This patient’s clinical presentation indicates an infection of the central nervous system (CNS) and systemic compromise, as signs of sepsis and septic shock are evident by the initial assessment performed in the ED. The initial evaluation of patients with a suspected CNS infection should include a detailed clinical history, assessment of epidemiologic factors, risk factors for infection, and medical comorbidities. The initial neurologic assessment provides important prognostic information and allows for comparison of serial neurologic examinations. This patient should be isolated in the ED, and droplet precautions should be maintained until a final etiologic diagnosis is made. After initial assessment and determination of a potential CNS infection, initial steps in the management of this case should include an evaluation of the ABCs (airway, breathing, and circulation), assessment of the hemodynamic status, collection of blood and cerebrospinal fluid (CSF) samples, and initiation of appropriate antimicrobial therapy. Patients with suspected meningitis who present with abnormal mental status or neurologic deficits, especially those with a GCS ≤ 12, require intensive care unit (ICU) admission for observation (Table 8-1).

Airway

Rapid neurologic deterioration and ensuing loss of consciousness with impairment of reflexes that maintain the airway mandate permanent airway control (Table 8-2).¹ Failure to recognize imminent airway loss may result in complications such as aspiration, hypoxemia, and hypercapnia. Preferred induction agents for rapid sequence intubation (RSI) in the setting of suspected brain injury with high intracranial pressure (ICP) include propofol² and etomidate,³ both of which are short-acting agents that will not obscure the neurologic examination for a prolonged period of time. Adverse effects of propofol include drug-induced hypotension that usually responds to fluid infusion.³ Adverse effects of etomidate include nausea, vomiting, myoclonic movements, seizures (by lowering of seizure threshold),³ and adrenal suppression.⁴ Midazolam may be an alternative, but unfavorable effects on the ICP have been reported with the use of this agent.^2,5 Succinylcholine is the most commonly administered muscle relaxant for RSI owing to its rapidity of onset (30-60 s) and short duration (5-15 min).⁶ However, side effects of succinylcholine include hyperkalemia, cardiac arrhythmias, exacerbation of neuropathy or myopathy, malignant hyperthermia, and elevation of ICP in patients with intracranial mass lesions.^3,7 For this reason, in neurologic patients, a nondepolarizing neuromuscular blocking agent such as cis-atracurium,⁸ rocuronium,³ or vecuronium is preferred if needed.⁹ In patients with increased ICP, premedication with IV lidocaine for RSI is of questionable use, but sometimes has been suggested.¹⁰

Breathing

The goal of treatment is to ensure oxygenation and ventilation at an adequate oxygen level and maintenance of normocarbia, as this is associated with good effects on the cerebral blood flow (CBF). Both hypoxia and hyercarbia are detrimental for the CBF and have the potential for increasing cerebral edema and development of high ICP.¹¹

Circulation

Isotonic fluid resuscitation and vasopressors are indicated for brain-injured patients in shock.^12,13 Dextrose-containing solutions should be avoided as hyperglycemia may be detrimental to the injured brain.¹⁴ Initial assessment of the volume status by placement of a central venous catheter for measurement of central venous pressure (CVP) is recommended by current Surviving Sepsis Campaign Guidelines to achieve a CVP of 8 to 12 mm Hg and a goal mean arterial pressure (MAP) of 65 mm Hg or more with crystalloids, colloids, or vasopressors.¹³ Urine output should be greater than 0.5 ml/kg/h, and superior vena cava oxygenation (ScvO₂) should be greater than 70%.¹³ Although the recently published Protocolized Care for Early Septic Shock (ProCESS) trial¹⁵ failed to show benefit from early goal-directed therapy, the Surviving Sepsis Campaign committee only recognized the results but has not formally updated the guidelines, and thus the above recommendations are still encouraged. The influence of such treatment on the CBF of these patients is unknown. In the healthy person, cerebral autoregulation maintains a constant CBF with MAP ranging from 60 to 130 mm Hg. However, when autoregulation is impaired, as it may be in severe CNS infections, there is a risk of cerebral hypoperfusion as well as ischemia when MAP decreases and a risk of hyperperfusion leading to vasogenic edema when MAP increases. An association between cerebral ischemia and poor neurologic outcome or death has been demonstrated in various studies of bacterial meningitis.^16,17 Cerebral perfusion pressures (CPPs) < 30 mm Hg were strongly correlated with death or major neurologic sequelae in infants and children with meningitis.¹⁸ Taken together, these findings suggest that maintenance of an adequate CPP, primarily by manipulating MAP, would prevent cerebral ischemia, attenuate brain damage, and improve the prognosis.

