approach or tactics should immediately be changed. This presumes that an ICP monitor is in place; in the modern NCCU ICP treatment should be guided by such technology and not given empirically. There is no one best way to treat increased ICP and ideally, a pragmatic approach, guided by multimodality monitoring and coupled with a view toward balanced critical care of the entire patient, must form the guiding principles.
The Role of ICP Monitoring and Treatment in TBI Management
The use of an ICP monitor and ICP treatment is best studied in TBI and the techniques and expectations used to treat elevated ICP in other conditions such as subarachnoid hemorrhage, intracerebral hemorrhage, ischemic stroke, encephalitis, and metabolic disorders among others are based in large part on the experience from TBI. The focus of this chapter therefore is on ICP and TBI. ICP data can be used to predict outcome and evolution of intracranial pathology, calculate and manage cerebral perfusion pressure (CPP), direct management strategies, and limit the use of potentially deleterious therapies. An ICP monitor is considered be a conditio sine qua non of severe TBI care. Controversy about ICP monitoring, however, is reflected in the variability of its clinical use. In addition are various philosophies about how to integrate ICP data into treatment—ICP-based therapy, CPP-based therapy, Lund therapy, and optimized HV among others. Insertion of an ICP monitor is the first step from “observational,” “clinical,” or “empirical” management of severe TBI (or pathologies in which increased ICP occurs) to monitor-based, physiologic care. Despite experience dating back to the 1960s, considerable controversy surrounds ICP-monitor–based care in TBI.
Is High Intracranial Pressure Associated with Poor Outcome?
Since the seminal work of Lundberg and Guillaume and Janny in the 1950s and 1960s, a large body of observational data demonstrates that ICP elevation is associated with poor outcome and that ICP cannot be estimated accurately or reliably on clinical grounds. The association between intracranial hypertension and poor outcome appears to be independent from other confounding variables and proportional to the amount of time that the ICP is greater than 20 to 25 mm Hg. In SAH, elevated ICP is associated with poor outcome but on multivariate analysis is not an independent factor. Outcome also may be poor because treatment for high ICP can be harmful in some circumstances. For example, mannitol administration can produce renal damage or hypotension and HV may induce cerebral ischemia, whereas barbiturates and hypothermia both increase the risk of infection and hypotension. Therapies to augment CPP are associated with increased systemic complications, and even the most basic treatments, sedation and mechanical ventilatory control, may increase ICU length of stay and exposure to iatrogenic injuries such as ventilator-associated pneumonia or skin breakdown. In particular when the proper ICP threshold is unclear (vide infra) or there is mild ICP elevation without a perfusion deficit, a patient may be at more risk of a treatment complication than of harm from ICP. In these circumstances other monitors or information may help guide ICP management.
Does Treating Intracranial Hypertension Improve Outcome?
Whether treatment of intracranial hypertension is associated with improved outcome is unclear. Certainly in individual patients an effect can be seen, but there is no Class I evidence that demonstrates ICP management is associated with better outcome. Such a trial may be difficult to perform because ICP monitoring and management have become standard of practice. Early retrospective observational data not adjusted for other variables suggest that ICP control may be beneficial. For example, Marshall observed better outcome in patients whose ICP was less than 15 mm Hg than those whose ICP was greater than 40 mm Hg for more than 15 minutes. Saul and Ducker also observed that patients with severe TBI whose ICP was treated when it was greater than 15 mm Hg did better than those treated when ICP was greater than 25 mm Hg. Eisenberg et al. conducted a randomized clinical trial (RCT) that examined high-dose barbiturate therapy for refractory intracranial hypertension in severe TBI; ICP control rather than clinical outcome was the primary objective. An association between ICP control and improved outcome for all patient groups (barbiturate group, control group, and control group that crossed over to barbiturate treatment) was observed. Whether these results apply to patients with mild elevated ICP or ICP that is responsive to management is uncertain. In SAH, Heuer et al. in a retrospective analysis of 433 patients observed that among 21 patients whose raised ICP did not respond to mannitol, all had a poor outcome. By contrast, among 145 patients whose elevated ICP responded to mannitol, 66.9% had a favorable outcome. Most recently Stein et al. performed an extensive literature review of TBI trials and case series reported after 1970 and divided patients into those with and without ICP monitoring and intensive therapy. Meta-analysis showed that mortality was 12% lower and favorable outcome 6% greater in the aggressively treated group.
