Ventilated-associated pneumonia (VAP) is one of the most common complications in mechanically ventilated critically ill patients. VAP was previously defined by signs of systemic infection developing within 48 h after intubation and a new infiltrate on chest x-ray. In 2015, the Centers for Disease Control introduced new definitions for ventilator-associated condition (VAC) and VAP. The new definitions require a sustained increase in oxygenation requirements after a period of stability coupled with signs of inflammation or infection. Prolonged intubation, which is the greatest risk factor for VAP, is frequently required for the neurocritically ill patient due to conditions such as severe TBI or status epilepticus requiring pharmacological coma. Decreased level of consciousness, loss of airway protective reflexes, and positive pressure provided by mechanical ventilation are factors that contribute to the risk of developing VAP [10].

In neurocritical care, an accurate diagnosis of VAP can be challenging. The patient may be febrile due to noninfectious causes, and bronchoalveolar lavage (BAL) is less commonly performed in patients at risk for elevated intracranial pressure [4]. For diagnosis and treatment purposes, it is important to obtain a sputum sample, either by endotracheal suctioning, BAL, or mini-BAL. The mini-BAL is often preferred due to the lower risk of increasing intracranial pressure. Calculating the clinical pulmonary infection score (CPIS) is useful in diagnosing VAP. A CPIS of greater than 6 has a strong correlation with VAP. Variables used to calculate the CPIS include:

Common organisms that cause VAP are listed below (Table 23.1). Antibiotic sensitivity of individual organisms varies by institution. Broad-spectrum initial antibiotic therapy with narrowing of coverage after isolation of the organism helps to reduce development of multidrug-resistant organisms. Organisms that tend to be drug resistant are typically seen in late-onset VAP, which is defined as VAP occurring after more than 4 days of intubation [10].

Bundles are widely used to standardize practices and help decrease the incidence of hospital-acquired infections. Elements in the Institute of Healthcare Improvement’s VAP bundle are listed in Fig. 23.3 [8].

Historically, the endotracheal tube has been maintained at 2–4 cm above the carina on chest x-ray and adjusted daily after reviewing the radiograph. In recent studies, endotracheal tube repositioning has been associated with approximately a threefold increased risk in VAP [17]. A wider margin for endotracheal tube depth can be tolerated to reduce the occurrence of tube repositioning, which decreases the chance of VAP from oral secretions draining into the airway. Other factors that have been identified to decrease the incidence of VAP include [10]:

External ventricular drains (EVDs) are frequently used in neurocritical care to treat a variety of conditions. While the drainage of CSF from an EVD can be lifesaving and necessary, the presence of an EVD increases a patient’s risk of developing a HAI. The CDC definition of ventriculostomy-related infection (VRI) states that the patient must have clinical symptoms of infection, abnormal laboratory findings indicative of infection, and positive CSF cultures, whereas some studies define VRI as the presence of positive CSF cultures alone [7]. Due to the urgent nature of EVD placement and previous absence of formalized guidelines, insertion technique and care of EVDs tend to vary widely and have not been evaluated in large studies. In 2016, the Neurocritical Care Society published an evidence-based consensus statement on the insertion and management of EVDs.

Increased risk of infection has been associated with longer duration of EVD; thus it is recommended that the EVD be removed as soon at the clinical situation allows [5]. Despite striving to remove EVDs as soon as possible, many neurocritically ill patients do require EVDs for an extended period. Therefore, to mitigate the risk of infection, the use of antimicrobial-impregnated EVDs, in addition to the administration of a single dose of a prophylactic antibiotic prior to the EVD insertion, is recommended. Historically, ongoing prophylactic antibiotics were prescribed throughout the duration of the EVD. This is not currently recommended due to the associated increased risk of the development of Clostridium difficile colitis and infections from drug-resistant organisms [5]. The use of a care bundle for EVDs has been shown to reduce rates of VRI (Fig. 23.4).

If a VRI is detected in the neurocritical care patient, antibiotics that cross the blood-brain barrier must be chosen. At times, it may be necessary to treat VRI with intraventricular antibiotics to achieve high CSF concentrations in multidrug-resistant infections or in patients who do not have an adequate response to intravenous antibiotics [5].

