CHAPTER 23 Intracranial Pressure Monitoring
Historical Perspective
The importance of intracranial pressure (ICP) was first recognized by Alexander Monro more than 200 years ago and is now referred to as the Monro-Kellie doctrine or the Monro-Kellie hypothesis.1 The Monroe-Kellie doctrine states that (1) the brain is housed in the nonexpandable skull, (2) brain parenchyma is fairly noncompressible, and (3) the volume of blood is relatively constant and outflow of venous blood is necessary for the inflow of arterial blood.1 Later, cerebrospinal fluid (CSF) was recognized as a component of brain volume in addition to brain parenchyma and blood and was incorporated into the doctrine. If there is a new intracranial mass lesion such as tumor or hematoma or an abnormal increase in the volume of any of the components, such as CSF (during hydrocephalus) or parenchyma (during brain edema), the volume of venous blood or CSF, or both, will decrease to accommodate. However, this compensatory reserve is limited, and any further increase in the volume of the pathologic lesion will lead to an increase in ICP because of the rigid, nonexpandable skull. An increase in ICP will then result in a decrease in perfusion pressure and cerebral blood flow and eventually cerebral herniation and death.1
For more than a century there has been clinical interest in measuring ICP. Early efforts to measure ICP were based on the observation that because the cranial and spinal CSF compartments communicate with each other, their pressure should be equal. Measurement of spinal CSF pressure through lumbar puncture should therefore reflect cranial CSF pressure, or ICP.1,2 However, it was soon recognized that measurement of opening pressure via lumbar puncture is associated with a risk for cerebral herniation in the presence of an intracranial mass lesion and that it may not reflect ICP if there is any obstruction to CSF flow between the cranial and spinal CSF compartments.
During the first half of the 20th century, several investigators had measured ventricular fluid pressure in a small number of patients.2 Its clinical use, however, was limited until the 1960s, at which time the pioneering neurosurgeon Nils Lundberg started to measure ICP with a ventricular catheter connected to an external strain gauge pressure transducer and a standard ink-writing potentiometer recorder.2 Drainage of CSF was also used to reduce ICP.2 This method of ICP monitoring and CSF drainage was used in more than 400 patients, many of whom had traumatic brain injury, and marked the beginning of the modern era in ICP monitoring.3
General Principles and Standard of Intracranial Pressure Monitoring Technology
Today, ICP monitoring is an integral part of neurocritical care. ICP monitoring has been used in the management of patients with traumatic brain injury, subarachnoid hemorrhage, intracranial tumor, intracranial hemorrhage, stroke, hydrocephalus, central nervous system infection, and fulminant hepatic failure.4
Since the 1960s, there has been a continuous effort to develop new technology for ICP monitoring. Nils Lundberg outlined the basic requirements for an ICP monitor, which still apply today: minimal trauma during placement, negligible risk for infection, no CSF leakage, easy to handle, reliable, and able to continue to function during various diagnostic and therapeutic procedures.3 The Association for the Advancement of Medical Instrumentation has developed the American National Standard for Intracranial Pressure Monitoring Devices, which specifies that an ICP monitoring device should have a pressure range between 0 and 100 mm Hg, accuracy of 2 mm Hg in the range of 0 to 20 mm Hg, and a maximal error of 10% in the range of 20 to 100 mm Hg.5 Throughout the years, many different ICP monitors have been developed, but only very few are in active clinical use today.
Current Intracranial Pressure Monitoring Technology
External Ventricular Drain
An external ventricular drain (EVD), or ventriculostomy drain, connected to an external strain gauge is currently the “gold standard” for measuring ICP.5 It remains the preferred method for monitoring ICP among U.S. neurosurgeons.6 An EVD can be placed at the bedside in the emergency department, intensive care unit (ICU), or the operating room, depending on local practice tradition. Most practitioners use anatomic landmarks (freehand technique) to insert the ventricular drain into the lateral ventricle with the tip in the foramen of Monro.6,7 The catheter can then be tunneled subcutaneously to minimize CSF leakage and infection.8 Ventricular fluid pressure, which represents ICP, is transmitted to an external strain gauge transducer via the fluid-filled EVD. The strain gauge transducer can be recalibrated without manipulation of the EVD. It can be connected to many standard ICU monitoring systems and allows ICP measurements to be displayed along with other physiologic data such as pulse, blood pressure, or central venous pressure.