There also are several reports that use of an ICP monitor is not associated with improved outcome after TBI or that “ICP monitoring in accordance with the guidelines for the management of severe traumatic brain injury is associated with worsening of survival in TBI patients.” These studies have several flaws that limit meaningful conclusions but they emphasize two important issues about ICP monitors: (1) continued equipoise about the role of ICP monitors in TBI management, and (2) ICP monitoring exists only in the context of an associated treatment philosophy, which confounds the analyses of the value of ICP monitoring versus the efficacy of a specific treatment strategy. In the two studies that reported no benefit to an ICP monitor, all the patients were treated for suspected intracranial hypertension, whether monitored or not. Because insertion of an intraparenchymal ICP monitor is of very low risk, it is unlikely that display of the ICP will adversely influence outcome or improve recovery per se. Implicit in any discussion about the role of an ICP monitor in TBI management is that the interventions generated by the ICP data improve outcome. Embedded within such a monitor-driven therapy question are several issues, including (1) the validity of the treatment threshold and whether that threshold applies to all patients, (2) whether a given treatment regimen is appropriate in all monitored patients, (3) the choice of therapy, (4) the efficacy and risk-to-benefit ratio of various treatments alone or in combination, and (5) an understanding of the interaction between therapy and management interventions made in a given patient for other reasons. Consequently, although use of an ICP monitor may make sound physiologic sense, the ability to demonstrate that its use makes a difference to outcome may be confounded by several reasons: (1) wrong treatment threshold, (2) wrong individual therapeutic agent, (3) wrong therapeutic approach, (4) toxic effects of ICP treatment, (5) misinterpretation of ICP data, (6) wrong patient population, (7) lack of utility of ICP monitoring, or (8) a combination of all. An example of how such confounding may affect whether a monitor is useful are the RCTs that demonstrated that routine use of right heart catheterization in the ICU is not associated with better outcome (i.e., use of a monitor does always mean better outcome unless effective management strategies exist).
An Intracranial Pressure Threshold
Among several areas on how best to interpret and treat ICP is the threshold value used for treatment and in particular whether a single threshold applies to all patients or in one patient at all times. Too high a value may allow unrecognized neural injury, whereas too low a value may result in overtreatment and iatrogenic complications. The most frequently used threshold of 20 mm Hg was adopted from the accepted upper limit of normal for lumbar CSF pressure measurements and from early work by Lundberg. Schreiber et al. used multiple regression modeling and found that ICP greater than or equal to 15 mm Hg was associated independently with mortality after severe TBI. However, they did not analyze treatment thresholds per se. Marmarou et al. analyzed 428 monitored patients from the Traumatic Coma Data Bank and found that the proportion of total monitoring time that the ICP was greater than 20 mm Hg was an independent predictor of 6-month outcome. However, ICP treatment thresholds of 20 or 25 mm Hg were used, which confounds separation of the effects of ICP of greater than or equal to 20 mm Hg from treatment toxicity. Nevertheless, an ICP threshold of more than 20 mm Hg is now central to ICP management based largely on post hoc analysis of studies in which 20 mm Hg was the treatment trigger (i.e., there are no controlled studies on of the “best” ICP treatment threshold). This is important for several reasons. First, observational data for which other monitors are used—jugular bulb catheters, brain oxygen, or microdialysis—suggest that brain metabolism may be abnormal even when ICP and CPP are normal. Second, an ICP threshold of 20 mm Hg was established during a time period when systemic hypertension was thought to be a risk factor for intracranial hypertension, patients were routinely kept “dry,” and CPP was not a monitored variable. Consequently, many TBI patients in the intensive care unit (ICU) had relatively low blood pressures and the only real limit was to avoid systolic pressures less than 90 mm Hg. This represents a mean arterial pressure of about 70 mm Hg. Because normal autoregulation becomes impaired at approximately 50 mm Hg, an ICP value of 20 mm Hg therefore may represent borderline ischemia; that is, the threshold value may be a by-product of the treatment regimens of that period. At the time of this publication, the management focus is on not only ICP but also CPP and in some centers brain oxygen or other parameters. This combined with sedatives that lower the cerebral metabolic rate mean that the flow-to-perfusion consequences of an ICP that is 20 mm Hg or just elevated may be different today compared with the time period when this threshold was decided upon.
