Definition, Diagnosis, and Clinical Features

Acute brain dysfunction can occur when there is an imbalance of the brain’s homeostatic reserve and acute stressors. The term acute brain dysfunction most commonly refers to delirium but may also include coma. Delirium is an acute disorder of attention and cognition [5]. A comprehensive psychiatric evaluation using criteria based on the Diagnostic and Statistical Manual of Mental Disorders (DSM), 5th edition, from the American Psychiatric Association [6], is considered the gold standard for diagnosing delirium. Important diagnostic features of DSM-5 delirium include sudden onset of altered consciousness, reduced capacity to maintain one’s attention and awareness, and disorganized thought process, all of which cannot be better explained by another neurocognitive disorder or severely reduced arousal. Pragmatically, the DSM-5 definition may be challenging to apply in clinical settings, especially in geriatric critically ill patients who may have baseline cognitive impairment or neurologic injury [7]. The mental status therefore must be an acute change from baseline impairments and fluctuate throughout the day.

Clinically, delirium is often suspected when one has fluctuating attention and consciousness causing impaired awareness. One may also exhibit disorientation, episodic memory disorders, and delusions or hallucinations that lead to an abnormal perception of the environment. A number of well-validated approaches for diagnosing delirium in the critical care setting exist, with the most widely recognized methods being the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) [8] and the Intensive Care Delirium Screening Checklist (ICDSC) [9]. The CAM-ICU, originally described by Ely et al., is a four-element diagnostic algorithm that trained healthcare professionals can easily employ in less than 2 minutes [8]. It assesses for acute changes/fluctuations in mental status, inattention, disorganized thinking, and an altered level of consciousness (Table 5.1). The ICDSC assesses eight diagnostic features of delirium over an entire nursing shift (altered level of consciousness, inattention, disorientation, psychosis, altered psychomotor activity, inappropriate speech/mood, sleep disturbance, and symptom fluctuation). Given the fluctuating nature of delirium, it is imperative that delirium assessments be performed serially. Of special challenge in geriatric critical care is delirium superimposed on dementia and delirium in those with primarily neurologic insult (i.e., stroke). The CAM-ICU has demonstrated high sensitivity and specificity, and the Richmond Agitation-Sedation Scale [10] has demonstrated moderate sensitivity and high specificity for delirium superimposed on dementia [11,12]. The CAM-ICU has also been validated for poststroke delirium assessment and should be employed for the detection of delirium in this subset of elderly patients [13].

Delirium can have varying clinical presentations, and three different motor subtypes of delirium (hypoactive, hyperactive, and mixed) have been described. Hypoactive delirium is characterized by symptoms of lethargy, decreased movement, and slowed mentation, whereas the hyperactive subtype manifests as agitation, heightened arousal, or aggression [14]. Additionally, a patient may have features of both subtypes, which is referred to as a mixed delirium [14]. Hypoactive delirium has been found to be the predominant subtype among elderly patients [15]. It is less clinically apparent than the hyperactive subtype, which may lead to a delayed diagnosis. In addition to diagnosing delirium and subtypes, the severity of the delirium may be rated using the Delirium Rating Scale–Revised-98 [16] or the Confusion Assessment Method–Severity (CAM-S) [17].

Postoperative delirium (POD), a commonly used clinical term, is recognized as a medical diagnosis code that is a subset of delirium in the tenth edition of the International Statistical Classification of Diseases and Related Health Problems (ICD-10) system. Postoperative cognitive dysfunction (POCD), another widely used term, is not recognized in the ICD-10 system. POCD typically refers to deficits that have a longer duration (weeks to years) than delirium (hours to days).