Dopamine, norepinephrine, and phenylephrine are frequently used in the ICU to restore CPP by increasing the MAP. Selection of the initial vasopressor is frequently guided by the clinical characteristics of the patient as well as the goals of therapy. Consideration should be given to the effect of vasopressors on cerebral hemodynamics. The disruption of the blood-brain barrier (BBB) in CNS infections may allow for these agents to have direct effects on the cerebral vasculature, although this may be theoretical. Several studies have demonstrated that norepinephrine may increase the CBF with no net effect on ICP, global cerebral metabolism, and oxygen consumption, suggesting that the net effect on CBF is related to abnormal autoregulation.¹⁹ Norepinephrine should be used as the first choice vasopressor in patients with septic shock and can be complemented with vasopressin if needed.¹³ Epinephrine may be added or can be a substitute for norepinephrine.¹³ Dopamine can be used as an alternative but only in highly selected patients who have low risk for tachyarrhythmias and/or bradycardia.²⁰

A LP is essential to obtain CSF and to make the definitive diagnosis of a CNS infection. Opening pressures are usually elevated and may be in the range of 20 to 50 cm H₂O (15-35 mm Hg).²¹ Complications after LP are variable, but the most feared complication is life-threatening herniation. A theoretical pressure gradient with downward displacement of the cerebrum and brainstem can be triggered by an LP. However, studies addressing this phenomenon have found that herniation occurs > 8 hours after the LP.²² In a study of 129 adult patients with elevated ICP, 1% with papilledema and 12% without papilledema had unfavorable outcomes within 48 hours after LP.²³ This supports the possibility that herniation after LP in the setting of a space-occupying lesion may occur in patients who are going to herniate even if the LP is not done. In a recent study of 301 adults with bacterial meningitis, abnormal findings in the physical examination associated with an abnormal head CT scan were age 60 years or older and had history of CNS disease (tumor, stroke, focal infection), an immunosuppressed state (human immunodeficiency virus [HIV], acquired immune deficiency syndrome [AIDS], chemotherapy, and posttransplantation), a history of new-onset seizures, abnormal level of consciousness, and abnormal neurologic examination (eg, gaze preference, visual field defects, and paralysis).²⁴ On the basis of this study, the Practice Guideline Committee for the Treatment of Bacterial Meningitis recommends a CT scan of the brain in all patients with these abnormalities.²¹ CT scans are essential to evaluate the patient for complications of CNS infections such as cerebral edema or collections (abscess or hemorrhage) and should be conducted in all patients even though some data suggest that CT scans are not very sensitive when predicting herniation after an LP.^25,26 The decision to obtain a CT scan prior to an LP should not affect the timing of administration of antibiotics.

After obtaining an LP, cerebrospinal fluid should be examined for glucose content, protein level, and cell count with differentials. Hypoglycorrhachia, polymorphonuclear pleocytosis, and elevated protein levels are usually indicative of a bacterial infection. Viral infections, on the other hand, tend to have milder protein elevations and normal glucose concentrations. Cell count differentials tend to demonstrate a mononuclear pleocytosis, although polymorphonuclear cells may initially predominate.²⁷ Findings that predict bacterial origins are as follows: a glucose concentration < 34 mg/dL, a CSF-to-serum glucose ratio < 0.23, a protein level > 220 mg/dL, and a pleocytosis of > 2000 cells per μL.²⁸ Gram staining is reported be positive in 60% to 90% of cases. When there are more than 10⁵ colony-forming units (CFUs) per mL, 97% of stains are positive, decreasing to 60% with 10³ to 10⁵ CFUs per milliliter.²⁹ Along with performing a Gram stain, India ink capsule and acid-fast stains should be considered, depending on likelihood and exposure. Cultures will provide the identity in 70% of bacterial cases. Viral and fungal cultures can also be obtained, depending on the patient’s history. If viral infection is suspected, a panencephalitis panel should be ordered, which should include herpes simplex virus types 1 and 2 (HSV-1 and HSV-2), West Nile virus, and enterovirus. The sensitivities vary for these PCR tests based on the organism, from 61% to 88%, and have specificities of > 95%.²⁹

In addition to the LP, blood cultures should be obtained, as they will be positive in close to 50% of cases.³⁰ Laboratory tests for inflammatory markers such as erythrocyte sedimentation rate (ESR) and C-reactive protein, as well as procalcitonin levels, can be ordered as these may help differentiate a bacterial from a viral infection. On the basis of the patient’s history, a test for HIV may help demonstrate an immunocompromised state, which would change treatment for the suspected organism.