Perhaps more important than a single ICP threshold may be a trend over time, ICP waveform analysis, or whether the ICP value is associated with other detrimental effects. The concept that the ICP threshold may vary among patients and within patients over time (i.e., targeted) is not new. In the 1970s and 1980s, professor Douglas Miller often used an ICP threshold of 25 to 30 mm Hg in TBI patients with acceptable indicators of cerebral oxygenation and compliance. This concept of permissive intracranial hypertension is practiced at several institutions including the author’s own but is best guided by information from multimodality monitoring. For example, a brain oxygen monitor may complement the information from an ICP monitor. Some but not all observational studies suggest that when both ICP and brain oxygen are treated, the outcome may be better than if just ICP is treated after TBI. This effect may occur in part because the deleterious consequences of ICP overtreatment may be avoided. Consistent with this Chambers et al. calculated receiver-operating characteristic curves for ICP and CPP in 227 patients and found that the sensitivity of ICP for outcome increased at an ICP of 10 mm Hg, but by 30 mm Hg it had only reached 61%, with a cutoff value of 35 mm Hg that varied with the CT classification.
The Value of Intracranial Pressure Monitoring
Although understanding of ICP is incomplete and treatment approaches still need to be refined, there are many other potential benefits from ICP monitoring. First, without an ICP monitor, CPP is not known. Even transient episodes of ischemia can be devastating to the traumatized brain, making it critical to accurately and continuously monitor CPP. Because insertion of intraparenchymal ICP monitors is safe, the ability to monitor CPP per se is a supportable argument for widespread ICP monitoring. Second, cerebral herniation is a pressure issue and an ICP monitor may allow early detection. It is impossible to determine the pressure (pressure gradient) empirically (i.e., the neurologic examination) when brain herniation will occur although it is clear once it has occurred. It is preferable to avoid herniation than to treat it post hoc. Third, information from an ICP monitor may provide useful acuity information and so guide patient care and perhaps ICU resources. For example, a patient with a worrisome-appearing CT scan who does not have intracranial hypertension may not require the same degree of treatment as a patient with a similar scan but elevated ICP. Similarly, a patient with elevated ICP that is refractory to escalating management becomes an early candidate for “second tier” treatments or if very high, even withdrawal of care. Fourth, ICP trends can be an early warning of mass lesion expansion, the appearance of new lesions, or evolution of edema, ischemia, or hydrocephalus and allow these conditions to be effectively managed before clinical findings change or they are detected on periodic imaging studies. Finally, because ICP values have prognostic value it can guide management and prognosis discussions with the family.
Intracranial Pressure Monitoring Technology
Technology evaluation is not an ideal topic for evidence-based medicine. However, the various ICP monitor systems are well described in the ICP monitoring technology section of the guidelines for the management of severe traumatic brain injury. This report concluded that at the time of its writing, three technologies are sufficiently accurate to be interchangeable in use: (1) a ventricular catheter connected to an external strain gauge; (2) catheter tip strain gauge devices, or (3) catheter tip fiber-optic technology. Each of these systems can be placed into the ventricle. Catheter tip strain gauges or fiber-optics also can be placed into the parenchyma. The most commonly used devices include the Camino or the Ventrix (Integra Neurosciences, Plainsboro, NJ), Codman microsensor (Codman, Raynham, MA), the Spiegelberg ICP sensor and compliance device (Spiegelberg KG, Hamburg, Germany), and the Raumedic ICP sensor and multiparameter probe (Raumedic AG, Munchberg, Germany). The pneumatic Spiegelberg ICP monitor unlike the other parenchymal monitors also allows in vivo calibration and intracranial compliance monitoring. The evidence report also concluded that fluid-coupled or pneumatic devices placed in the subarachnoid, subdural, and epidural compartments are less accurate. Noninvasive technologies for ICP monitoring based on a variety of techniques such as ultrasound, transcranial Doppler, acoustic properties of cranial bones, tympanic membrane displacement, heart rate variability and cardiac coupling, ophthalmodynamometry, and measurement of optic nerve diameter are under investigation. Although more than 30 noninvasive ICP monitors have been patented since the 1990s, most are still too cumbersome or not sufficiently accurate to be used in clinical practice.
