Introduction
Infections in the intensive care setting are common, both as the reason for the intensive care unit (ICU) admission and as a nosocomial complication of critical illness. Nosocomial infections, now referred to as hospital acquired infections (HAIs), are thought to complicate between one quarter and one half of all ICU admissions. In the Extended Prevalence of Infection in Intensive Care (EPIC II) study, a 1-day, prospective, point prevalence study, 51% of 13,796 adult (older than 18 years) patients in 1265 participating ICUs from 75 countries were considered infected; 9084 (71%) were receiving antibiotics. The ICU mortality rate of the infected patients was twofold greater than that of noninfected patients (25% vs. 11%). Although central nervous system (CNS) infections are far less common as admission diagnoses or as complications of critical care, these infections, including the hospital-acquired complications of neurosurgical intervention and the community-acquired infections of meningitis, encephalitis, and brain abscess are the focus of this chapter. Each of these infections is discussed separately to provide insights on what tests are needed to make the diagnosis. Infections such as catheter-associated bloodstream infections (BSIs) and ventilator-associated pneumonia (VAP) are reviewed in Chapter 6 . Finally, the role of infection surveillance, prevention, and control is reviewed.
Bacterial Infections
Bacterial Central Nervous System Space-Occupying Lesions
Examples of space-occupying lesions of the CNS include parenchymal abscesses, epidural abscesses, and empyemas. The most common sources of brain abscesses are from hematologic spread (most common lung), direct extension from a parameningeal source (including the sinuses, middle ear, or after dental procedures, head trauma, or surgery), suppurative lung disease, or congenital heart disease (e.g., patent foramen ovale or patent ductus arteriosus, i.e., right to left shunting). Rare causes of brain abscess have been reported after tongue piercing or endovascular occlusion of aneurysms using Guglielmi detachable coils. Infective endocarditis occasionally presents first with a brain abscess.
Clinical Presentation of Bacterial Central Nervous System Space-Occupying Lesions
Parenchymal brain abscess may present with many symptoms. Although the classic triad of fever, headache, and focal neurologic signs is helpful, this occurs in less than half of patients, and the presentation may be more of a mass lesion than an infection. Other manifestations, including symptoms and signs of the original infection (otitis or sinusitis), may be present and more impressive. Although the average course of brain abscesses from the time of symptom presentation to hospital admission can be as short as 5 days, the course may be more indolent. Subdural empyema also may present with the triad of sinusitis, fever, and neurologic deficit. However, signs of cortical irritation such as seizures (50%), raised intracranial pressure (headache, vomiting, and papilledema [50%]), or focal deficits (75%) are a frequent presentation. Spinal epidural abscess is a neurosurgical emergency, and the typical presenting features include back pain, malaise, and fever. Neurologic deficits associated with spinal epidural abscesses often are associated with venous thrombophlebitis and so may evolve rapidly or be greater than the lesion size suggests.
Epidemiology
The frequency of brain abscesses in the United States ranges from 0.3 to 1.3 cases per 100,000 persons per year, with a predominance occurring in men between the ages of 30 to 40. Twenty five percent of all brain abscesses occur in children between the ages of 4 to 7, resulting from either cyanotic heart disease or extension from an otic source. The organism causing the infection depends in part on the source of the infection and the patient’s age. Common organisms in adults include anaerobic organisms, in particular Bacteroides and Peptostreptococcus in some series, whereas in others, aerobic organisms especially Staphylococcus, Streptococcus (most common), Enterobacteriaceae , and Haemophilus are more common. In infants Proteus and C itrobacter are the most frequent cause, whereas in children (3-5 years old) Haemophilus , Streptococcus pneumoniae , and Bacteroides fragilis (otogenic spread) are common. Staphylococcus, S. pneumoniae, and Haemophilus influenzae are common from sinus spread and Streptococcus or Staphylococcus aureus from cyanotic heart disease. Since 1981 the incidence of human immunodeficiency virus (HIV)–related brain abscesses due to toxoplasmosis has become the most common cause of a protozoal space-occupying brain lesion and must be considered in any patient presenting with a ring-enhancing brain lesion. These lesions may be single or multiple and should lead to expeditious HIV serologic testing. Aspergillus is the most common fungal organism to cause a brain abscess.
