9 

ICP monitoring was first introduced by Lundberg et al1 in the early 1960s. Because such monitoring requires the expenditure of some resources, increases the work involved, and carries some risk of complications, documentation of its efficacy would be desirable. To date, there has been no prospective randomized trial proving that ICP monitoring improves outcome in patients with severe head injury. A study of this nature has not been conducted primarily because of ethical issues surrounding the need for a control group that would be treated empirically, without ICP monitoring. In addition, such a study would require approximately 349 patients in each treatment arm to demonstrate a 10% decrease in mortality [with an alpha (probability of a type I error, or erroneously rejecting a null hypothesis that is really true) of 5% and a beta (probability of a type II error, or erroneously failing to reject a null hypothesis that is false) of 20%] and cost over $5 million. In fact, a proposal for such a study in head-injured patients was submitted to the National Institutes of Health several years ago but did not receive funding.

Enthusiasm for ICP monitoring has grown because of a large body of clinical experience and anecdotal and historical evidence that it facilitates rational management, provides prognostic information, and improves outcomes. In 1991, Marmarou et al² (class II data) reported the results of an analysis of 428 severely head-injured patients from the Traumatic Coma Data Bank (TCDB). They noted that one of the strongest predictors of outcome in patients with severe head injury was the percentage of time ICP exceeded 20 mmHg. Specifically, as the amount of time ICP remained above 20 mmHg increased, poor outcomes became more likely, whereas favorable outcomes became less likely. Further evidence in favor of ICP monitoring has also come from studies focusing on cerebral perfusion pressure (CPP) (class II data), which is defined as mean arterial pressure minus ICP.³ Of course, treating elevated ICP and maintaining adequate CPP are impossible without continuous monitoring of blood pressure and ICP.

Most experts in the field of neurotrauma agree that ICP monitoring should be an integral part of the management of patients with severe TBI. However, when Ghajar et al4 (class III data) conducted a survey of 219 randomly selected trauma centers in 1991, they found that ICP monitoring was routinely used in the management of patients with severe TBI in only 28% of these centers, and 7% never used ICP monitoring, even though all surveyed hospitals indicated that they provided care for severely head-injured patients. Thus, in the early 1990s, the majority of trauma centers were not routinely treating their severe TBI patients in ways that could potentially improve outcome. These statistics improved after the publication of the Guidelines for the Management of Severe Head Injury⁵ Guidelines in 1995. In 1997, Marion and Spiegel⁶ (class III data) reported a survey of 3156 neurosurgeons who were surveyed by the American Board of Neurological Surgery. Of the 1262 respondents, 83% felt that ICP monitoring should be performed in patients with severe head injury.

Literature Review: Indications for Intracranial Pressure Monitoring

Normal ICP is 0 to 10 mmHg (0–135 mmH₂O), and most centers use 20 mmHg as the upper limit beyond which treatment is initiated. TBI patients with a GCS of 8 or less appear to be at risk for developing intracranial hypertension (ICH). Because accurate prediction of ICP based solely on the neurologic examination is not possible, measurement of ICP in patients with severe TBI gives the clinician objective data that may profoundly influence management. In 1979, Marshall et al⁷ (class III data) reported the results of 100 consecutive patients with severe head injury treated with a standard aggressive protocol that included ICP monitoring. They found that ICP exceeded 15 mmHg in 55% of patients, causing them to recommend ICP monitoring in patients with severe head injury. In 1982, in a prospective series of 207 consecutive patients with severe closed head injury, Narayan et al⁸ (class III data) reported that the incidence of ICH was 53 to 63% in comatose patients with an abnormal CT scan. Only 13% of patients with a normal admission CT scan developed ICH. However, if a patient had a normal admission CT scan but demonstrated at least two of three adverse features (age over 40 years, unilateral or bilateral motor posturing, or SBP <90 mmHg), the risk of developing ICH was similar to that of patients with abnormal CT scans (60%). Thus, head-injured patients who do not obey commands are at increased risk of developing ICH if their admission CT scans are abnormal or if they exhibit at least two of the adverse features listed above despite a normal admission CT scan.

On the other hand, patients with mild to moderate head injury (GCS 9–15), that is, those who are generally able to follow simple commands, are believed to have a relatively low risk for ICH and can be followed with serial neurologic exams. Less than 3% of patients with mild head injury (GCS 13–15) and approximately 10 to 20% of those with moderate head injury (GCS 9–12) will deteriorate into coma.⁹ Therefore, ICP monitoring is not indicated in most patients with mild to moderate head injury, although physicians may opt to place an ICP monitor in certain conscious patients with traumatic mass lesions.⁹

At some centers, therapies to reduce ICP have been initiated routinely even though ICP was not being measured. However, several of the therapies used to control ICP, including hyperventilation, mannitol, sedation, and paralysis, may have deleterious side effects.⁹ For example, in 1991, Muizelaar et al¹⁰ (class II data) reported the results of a prospective randomized study that found that severe prolonged hyperventilation resulted in poorer outcomes in patients with severe head injury. Regarding mannitol treatment, Kaufmann et al¹¹ in 1992 reported the results of a laboratory study that demonstrated that cumulative doses of mannitol can worsen cerebral edema by leaking into the interstitium. Lastly, although sedation, analgesia, and chemical paralysis reduce blood pressure elevations and ICP surges, their use makes it nearly impossible to follow and interpret the clinical exam. In such situations, ICP monitoring would be most valuable to allow early detection of brain swelling or intracranial hematoma formation.⁹ Furthermore, because the routine use of paralytic agents may increase the risk of pulmonary complications (class II data),¹² it has been recommended that paralysis be used to control ICP only if sedation is inadequate. ICP monitoring can eliminate unnecessary interventions and their potentially adverse effects.