The key to the treatment of CNS infections is early delivery of an appropriate empiric antibiotic regimen.³¹ Selection of the empiric regimen will vary depending on the suspected organism and the local drug-resistance patterns for that organism. The suspected organism type varies based on several patient factors including age, immune status, predisposing conditions, and other comorbidities. It is extremely important that patients are assessed for the most likely causative organisms and that IV treatment is delivered for those potential causes. Because of the lack of prospective data, current guideline recommendations do not provide a recommended time frame from onset of symptoms to delivery of an appropriate antibiotic regimen; however, experience in patients with septic shock suggests the earlier the better.^32,33 Available studies demonstrate a reduced mortality rate and fewer neurologic complications in patients who received timely administration of antibiotics,^34,35 and retrospective observational data show that delay in antibiotic therapy was independently correlated with unfavorable outcomes.³⁶ If antibiotics are administered prior to LP, they may diminish the yield of the bacterial Gram stain or culture of the CSF by 20%.²¹ Therefore, when LP cannot be conducted prior to antibiotic delivery, it should be performed as soon as possible after antibiotic administration to minimize the effect on the Gram stain and culture (see Chapter 53, Figure 53-2. Example guideline for selection of appropriate empiric antimicrobials in severe sepsis and septic shock).

In addition to timely administration of an appropriate antibiotic regimen, the dosage of the antibiotic is extremely important. Transport of drugs across the BBB is dependent on several factors, including, but not limited to, lipophilicity, protein binding, and the presence of inflammation, and many commonly used antibiotics exhibit poor penetration into the CNS.³⁷ Owing to limited penetration and systemic toxicities associated with some IV therapies, some practitioners have utilized the intraventricular route of administration to treat severe infections.²¹ After identification of the causative organism, the antibiotic regimen should be reviewed and adjusted to the most effective available antibiotic.

Empiric treatment regimens should also include treatment for viral encephalitis pending results of diagnostic testing. Numerous viruses have been reported as causes of encephalitis, although delivery of empiric antiviral therapy is typically limited to infections due to the herpes viruses. In addition, the empiric regimen may need to include doxycycline in patients who present with signs and symptoms suggestive of rickettsial or ehrlichial infections during the appropriate seasons. Empiric treatment regimens should not routinely include drugs for other causes of encephalitis. These treatment regimens should only be started once a specific viral cause is identified. Clinical practice guidelines are available for the treatment of encephalitis and should be used to determine the appropriate treatment for these infections.³⁸

Duration of treatment should last for a minimum of 7 days regardless of the organism isolated and can last 3 to 4 weeks if Listeria is being treated.²¹ Streptococcus pneumoniae should be treated for 10 to 14 days, whereas Streptococcus agalactiae should be treated for 21 days. Neisseria meningitides and Haemophilus influenzae can be treated with 7 days of antibiotics.²⁹ Patients who are immunocompromised will likely need longer treatment times depending on their clinical response.

Experimental models of bacterial meningitis have demonstrated that the inflammatory response in the subarachnoid space is a major contributing factor to the associated morbidity and mortality.²¹ A study in rabbits demonstrated that hearing loss develops early in the course of meningitis and is preceded by the abrupt onset of inflammatory changes in CSF.³⁹ The majority of the data pertaining to use of corticosteroids in CNS infections comes from the pediatric population. A meta-analysis of all studies conducted in infants and children from 1988 to 1996 demonstrated reductions in hearing impairment in patients with infection due to H influenzae and protection against severe hearing loss in patients with S pneumoniae if corticosteroid therapy was started prior to or with the first dose of antibiotics.⁴⁰ To date, only one randomized, double-blinded, placebo-controlled trial has demonstrated a significant mortality rate reduction in adults.⁴¹ In this study, patients received dexamethasone, 10 mg IV every 6 hours for 4 days, with the first dose being administered 15 to 20 minutes prior to antibiotics. Patients randomized to dexamethasone had significant reductions in unfavorable outcomes and death, but notably, the only subgroup of patients who experienced statistically significant reductions in unfavorable outcomes (26% vs 52%, P = .006) and mortality rates (14% vs 34%, P = .02) were those with pneumococcal meningitis; however, the other subgroups had relatively few patients. A follow-up to this study was designed to examine the potential harmful effect of adjunctive dexamethasone treatment on long-term neuropsychological outcome.⁴² The overall rate of cognitive dysfunction was not affected by dexamethasone treatment. The study did find a statistically significant higher rate of cognitive dysfunction in patients after pneumococcal meningitis (21% vs 6%; P = 0.05) than those after meningococcal meningitis.