Fever is defined as elevation of core body temperature above 38.0–38.5 °C. Fever is a very common systemic complication encountered in neurocritical care patients and has a higher incidence in the neurocritical care unit compared to other ICUs [15]. For patients with intracranial temperature probes, brain temperature should be monitored and recorded. For patients without intracranial temperature probes, the core temperature should be monitored from esophageal or bladder temperature probes. Independent of the underlying cause, fever is associated with both acute and chronic neurological problems. For patients with brain injury and encephalopathy, elevated brain temperature is associated with worsening encephalopathy and lower seizure threshold [3]. In addition, fever increases cerebral metabolic demands which can worsen ischemic in the setting of brain hypoperfusion due to cerebral vasospasm and stroke. In observational studies, fever has been independently associated with increased chance of death and worse neurological outcomes in ischemic stroke, intracerebral hemorrhage, subarachnoid hemorrhage, traumatic brain injury, and hypoxic-ischemic encephalopathy from cardiac arrest. The area under the fever curve – referring to both the degree and duration of fever – is strongly associated with poor outcomes in patients with severe brain injuries. Aside from cardiac arrest survivors, there are currently limited data to prove that fever control results in improved outcomes so few guidelines exist. The topic of targeted temperature management can be referenced in Chap. 17 “Hypoxic Ischemic Injury After Cardiac Arrest.”

Fever can be produced by a variety of systemic and neurological causes in patients with injuries to the brain and spinal cord. These can include both infectious and noninfectious etiologies. See the Table 23.2 for the differential diagnosis of fever in the neurocritical care unit. Patients with acute brain injury can be particularly susceptible to noninfectious causes of fever, in particular patients with subarachnoid and intraventricular hemorrhage. The term “central fever” is used to describe fever caused by dysregulation of temperature homeostasis caused by injury to the hypothalamus either due to direct involvement or inflammatory hemolysis in cerebrospinal fluid bathing the hypothalamus located in the walls of the third ventricle. It can be challenging to differentiate patients with infectious and noninfectious causes of fever, which contributes to overuse of empiric antibiotics in neurocritical care patients. One group found that nearly half of neurocritical care patients with fever had noninfectious etiologies [6]. They found that independent predictors of central fever – as opposed to infectious fever – were:

Based on these findings, the combination of negative cultures, absence of infiltrate on chest x-ray, diagnosis or subarachnoid/intraventricular hemorrhage or brain tumor, and fever onset within 72 h was strongly predictive of central fever (positive predictive value 90%) [6]. In the absence of guidelines, these criteria provide a practical approach to discontinuation of empiric antibiotics. Other groups have explored using serum procalcitonin levels as a biomarker for infectious causes of fever, but this approach remains investigational.

Suppression and treatment of fever can be challenging in neurocritical care patients and often require multiple pharmacological and non-pharmacological strategies deployed in concert. Shivering represents the greatest impediment to fever control. Shivering results from involuntary systemic muscle contraction which dramatically raises resting energy expenditure and the systemic rate of oxygen consumption which produces heat as a byproduct of metabolism [1]. Using indirect calorimetry, the resting energy expenditure can increase by as much as 2.5-fold during severe shivering [1]. The increase in metabolic demand and associated temperature increase may negate the potential salutary effects of temperature modulation in patients with acute brain injuries. The recognition and treatment of shivering should be an integral part of any neurocritical care unit’s targeted temperature management protocol. Hospitals that are most successful at fever and shivering control utilize a standard nurse-driven assessment and treatment algorithm and physician order set. The most commonly applied recognition tool for shivering is the Bedside Shivering Assessment Scale (BSAS), outlined in Table 23.3 [1]. The BSAS requires regular nursing assessments to evaluate for the presence and severity of shivering. Shivering can be insidious and surprisingly difficult to detect, especially in patients who are clothed or covered by bed linens or surface temperature management devices. In addition to physical examination, display of the EKG and bispectral index (BIS) EEG waveforms on the bedside monitor can assist in recognition of shivering.

Institutions should adopt a standard treatment protocol for targeted temperature management and shivering control. As an example of a fever control algorithm, see below. Note, this protocol is intended for intubated and mechanically ventilated patients with severe brain injury and persistent temperature >38.0 °C.