Subdural empyemas (SDEs)—a collection of purulence located between the arachnoid and dura—account for 15% to 20% of all space-occupying infections and result from extension of a suppurative process in a paranasal sinus or otorhinologic infection. Less commonly, subdural empyemas result from head trauma or after a neurosurgical procedure. Tewari et al., found the majority of SDEs are identified in infancy through the third decade of life (60%).
Spinal epidural abscesses usually occur as the result of hematogenous seeding of the epidural space from a distant suppurative focus such as infective endocarditis, from an infected central line catheter, or from the injection of intravenous drugs. S. aureus , and less commonly aerobic and anaerobic streptococci, and gram-negative rods cause the vast majority of epidural infections.
Diagnosis of Central Nervous System Space-Occupying Lesions
Fever with or without a peripheral white count and symptoms of new CNS deficit suggest an intracranial infection. Other signs and symptoms such as headache, vomiting, and papilledema are not specific to these diagnoses. Cranial imaging with computed tomography (CT) or magnetic resonance imaging (MRI) are sensitive tests; MRI is able to provide more information and better resolution. On MRI diffusion-weighted imaging (DWI) an intraparenchymal abscess demonstrates high signal on DWI and low apparent diffusion coefficient (ADC) in the abscess cavity. In subdural empyema, there may be a disproportionate amount of underlying cortical and white matter edema and enhancement relative to the size of the fluid collection. In patients with spinal epidural abscess there may evidence of diskitis and vertebral body infection. However, even imaging does not have the sensitivity required to definitively diagnose all infections. Lumbar puncture should not be performed in those with brain abscesses given the risk of herniation.
Laboratory and Microbiologic Analysis
Laboratory Studies
The most useful blood test is the complete blood count with a manual differential. Although an elevated white blood cell (WBC) count is always helpful, patients presenting with a significant suppurative process may have a normal peripheral WBC count. In these patients bandemia may be the only abnormal finding to guide the clinician toward making the diagnosis of an infectious process. It is imperative that the treating team request a manual differential on presentation. Furthermore, a markedly elevated platelet count such as one greater than 500,000 can also help suggest the presence of a suppurative CNS process. An elevated erythrocyte sedimentation rate (ESR) and C-reactive protein (CRP) may be useful but are nonspecific.
Microbiologic Analysis
Blood cultures should always be obtained on admission to identify a potential hematogenous source for the CNS or epidural infection. CNS-obtained material must be sent to a laboratory for Gram stain, routine aerobic and anaerobic culture, and in select patients for mycobacterial and fungal cultures. In the immunocompromised patient, a silver stain should be requested to identify Aspergillus species along with a request for modified acid-fast bacillus (AFB) staining to potentially identify a Nocardia species. In an HIV-infected patient pathologic specimens should be sent to a lab to look for Toxoplasma gondii tachyzoites. In the event that routine cultures fail to grow after 4 to 5 days, the clinician must consider whether he or she is dealing with a routine anaerobic process that failed to grow or a more unusual pathogen such as a Nocardia species, fungal or mycobacterial infection, or T. gondii in an HIV-infected patient. In these patients, consultation with an infectious diseases expert and the microbiology laboratory is advised.
Follow-up
The most effective approach to evaluate a patient’s response to therapy is to follow the patient’s temperature and WBC trend and the patient’s mental status and to obtain CNS imaging on a regular basis to assess structural improvement.
Management of Space-Occupying Infections
Treatment of brain abscess is both surgical and medical. Abscesses greater than 2.5 cm in diameter or those associated with mass effect should be excised or aspirated. Aspiration may be performed with image guidance and through a twist drill or burr hole. Brain imaging should be used to follow the treatment response, with repeat scans every 1 to 2 weeks.