A recent Cochrane meta-analysis compared outcomes from 4121 individual patients entered into 25 randomized trials. Glucocorticoids were associated with lower rates of severe hearing loss (relative risk [RR], 0.67; 95% CI, 0.51-0.88) or any hearing loss at all (RR, 0.74; 95% CI, 0.63-0.87).⁴³ When outcomes were divided between studies in high- vs low-income countries, steroids were only beneficial in high-income countries. No difference was recorded in long-term neurologic sequelae between glucocorticoid-treated patients and controls (RR, 0.90; 95%, CI 0.74-1.10). For difference in mortality, there was a nonsignificant trend toward decreased mortality in adults who received glucocorticoids (RR, 0.74; 95%, CI 0.53-1.05); however, there was a reduced mortality in patients with meningitis caused by S pneumoniae (RR, 0.84; 95%, CI 0.72-0.98).

Despite the available data, there is still controversy about the use of steroids in meningitis based on the fact that antibiotic penetration into the CSF relies on meningeal inflammation, and administration of dexamethasone may decrease the inflammatory response. A rabbit study demonstrated that dexamethasone significantly reduced the penetration of vancomycin into CSF.⁴⁴ It is uncertain whether these results can be applied to human models, but in a study that examined concentrations of vancomycin in CSF in which 93% of patients were treated with steroids, the mean CSF concentration was higher than the minimum inhibitory concentration.⁴⁵

Based on 2013 Cochrane meta-analysis⁴³, dexamethasone should be given to adults in developed countries because the data regarding developing countries is not as convincing. This may be secondary to delayed clinical diagnosis, access to healthcare, higher incidence of HIV infections, more malnutrition, and reduced life expectancy. Nevertheless, current guidelines recommend the initiation of dexamethasone at a dose of 0.15 mg/kg IV every 6 hours for 2 to 4 days 10 to 20 minutes before or with antibiotic therapy in all infants, children, and adults with suspected bacterial meningitis.²¹ Steroids should not be administered following antibiotics because they are unlikely to be of benefit. In adults without evidence of pneumococcal meningitis on CSF Gram stain or blood culture, corticosteroids should be discontinued. Current data do not support a recommendation for their use in herpetic encephalitis.³⁸

Elevated ICP with herniation and compression of the brainstem is the most frequent cause of death in patients with CNS infections.^46,47 Increased ICP in bacterial meningitis is secondary to the development of cerebral edema. Additionally, high ICP could be the result of increased intracranial blood volume from venous congestion due to thrombotic venous occlusion of the cerebral sinuses or due to arteriolar dilation from impaired autoregulation and high CBF.^48,49 It is important to remember that neither CT scan nor papilledema can predict a high ICP in the acute setting, but intracranial hypertension may be suspected in stuporous or comatose patients, those who present acutely with clinical signs of brainstem herniation (ie, pupillary abnormalities or motor posturing), and those with abnormal CT scans that suggest brain shifts and mass effect.

If intracranial hypertension with imminent herniation is suspected, the head should be elevated to 30°, 1.0 to 1.5 g/kg of 20% mannitol should be administered by a rapid infusion, and the patient should be hyperventilated to a Paco₂ of 26 to 30 mm Hg. As an alternative, or if the patient is relatively hypotensive, 0.5 to 2.0 mL/kg of 23.4% hypertonic saline (HS) solution can be administered through a central venous line;⁵⁰ however, there is no evidence for the use of either agent in the treatment of high ICP in CNS infections. HS does have some advantages over mannitol in infection-related cerebral edema.