The ventricular catheter connected to a fluid-coupled external strain gauge is considered the gold standard ICP monitor in part because it can be used also to treat ICP through CSF drainage. Control of elevated ICP is observed in about 50% of patients in whom an external ventricular drain (EVD) is inserted after other initial measures fail. Traditional ventricular catheter external transducers only permit intermittent ICP monitoring when the drain is closed (i.e., not draining), although some catheters have internal transducers that permit CSF drainage at the same time ICP is monitored. An EVD may miss episodes of increased ICP when it is draining. This gold standard status, however, may take precedence because it was the first monitoring system available. Furthermore, there are no clinical outcome studies that show that one monitoring technology is superior to others. Because a ventricular catheter is placed into the ventricle, it is “logically” felt to optimally represent the ICP. However, even though gradients have been demonstrated between parenchymal and ventricular systems, none of these has ever been found to be of clinical significance. Parenchymal devices are easier to place, particularly when altered ventricular anatomy may limit ventricular catheter placement. However, intraparenchymal fiber-optic and electronic strain gauge systems are more expensive and cannot be recalibrated once in situ. Consequently, outside the ability of ventricular-catheter monitors to drain CSF as a potentially therapeutic maneuver, the choice of an ICP monitor is probably best decided based on factors such as accuracy, reliability, complication rates, ease of insertion, and cost.
Potential Complications of Intracranial Pressure Monitors
Ventricular catheters and ICP bolts traditionally have been placed by neurosurgeons. However, neurointensivists now are inserting these monitors on a more frequent basis. The complications rate appears to depend more on the technique used and the device (ventricular catheter vs. parenchymal monitor) than on who places the monitor once he or she has been adequately trained.
Infection
The definition of infection of an ICP monitor is somewhat controversial. It is easier to detect bacterial involvement of a ventricular catheter rather than a parenchymal monitor because CSF sampling can be periodically performed. In the absence of signs of ventriculitis, positive cultures may be better termed colonization . When this definition (colonization rather than infection) is used, positive cultures are observed in 8% of ventriculostomy CSF cultures and 14% of the cultures obtained from the tip of an intraparenchymal device once it is removed. In clinical practice, however, routine surveillance of intraparenchymal monitors is not performed. By contrast CSF sampling is common with ventricular catheters. Positive CSF cultures may prompt removal of the device and antibiotics that may lengthen hospital stay, expose patients to the potential hazards of parenteral antibiotics, and expose them to the risk of catheter replacement. Therefore although both types of monitors may be colonized, the consequences are markedly different. This must be balanced against the value of CSF drainage.
Iatrogenic catheter-related ventriculitis and meningitis may occur in 5% to 20% of EVDs through direct catheter contamination at insertion, through retrograde bacterial colonization, or if introduced when the system is accessed or flushed. The overall device infection rate is between 6 and 8 per 1000 drainage days. Risk factors for infection include the presence of concurrent systemic infections, longer duration of monitoring, presence of intraventricular hemorrhage or SAH, open skull fracture (with or without CSF leak), trauma, flushing of the catheter, and CSF leakage at the insertion site. Because most data are from case series, it remains uncertain how best to prevent or manage infections. However, strategies such as use of closed drainage systems, a long subcutaneous tunnel, reduced duration of ventricular cannulation (i.e., prompt removal when not needed), avoiding catheter irrigation or flushing or if needed done using aseptic techniques, and avoiding CSF leakage from any site are recommended. There does not appear to be a role for continuous antibiotic prophylaxis. Silver- or antibiotic-impregnated catheters may decrease the incidence of catheter-related CSF infections, although the exact role for these catheters is still debated. Because CSF sampling may predispose to contamination or infection, sampling should be indicated by specific clinical criteria rather than be a routine practice. Even a reduction of CSF sampling from daily to every 3 days can reduce the rate of ventriculitis from 10% to 3%. Use of standard management protocols particularly with a bundled approach also can help reduce the infection rate. There does not appear to be a role for routine replacement, because it appears that the risk of replacement outweighs any potential benefit. The use of parenchymal ICP monitors should be considered when there is systemic infection or an open skull fracture because there are no reports in the literature of infections associated with these devices in children or adults. There is, however, one report of an abscess that developed several weeks after an ICP monitor was removed from a 35-week-old child who required both prolonged ICP monitoring and steroids.