In patients with ventriculitis, meningitis, or hydrocephalus that requires cerebrospinal fluid (CSF) drainage, or those with inaccessible abscesses or early abscess formation (cerebritis), medical treatment alone can be attempted preferably after microbiologic analysis on blood or CSF. Broad-spectrum antibiotics are started and refined when culture results are available. Empiric drug regimens for immunocompetent patients should include coverage for methicillin-resistant S. aureus (MRSA), anaerobes, and gram-negative bacilli (including Pseudomonas coverage in the setting of trauma or post neurosurgery) with a regimen such as vancomycin, metronidazole, and cefotaxime, typically for 6 to 8 weeks. The treatment response can be followed initially with weekly brain imaging. This can be spaced out to every 2 weeks during the remainder of the 6- to 8-week antibiotic course, with follow-up scans every 2 to 4 months for the following year to assess for recurrences.
The treatment of subdural empyema is surgical either through a craniotomy or burr holes. Early treatment (within 72 hours of symptom onset or sooner) is preferable. Broad-spectrum empiric antibiotics (similar to that described for brain abscess) should be started early. Once the cause of the infection is defined from culture analysis, antibiotics can be tailored to the organism and continued for 3 to 4 weeks after drainage. Spinal epidural abscess is a neurosurgical emergency. Without treatment, severe and permanent spinal cord damage can occur in just a few hours in part because of venous thrombosis rather than compression alone. Treatment includes surgical decompression and drainage, a search for the infectious source (if possible), and antibiotics (empiric as described earlier, then specific antibiotics based on the pathogens).
Outcome
The mortality of patients with brain abscess is between 20% and 30% and greater in those with a depressed level of consciousness or in coma at admission and distant metastatic focus of infection. Factors such as age, multiple abscesses, and steroid use do not appear to alter outcomes. Mortality in subdural empyema is between 10% and 40%. Younger age and early treatment is associated with better outcome. For example, Renaudin and Frazee observed that patients treated in less than 72 hours had less than 10% mortality, whereas 70% of patients died or were disabled when treated more than 72 hours after symptom onset. Outcome also is associated with neurologic status at presentation and whether sinus thrombophlebitis occurs and its extent. Outcome after spinal epidural abscess is associated with how quickly the diagnosis is made and surgical decompression is achieved and the severity of the spinal cord symptoms at the time of decompression. The more symptoms at the time of decompression, the worse the outcome, although compared with similar neurologic presentations in traumatic spinal cord injury, those with epidural abscess tend to do better.
Bacterial Meningitis
Epidemiology
S. pneumoniae (or pneumococcus) and Neisseria meningitidis remain the most common etiology for community-acquired bacterial meningitis. With the widespread use of H. influenzae –conjugated vaccine in children, the incidence of H. influenzae meningitis has decreased. S. pneumoniae (or pneumococcus) is now the most common cause of meningitis in the United States, although its incidence too is decreasing with the use of the multivalent pneumococcal vaccine. Group B streptococcus is the most common cause of bacterial meningitis in newborns but rare after the neonatal period, N. meningitidis in children, teens, and young adults and pneumococcus is the most common cause of bacterial meningitis in adults.
Clinical Features
The typical presenting signs and symptoms of bacterial meningitis are headache, fever, stiff neck, and altered mental status that each occur in 80% of patients. Kernig’s sign and Brudzinski’s sign are found in 50% of patients. In most cases of bacterial meningitis there also is associated inflammation of the underlying brain tissue (cerebritis). In addition, secondary venous sinus thrombosis, vasculitis, or cranial nerve inflammation may occur. Therefore focal neurologic deficits are seen in about 20% of patients and seizures in about 30% of patients. Papilledema at presentation is uncommon (<1%).
Diagnosis
The diagnosis of bacterial meningitis is made based on the clinical presentation and CSF examination. At the time of a lumbar puncture many patients have an increased opening pressure (>15 cm H 2 O). CSF total WBC count is in the 1000 to 10,000 range, and almost all patients with bacterial meningitis have a predominance of neutrophils. CSF glucose is usually decreased, although the CSF-to-serum ratio is more accurate. The CSF protein is almost always elevated (>100 g/dL). C-reactive protein (CRP), either in the serum or the CSF, or procalcitonin can help differentiate bacterial meningitis from viral meningitis. The specific diagnosis is made on CSF Gram stain and culture.