The osmotic reflection coefficient of the brain capillaries to sodium is 1.0 compared with 0.9 in the case of mannitol, indicating that HS does not effectively cross the brain capillaries. During the first few hours of a bolus of HS, the concentration of sodium in the CSF does not change, forming the basis of efficacy of HS as an effective osmotic agent in brain edema.⁵¹ Therefore, in infection-related cerebral edema, where permeability or integrity of the BBB is disturbed, less-permeable agents such as HS provide more osmosis than higher permeable agents such as mannitol. HS has neuroprotective effects due to its anti-inflammatory action,^52,53 whereas mannitol prevents biochemical injury owing to its free-radical scavenger effect.⁵⁴ Moreover, mannitol is relatively contraindicated in hypovolemic patients because of the diuretic effect; HS is superior in hypovolemic and hypotensive patients and would be the preferred agent in this patient and in those with severe sepsis or in septic shock. Adverse effects of HS include fluid overload, hematologic and electrolyte abnormalities, such as bleeding secondary to decreased platelet aggregation and prolonged coagulation times, hypokalemia, and hyperchloremic acidosis.⁵⁵ The serum sodium level should never be allowed to drop > 12 mEq/L over 24 hours, as rapid withdrawal of hypertonic therapy may result in rebound cerebral edema, leading to elevated ICP and/or herniation syndromes.^55,56 Additional adverse effects of mannitol include paradoxical cerebral edema and rebound high ICP, based on its propensity to cross the BBB, drawing free water into the CNS. Other side effects include hyperosmolarity and renal failure, specifically when serum osmolarities are > 320 mOsm/L, which will wash out the renal medullary gradient.

An ICP > 20 cm H₂O (15 mm Hg) should be aggressively treated to prevent cerebral herniation and irreversible brainstem injury. Studies in patients with meningitis have shown that the mean ICP was significantly higher in nonsurvivors compared with survivors.⁴⁷ When an ICP monitoring device is in place, maintenance of an adequate CPP and a normal ICP should be attempted at all times. CPPs ≤ 50 mm Hg were found to be associated with a 100% death rate in patients with bacterial meningitis,⁵⁷ and death occurred in 75% of the patients in whom the initial ICP was > 40 mm Hg.⁴⁷ Therefore, it is important to recognize signs of intracranial hypertension and institute adequate monitoring and aggressive treatment. High ICPs may be lowered successfully in most patients with bacterial meningitis by different measures and using unconventional volume-targeted (Lund concept) ICP management.⁵⁸

For cases of severe or intractable elevated ICP, barbiturates and induced therapeutic hypothermia are also effective tools to control refractory elevated ICP by decreasing cerebral metabolic activity, which translates into a reduction of the CBF. These two techniques require expertise, advanced tools, and continuous monitoring of cerebral electrical activity and may be associated with significant complications.

The frequency of clinical seizures in bacterial meningitis is 5% to 27%⁵⁹ and in viral encephalitis is 62% to 67%,⁶⁰ and the presence of encephalitis and a GCS < 12 were independent predictors for the occurrence of clinical seizure.⁶¹ The incidence of seizures is significantly increased in patients with pneumococcal meningitis (24% vs 5% with meningococcal meningitis).⁶² Additional risk factors for seizures are tachycardia, a low GSC score on admission, infection with S pneumoniae, and focal cerebral abnormalities.⁶² In this study, patients with seizures had a CSF leukocyte count < 1000 cells/mm³ (36% vs 25%; P = 0.01), had higher median CSF protein levels (4.8 vs 4.1 g/L), and higher median ESR (46 vs 36 mm/h; P = 0.02). There are no data to suggest the role of prophylactic antiseizure medication, but suspicion of ictal activity should be raised in those patients with abnormal movements, persistent coma, or altered sensorium.⁶³ In fact, seizures in acute meningitis are often a poor prognostic sign. An observational cross-sectional study, prospective nationwide cohort of 696 of patients with community-acquired bacterial meningitis, death occurred in 41% of patients with seizures compared with 16% of patients without seizures (P < 0.001).⁶² Continuous electroencephalographic (cEEG) monitoring may be indicated in these patients and may help with the identification of subclinical ictal activity and patterns associated with poor outcome, such as periodic lateralizing epileptiform discharges (PLEDs),⁶⁴ but there is still no consensus on whether to treat these patterns and, if so, how aggressively.⁶⁵ Patients who develop seizures during meningitis do have a risk of developing epilepsy after resolution. A population-based cohort of 199 survivors of meningitis developed a 20-year risk of unprovoked seizures of 13% for patients with bacterial meningitis and early seizures and 2.4% for patients with bacterial meningitis without early seizures.⁶⁶