Advantages of the EVD as an ICP monitoring device include its extensive history, low cost, and reliability.5,9 Most importantly, an EVD can also serve as a therapeutic device to remove CSF and lower ICP.2,5 In patients with subarachnoid or intraventricular hemorrhage, in which the elevated ICP is frequently due to hydrocephalus, an EVD is the most appropriate ICP monitoring device given its monitoring and therapeutic capabilities. However, an EVD has several weaknesses. Accurate placement of an EVD may be difficult with the freehand technique. In a recent survey of practicing neurosurgeons and residents, the success rate of cannulation of the ventricle was just 82%, even in the hands of practicing neurosurgeons.6 Currently, there is an EVD placement guide available that may increase the accuracy of placement of EVDs, although it is not widely used in the neurosurgical community.7 In some patients, it is simply not possible to place an EVD because of the small size of some ventricles or ventricular shift as a result of a mass lesion or severe edema.
Complications from EVD placement for ICP monitoring and CSF diversion include malposition, occlusion, hemorrhage, and infection. The malposition rate of EVDs ranges between 4% and 20%.10–13 Most of the misplaced EVDs did not have any significant clinical sequelae, but about 4% of these EVDs did require replacement.10,11,13 Occlusion by brain matter or blood clot occurs frequently, especially in patients with intraventricular or subarachnoid hemorrhage.10 Most of the occlusions can be resolved by flushing the EVD catheter.10 Hemorrhage secondary to placement of an EVD occurs infrequently. Hemorrhage rates ranging from 0% to 15% have been reported in the literature, with an average rate of 1.1%.5,11,13–15 Fortunately, most patients are asymptomatic from EVD-associated hemorrhage.11,15 Clinically significant hemorrhage requiring surgical evacuation occurs about 0.5% of the time, and results in intracerebral, subdural, and epidural hematoma.11,13,15 Coagulopathy is thought to be associated with an increase in the hemorrhage rate, and therapeutic anticoagulants and antiplatelet agents are also known to be associated with an increased risk for hemorrhage.16,17 In addition, laceration of a cortical artery can lead to traumatic pseudoaneurysm formation, and this complication has been reported with placement of ICP monitors.18
The most significant risk associated with an EVD is infection. Lozier and coworkers performed an extensive review of all the literature on infection associated with EVDs.19 The range of infection in all the series was 0% to 22%, with a cumulative incidence of 8.8%.19 More recent studies have found a similar rate of infection as well.20–22 Clinical characteristics that have been identified to be associated with increased EVD-associated infection include intraventricular hemorrhage, subarachnoid hemorrhage, craniotomy, CSF leakage, systemic infection, and depressed skull fracture.19,20 Technical factors that may contribute to CSF infection include the duration of catheterization and irrigation of the catheter.19 In 17 studies reviewed by Lozier and colleagues, 10 were found to have an association between the duration of catheterization and infection, whereas 7 did not find such association.19 Careful inspection of the raw data of the latter group showed that there was an increased risk for infection after day 10 in one study.19 More recent studies also show an increased risk for infection with a longer duration of catheterization.20,22,23 Most studies report that there are few infections during the first 5 days of drainage and monitoring with an EVD but that the infection rate increases significantly after 5 to 10 days of catheterization.13,24,25
Because of the relatively high rate of CSF infection in patients with EVDs, multiple interventions have been used in an attempt to minimize the infection rate. However, most studies are retrospective in nature and often do not have enough statistical power to detect small absolute differences in the incidence of infection.19 Such interventions are discussed in the following sections.
Venue of External Ventricular Drain Placement
Lozier and coworkers analyzed five studies that looked at whether there is a difference in infection rates in EVDs placed in the operating room, ICU, or emergency department.19 All but one of the studies revealed no significant difference in infection rate whether the EVD was placed in the operating room, ICU, or emergency department.19 Two more recent studies also did not find statistically significant differences in infection rate related to the venue of ventriculostomy drain insertion.22,23
Extended Tunneling
Subcutaneous tunneling of the EVD catheter was reported by Friedman and Vries in 1980 as a way to reduce the infection rate.8 Other investigators then extended the distance of tunneling to the upper part of the chest or abdomen and had an infection rate of 4%.26 Two recent studies reported conflicting results. Sandalcioglu and Stolke reported that there was a significant difference in infection rate (83% versus 17%) for catheters that were tunneled less than 5 cm subcutaneously versus catheters that were tunneled more than 5 cm, respectively.27 However, patient details were not available, and the infection rate of 83% in this study is significantly higher than most reported infection rates. Leung and coauthors, in contrast, did not find a significant difference in infection rate with long-tunneled EVDs.28 Most EVD insertion kits today contain a trocar for subcutaneous tunneling in excess of 5 cm.