The Gram stain, developed by Hans Christian Gram in Berlin in the late 19th century, colors bacteria by using a crystal violet stain to colorize what are referred to as gram-positive organisms blue, namely streptococci and staphylococci. In a subsequent step the specimen is then decolorized with alcohol and counter-stained with the red dye, safranin. Gram-negative bacteria such as coliforms are unable to hold their crystal violet stain but take up the safranin stain, rendering these bacteria red. These procedures are now done exclusively in a microbiology laboratory. This procedure can guide antibiotic therapy until culture and sensitivity results are available.
A head CT scan is typically not required and not needed to diagnosis bacterial meningitis. However, it can be useful to exclude mass lesions that are a contradiction to lumbar puncture (LP). The literature supporting the necessity of doing a CT scan of the brain prior to LP is minimal. Between 3% and 4% of patients with meningitis develop herniation syndromes ; about 1% of these patients herniate within several hours of the LP (1%). However, a normal CT scan does not eliminate the risk for herniation nor does an abnormal CT scan predict herniation after LP. In 2004 Tunkel et al. wrote the “Practice Guidelines for the Management of Bacterial Meningitis.” A head CT is recommended before LP in patients with suspected bacterial meningitis who (a) have a history of CNS disease, (b) are immunocompromised, (c) present with a seizure, or have (d) papilledema, (e) an abnormal level of consciousness, or (f) a focal neurologic deficit. Furthermore, in the setting of suspected bacterial meningitis the decision to begin antimicrobial therapy should be made expeditiously and independent of when and if a head CT is performed because earlier treatment is associated with better outcome.
Management
Treatment of bacterial meningitis should not be delayed while waiting for a head CT scan. Empiric treatment of the patient should be provided based on the patient’s relative risk for specific organisms. Most immunocompetent adults are at highest risk for S. pneumoniae . Those greater than 50 years old are also at risk for Listeria monocytogenes . Teens and young adults are also susceptible to N. meningitidis (meningococcus). Empiric antibiotics should be administered based on these likely causes. Beta-lactam–resistant pneumococcus is becoming more prevalent in the community (up to 15% in some metropolitan areas in the United States) and so vancomycin should be used as an empiric treatment in the appropriate patient population. Once the Gram stain is obtained and organisms identified by morphology and staining characteristics the antibiotics can be modified to be more specific. The antibiotics can be changed once again if necessary once the organism has been grown and identified in culture and sensitivities have been obtained. Antibiotic treatment durations are based on the organism found. Both meningococcus and H. influenzae can be treated successfully for 7 days. S. pneumoniae requires 2 weeks of treatment, whereas Listeria and other gram-negative organisms require at least 3 weeks of treatment. In addition, an infectious disease consultant should guide antibiotic therapy.
The role of beta corticosteroids has been controversial since the early 1990s. A randomized clinical trial of dexamethasone (10 mg of dexamethasone every 6 hours for 4 days) that included 301 patients with bacterial meningitis has been published. The first dose of steroid was given before antibiotic administration. Outcome was better in those who received dexamethasone. This improvement in outcome was almost exclusively in the patients who had S. pneumoniae meningitis; however, due to a small sample size, conclusions regarding the effectiveness of steroids in H. influenzae and meningococcal meningitis are unclear. Because at the time of presentation the offending organism is unknown, the dexamethasone can be started before antibiotics. If an organism other than S. pneumoniae is found to be the offending pathogen, clinical guidelines for treatment of bacterial meningitis recommend against continued routine use of corticosteroids. Good supportive intensive care medicine is important in these patients. In bacterial meningitis increased intracranial pressure does occur, yet it is unclear whether monitoring of intracranial pressure or treatment with hypertonic agents such as mannitol or hypertonic saline is useful in this patient population.
Herpes Simplex Encephalitis
Epidemiology
Herpes simplex viruses (HSVs) are distributed worldwide, and humans are the sole reservoir of this virus. Between 25% and 60% of humans are seropositive for herpes simplex virus by adulthood. Herpes simplex encephalitis (HSE) is the most common cause of sporadic fatal encephalitis in the United States. It is estimated to occur in 1/250,000 to 1/500,000 in the United States or approximately 250 to 500 cases per year.