Hemorrhage
The precise incidence of hemorrhagic complications following ICP monitor insertion depends on what device is used, insertion technique, and how hemorrhages were identified. Ideally, imaging studies should be obtained before and after ICP monitor insertion, but few studies include such imaging. Retrospective case series suggest that the risk of intracranial hemorrhage after placement of an EVD in adults is between 2% and 10%. The risk may be greater in children (17.6%). Most hemorrhages are less than 15 mL in volume and therefore “clinically insignificant.” The incidence of clinically significant hemorrhages associated with ventriculostomy placement is about 1%. After insertion of an intraparenchymal ICP monitor in adults the incidence of intracranial hemorrhage is estimated to be between zero and 11% in adults. The incidence is similar in children. Gelabert-Gonzalez et al. described 1000 consecutive patients who had intraparenchymal ICP monitors; 922 had postinsertion CT imaging. The overall hemorrhage rate was 2.5%, and 0.66% required surgery for an intraparenchymal or epidural hemorrhage. Two studies have examined intraparenchymal and intraventricular devices concurrently. The hemorrhage incidence is significantly greater for ventriculostomies.
Technical Issues
Ventricular catheter displacement, accidental removal, or blockage may occur. When blockage occurs either by debris or blood that obliterates holes in the catheter or catheter displacement into the parenchyma, CSF drainage can generate a large pressure gradient between the catheter lumen and the ventricle. In this circumstance ICP is underestimated when it is monitored at the same time as CSF is drained and the CSF waveform becomes flattened. When this occurs the drainage system should be closed to see if the waveform returns. If not, the system may need to be gently flushed with 1 to 2 mL of normal saline. The role of thrombolytic agents to clear blood clots with intraventricular hemorrhage is unclear. Technical complications such as catheter breakage or dislodgement occur in about 4.5% of intraparenchymal devices. Most of these occur during transport, nursing maneuvers, or patient activities. None appear to influence the patient’s course or outcome. Different parenchymal devices including fiber-optic, strain gauge, and pneumatic technologies are available. Only the pneumatic Spiegelberg ICP monitor allows in vivo calibration. Bench testing shows that parenchymal monitors have excellent accuracy; however, studies suggest suspicion that zero-drift rates can be clinically important in strain gauge monitors. Drift is very rare in fiber-optic catheters. In addition, there are descriptions of disturbed baseline pressures of the Codman and Raumedic ICP sensors by electrostatic discharges.
Optimal Placement of Intraparenchymal Devices
Intraparenchymal ICP monitors are placed into the brain parenchyma through a small burr hole and a cranial access device or bolt that may allow placement of one to three monitors. Usually these monitors are placed into the nondominant frontal lobe when the injury or pathology is diffuse. However, in focal injury these monitors may be better placed on the side of the pathology or in pericontusional parenchyma because, unlike ventricular catheters, intraparenchymal devices have the potential to represent local, “compartmental” pressures, should supratentorial gradients exist. In particular, intraparenchymal devices placed in the hemisphere contralateral to a mass lesion may underestimate ICP even with brain herniation.
Some but not all clinical studies in TBI with older technology such as subarachnoid bolts (not recommended in the TBI evidence report ) often demonstrated interhemispheric gradients. Sahuquillo et al. using fiber-optic intraparenchymal monitors also found interhemispheric gradients in patients with focal lesions that could be greater than 10 mm Hg. One quarter of such gradients produced a difference in calculated CPP of greater than or equal to 5 mm Hg, which the authors felt had the potential to alter treatment. The authors therefore concluded that “to optimize ICP monitoring in patients with a mass lesion larger than 25 mL or midline shift, the measuring device should preferably be implanted in both the brain parenchyma and the side of the mass” ( Table 34.1 ).
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