Prophylactic Catheter Exchange
The observation that the infection rate rises with increased duration of drainage from an EVD prompted several investigators to advocate prophylactic catheter exchange.13,25 Several retrospective studies, however, showed that there was no benefit of prophylactic catheter exchange and that in fact there was a higher incidence of infection in the group in which catheters were routinely exchanged.19 This was also observed in one prospective, randomized trial comparing the infection rate in a group that underwent prophylactic catheter exchange versus a control group that did not undergo prophylactic catheter exchange.29
Prophylactic Antibiotic Use
Many studies have analyzed the use of prophylactic antibiotics for reduction of the infection rate of EVDs. Prophylactic antibiotics can be given periprocedurally only or administered during the entire duration that the catheter is in place. Studies in the 1970s suggested that prophylactic antibiotics did reduce the infection rate when compared with no antibiotics, but two studies conducted in the 1980s and one study in 2000 did not find any reduction in the EVD infection rate in patients who received periprocedural antibiotics versus patients who did not.19,30,31 Several other studies also compared prophylactic antibiotics given just periprocedurally versus during the entire duration when the EVD is in place. In a large retrospective study, Alleyne and colleagues did not find any significant difference in infection rates between the two groups.32 In a prospective, randomized controlled study, however, Poon and associates did find a reduction in CSF and systemic infection in the group that received prolonged antibiotic prophylaxis.33 It should be noted that infections that develop in patients who receive prolonged or broad-spectrum antibiotic prophylaxis, or both, for EVD placement are often caused by more virulent microorganisms such as Candida and gram-negative organisms.32–34 Currently, the “Guidelines for the Management of Severe Traumatic Brain Injury” does not recommend antibiotic prophylaxis for EVD placement/catheterization.35
Antibiotic-Impregnated Catheter
A recent development in EVD catheter technology is the antibiotic-impregnated catheter. An EVD catheter impregnated with rifampin is capable of releasing rifampin in a controlled-release manner. These EVD catheters have been shown to significantly reduce bacterial adhesion versus controls in vitro and in animal models.36 In one randomized controlled trial, Zabramski and coworkers showed that a catheter impregnated with minocycline and rifampin reduced the infection rate significantly from 9.4% to 1.3% when compared with the nonimpregnated catheter control group.37 Although antibiotic-impregnated catheters cost significantly more than nonimpregnated catheters, one also has to consider the overall expense and length of hospital stay for patients with EVD catheter–related infections.
Fiberoptic Intracranial Pressure Monitor
Fiberoptic devices for ICP monitoring in which the catheter tip measures the amount of light reflected off a pressure-sensitive diaphragm were developed in 1980s.38 The most widely studied fiberoptic device is the Camino fiberoptic ICP monitoring device (Integra Neuroscience, Plainsboro, NJ). The Camino fiberoptic ICP monitoring device can be placed in the subdural, intraparenchymal, and intraventricular space.
The intraparenchymal Camino ICP monitor has the most extensive clinical experience. The main strength of the intraparenchymal ICP monitor is ease of insertion. The most commonly used technique involves insertion of the monitor into the right frontal region, although it is possible to insert the probe into a region with pathology as well. However, it should be noted that compartmental pressure differentials had been observed in different regions of the brain, so the choice of where to insert the monitor is important.39,40 Because there is no need to cannulate the ventricular system, it is possible to insert the ICP monitor even in patients with severely compressed ventricles or those with a significant midline shift. Although this device has led to more widespread use of ICP monitoring in critically injured and ill patients, there are reports of some technical issues that should be kept in mind when using these devices.
A number of early studies demonstrated that there is a high correlation between ICP measured by the intraparenchymal Camino ICP device and ICP measured by an EVD, with a correlation coefficient (r) of greater than 0.9 in both studies.41 However, Schickner and Young found that the Camino overestimated ICP by an average of 9 mm Hg when compared with an EVD in 10 patients.42 Nevertheless, because of its ease of insertion and low complication rate, it has gained popularity since its introduction into the market. Many large clinical series involving the intraparenchymal Camino device have been published. In most clinical series it is the sole ICP monitoring device inserted, and overall clinical experience with the Camino as an ICP monitoring device has been positive.39–41,43,44 However, there have not been any large prospective or retrospective series that have compared the intraparenchymal fiberoptic device with an EVD as an ICP monitoring device in neurocritical care patients.