Pathophysiology
It is unclear whether herpes simplex encephalitis represents a primary or recurrent infection. The virus can lie dormant in neurons and ganglia once the patient has been infected. However, it has been demonstrated that patients with HSE and concomitant cutaneous herpes simplex may have two different strains of HSV. No specific triggers for HSE have been identified. Immunosuppression does not predispose patients to HSE. However, patients who are immunosuppressed are at risk for a more aggressive disease with a worse outcome.
HSE is always a cortical infection, and no cases of HSE have been reported to involve the brainstem. The temporal lobe is the most commonly infected brain region, although other lobes can develop encephalitis. The virus causes a hemorrhagic and necrotizing encephalitis.
Clinical Features
The common clinical manifestations of HSE are a change in personality, seizures, and decreased level of consciousness. Fever is common, and focal or generalized seizures occur in two thirds of patients. Focal neurologic findings such aphasia or neglect associated with a cortical abnormality or dysphagia and hemiparesis associated with cortical dysfunction may be seen. Headache, nausea, vomiting, and papilledema also are common. The clinical presentation may be subtle and somewhat slow moving.
Diagnosis
The diagnosis of HSE is made based on the clinical course, imaging (CT and MRI), and laboratory evaluation, in particular CSF analysis and CSF PCR. Electroencephalography and brain biopsy may be helpful in some patients, but brain biopsies are now infrequently performed in this setting. The CSF WBC count in patients with HSE is in the 10s to the 100s but values of 1000 to 2000 cell/mm 3 may be observed. Most of the WBCs are lymphocytes. However, 10% to 25% of patients can have high numbers of neutrophils in their CSF early in the disease course. Red cells in the range of 10s to 1000s/mm 3 are seen in the CSF of about 50% of patients and xanthochromia is common. CSF protein may be mild to moderately elevated in approximately 50% of patients. Glucose is usually normal but may be slightly reduced. Opening pressure is elevated in one third of patients. The CSF PCR for herpes simplex is now the test of choice for diagnosis and has a sensitivity of 96% and specificity of 99%. Very early in the disease (<24 hours) the test may be negative, and with treatment the sensitivity may decrease.
On MRI imaging, temporal lobe abnormalities, often hemorrhagic, are frequently found. Hypodensities in the temporal lobe or frontal lobes are found on CT in a severe case of herpes encephalitis. On MRI there are hyperintensities on T2 and gadolinium enhancement around the lesion in the lobe that is infected. Electroencephalograms (EEGs) are sensitive (84%) but nonspecific (32% specificity) for HSE. The characteristic EEG demonstrates spike and slow activity with periodic lateralized epileptiform discharge (PLED). Today brain biopsy rarely is required.
Treatment
Acyclovir (10 mg/kg every 8 hours for 21 days assuming normal renal function) is the drug of choice to treat HSE. Higher doses and a longer treatment duration may help prevent relapse. Outcome for HSE has improved significantly since the introduction of acyclovir, but mortality remains 19% at 6 months and 28% at 18 months. Predictors of poor outcome include age greater than 30 years and coma on presentation. In these patients mortality is nearly 70%. Significant neurologic sequelae are often observed in those that survive.
West Nile Viral Encephalitis
Characteristics of the West Nile Virus
The West Nile virus (WNV) is a single-stranded RNA virus that belongs to the family Flaviviridae. This family of viruses includes the Japanese encephalitis virus, the St. Louis encephalitis virus, and the Kunjin virus. There is serologic cross-reactivity with all of these viruses, and so those vaccinated against one may be seropositive against another. This virus was thought to have originated in the West Nile valley of Uganda. Genetic linkage of the WNV epidemic in the United States suggests that the viral serotype virus comes from the Middle East. Clinically, only the United States and Israeli WNV infections have caused severe human disease.
Epidemiology
WNV infection is now considered epidemic in the United States. Between 1999 and 2008 there were more than 31,000 cases and 1150 deaths (≈3.5% mortality). Of these cases nearly 12,000 (≈35%) were West Nile virus encephalitis (WNVE). The incidence of WNVE increases in patients greater than 50 years old; there is a 10-times higher risk of meningitis or encephalitis in those who are 50 to 59 years old and a 43-times higher risk of encephalitis in patients older than 80 years of age. There also is an increased risk of encephalitis in patients who are immunocompromised secondary to malignancy, transplant, or HIV.
Clinical Presentation
There is a clinical spectrum of West Nile viral infections that ranges from mild, almost asymptomatic cases to the most severe version, encephalitis. WNV infection without encephalitis is typically asymptomatic or a “flulike illness” with fever, malaise, myalgias, and headache. In those patients who develop WNVE, fever, fatigue, nausea, vomiting, headache, myalgias and altered mental status are common. Half of the patients may develop objective weakness that appears to be associated with anterior horn cell injury in the spinal cord appearing similar to a polio presentation. These patients also may develop compromised respiratory function or frank respiratory failure. A subset of WNVE patients can present with movement disorders associated with basal ganglia involvement.
Diagnosis
The diagnosis of WNV is made by examination of immunoglobulin M (IgM) antibodies in the serum or CSF. Because IgM does not cross the blood-brain barrier, the finding of IgM antibodies specific for WNV in the CSF indicates an acute infection with the virus. Vaccines to yellow fever and Japanese encephalitis are cross-reactive with the antibodies to WNV; however, this is IgG rather than IgM. The remaining CSF analysis is nonspecific: up to 2000 white cells, predominantly lymphocytes, a mild to moderately elevated protein, and normal glucose are found. Peripheral blood analysis typically is nonspecific; the WBC count is normal or slightly elevated, or lymphopenia may be observed. Patients who develop encephalitis may have hyponatremia associated with the syndrome of inappropriate antidiuretic hormone (SIADH) secretion. Imaging studies are nonspecific. Head CTs often are normal. However, one third of patients have enhancement of the meninges or periventricular areas on brain MRI. Abnormalities that involve the cortex, brainstem, basal ganglia, or spinal cord also may be seen on MRI.
Treatment and Prognosis
The care of patients who develop WNVE can be very resource intensive, particularly in those who develop motor weakness and use up many ICU patient-days. There is currently no proven treatment for WNV infection, although there are small case series using ribavirin, interferon alpha-2b, hyperimmune gammaglobulin, and a monoclonal antibody to the WNV. The mortality in those patients who develop WNVE (usually the elderly) is 10% to 18% and about half the survivors do not return to their premorbid function. In those who develop motor neuron damage up to two thirds have continued weakness. One year after discharge, 55% of patients report that they were not fully recovered; symptoms include fatigue, weakness, gait difficulty, and memory problems.
Infections After Neurosurgical Procedures
Background
Infections after neurosurgical procedures, including those associated with intracranial monitors or drains (e.g., external ventricular drains or lumbar drains), or after cranial or spinal surgical procedures are rare but when they occur contribute to prolonged length of stay in the ICU or hospital and adversely affect outcome.
External Ventricular Drains
The external ventricular drain or ventriculostomy was developed in the early 1950s. During the 1960s Lundberg reported positive CSF cultures in 6% of patients but no clinical infections in patients with ventriculostomies. Most reports that describe infections associated with external ventricular drains (EVDs) are retrospective: the incidence is between zero and 22%. Lozier et al. in a literature review of ventriculostomy-related infections, reported a combined infection rate between 8% and 9%. However, the manner in which ventriculostomy-associated infections are described varies greatly and not every paper defines what an “infection” is. There are several possible scenarios based on the ranked certainty of infectious ventriculitis: (1) ventriculostomy-related infection, (2) suspected ventriculostomy-related infection, (3) ventriculitis, (4) ventriculostomy colonization, and (5) contamination ( Table 17.1 ). According to this classification a positive CSF culture from an EVD means colonization or contamination when other CSF parameters are normal. When the CSF WBC count and protein also are abnormal, then there is infection in the face of a positive CSF culture. Infections associated with lumbar drain infection are about 3%, although fewer studies have addressed this question.
| Parameters | Ventriculitis | Ventriculostomy- Related Infection | Suspected Ventriculostomy-Related Infection | Ventriculostomy Colonization | Contamination |
|---|---|---|---|---|---|
| Cultures/GS | Variable | One or more culture or GS | Absence | Multiple positive cultures or GS | Isolated culture or GS |
| Glucose | Low | Declining | Declining | Normal/expected | Normal/expected |
| Protein | High | Increasing | Increasing | Normal/expected | Normal/expected |
| CSF WBC | Pleocytosis | Increasing pleocytosis | Increasing | Normal/expected | Normal/expected |
| Clinical | Fever, meningeal signs, photophobia, altered mental status | Fever | None | Sometimes fever | None |
Risk factors for infection associated with ventriculostomies include intraventricular hemorrhage, subarachnoid hemorrhage, open depressed skull fracture, basilar cranial fracture with CSF leak, neurosurgical operation, ventriculostomy irrigation, systemic infection, and longer duration of catheterization. However, the role of insertion duration remains unclear and it seems that if no infection has developed by 9 days, then the risk may decline. Most studies suggest that gram-negative bacilli and gram-positive cocci are the most common species that infect ventricular drains.
Intraparenchymal Pressure Monitors
The infection rate associated with intracranial pressure monitors (“bolts”) is reported to be between zero and 4%, much lower than that associated with ventriculostomies. In part this may result from the limited, if any, manipulation of a bolt once inserted, whereas ventriculostomies may be sampled and drugs injected into them.
Craniotomy Infections
Infections including meningitis or ventriculitis complicate about 4% of craniotomies. The bacteria associated with these infections are typically skin flora such as Staphylococcus , Streptococcus, and less frequently gram-negative bacilli. Risk factors for a craniotomy infection include emergency surgery, longer duration of surgery, reoperation during the same admission, bleeding complications, CSF leak, or the surgeon who performs the craniotomy.
Diagnosis
The diagnosis of true ventriculostomy-associated infection has the following characteristics: (1) progressive decline in CSF glucose, (2) increasing CSF protein, (3) progressive CSF pleocytosis, (4) one or more positive CSF cultures or Gram stain, and (5) fever. Similar findings are evident when there is ventriculitis, but the patient is far sicker. A reduced CSF-to-serum glucose ratio can help in diagnosis. A ratio of 0.4 has a sensitivity of 77% and a specificity of 87%. Other methods including CSF lactate, cytokine analysis, and cell index studies are useful but lack sensitivity and specificity. The routine analysis of CSF to help predict infection does not appear to be helpful and instead the frequent access to the system may increase the risk of contamination or infection. Furthermore, in a prospective patient cohort in which CSF was sampled every 3 days compared with a historical control group sampled more frequently, Williams et al. demonstrated a significant reduction in ventriculitis with less frequent sampling.
Prevention
The data supporting the role of antibiotic prophylaxis or changing the catheter to help prevent EVD associated infections are conflicting. In a randomized trial Poon et al. compared continuous prophylactic ampicillin/sulbactam plus aztreonam to a single dose at the time of the procedure. Continuous antibiotic prophylaxis was associated with fewer infections but greater mortality if an infection occurred. Retrospective data suggest that the risk of infection increases once an EVD has been in place between days 5 and 9 but that routine changes may not help prevent this. For example Wong et al. in a randomized trial of 103 patients observed a similar infection rate whether catheters were changed every 5 days or not. Some institutions use prophylactic vancomycin or other antibiotics in this setting but data as to its efficacy are lacking.
Treatment
When there is evidence of infection, treatment should begin immediately. If cultures are available, the antibiotics should be directed at the appropriate organisms. Without available cultures or Gram stains, antibiotic choice should be directed at gram-negative bacilli (including Pseudomonas ) and both Streptococcus and Staphylococcus (including S. aureus ).
Related posts:
Stay updated, free articles. Join our Telegram channel
Full access? Get Clinical Tree