Outcomes Associated with Acute Brain Dysfunction

Previously, acute brain dysfunction was thought to be a transient, reversible, and self-limited process [18–22]. More recently, it has become clear that acute brain organ dysfunction in critically ill patients is a predictor of worse clinical outcomes and that delirium may have significant long-term consequences. Patients with delirium take longer to wean from mechanical ventilation and have longer ICU and hospital lengths of stay [23]. Furthermore, they are more likely to require institutionalization or to be readmitted after discharge [23,24]. Patients with delirium therefore have higher hospital costs [25,26]. In general, costs associated with delirium are estimated to be over $160 billion per year in the United States alone [5]. Delirium in the ICU has also been associated with increased risk of death, in particular when it persists for multiple days or after sedation has been discontinued [23,24,27,28]. Studies in surgical patients focusing on POD have found significant associations with increased length of stay, higher cost of care, readmission to the hospital, higher rates of institutionalization after discharge, and increased mortality [26,29,30]. Surgery combined with early postoperative neurologic dysfunction, including even minor reductions in performance (defined as a Mini Mental Status Exam score < 24), has recently been shown to be a negative prognostic factor among elderly patients with hip fractures with a nearly 15 percent mortality rate [31], which is double that of patients without delirium. Studies have now begun examining the association of subtype of delirium and severity of delirium with clinical outcome. In a study of elderly patients admitted to the ICU postoperatively after an elective surgery, patients suffering from hypoactive delirium had increased 6-month mortality compared with patients suffering from other subtypes [32]. Increased hospital length of stay and mortality have been associated with increased severity of delirium [17]. Poststroke delirium has been associated with increased hospital length of stay [13].

Although delirium represents acute brain dysfunction, it has additionally been linked to long-term cognitive impairment [33–36]. Recent observational studies of critically ill patients have shown that nearly three-fourths of survivors have cognitive impairment 1 year after the hospitalization [33] and that advancing age and duration of acute brain dysfunction were significant risk factors for worse global cognition [34]. Although the relationship has yet to be fully elucidated, POD appears to be associated with POCD [34–36] Furthermore, delirium has been shown to accelerate cognitive decline in patients with Alzheimer’s disease [37,38] and be associated with worse cognitive decline in patients with and without dementia [39]. Thus long-term changes in cognition are a recognized complication associated with delirium in the geriatric population [33–36], but whether or not this reflects progression of the underlying pathology from delirium is unclear [40]. A better understanding of the disease process is needed prior to determining whether these are distinct entities or a single coherent syndrome. This is vitally important to patients because there is evidence that older adults with limited life expectancy may value preservation of cognitive status over survival [41]. Neither baseline nor post–critical illness cognitive status is routinely measured in a rigorous manner [42]; therefore, it has been difficult to determine which strategies are effective in preserving long-term brain function after critical illness, especially in the high-risk elderly patients.

Incidence

The incidence of delirium in elderly patients has been reported as high as 50 to 80 percent, with the highest incidence among critically ill patients on mechanical ventilation [5,33]. Data demonstrate that most delirium in the ICU goes undiagnosed without a regular screening tool [43], and it is universally acknowledged that the incidence of delirium is likely to be underreported, with up to 50 percent of cases being undiagnosed [44]. Hypoactive delirium in particular is easily and frequently overlooked by healthcare professionals due to a lack of awareness of the importance of its recognition and lack of education on how to perform a formal delirium assessment [45,46]. Current critical care guidelines therefore recommend routine delirium screening, which is particularly important in elderly patients given their increased risk [47].

Risk Factors

Delirium diagnosis identifies the constellation of acute brain dysfunction signs but does not identify the etiology. It should therefore prompt further investigation into potential risk factors for delirium. Delirium risk factors are numerous and can be stratified into predisposing and precipitating factors (Table 5.2). Diminished preoperative cognitive status is probably the biggest risk factor for delirium in the elderly population [48]. Age greater than 75 years and cerebrovascular disease have also been identified as risk factors for delirium specific to the elderly [48]. Frailty, which is common in the elderly and refers to critically reduced or impaired functional reserves that may involve multiple organ systems, has been shown to be independently associated with a greater risk of developing delirium [49–51].

Other risk factors for delirium include lower levels of education, major comorbid disease, major surgery, acute renal failure, vision or auditory disturbances, alcoholism, infection, and electrolyte disorders [5,52]. Use of physical restraints, use of urinary catheters, malnutrition, and acute pain have also been reported as risk factors [53]. High and low mean arterial pressures, as well as overall blood pressure fluctuations, have been associated with increased risk of delirium in elderly individuals, but the optimal blood pressure target has not been determined with regard to delirium [54–56]. While transfusion and blood loss can be associated with delirium, liberal versus restrictive packed red blood cell transfusion targets found no difference in either delirium rates or severity in patients undergoing hip surgery [57,58].

A handful of sedative medications has been associated with delirium development, including benzodiazepines and opioids. Longer-acting benzodiazepines and opioids with active metabolites are more likely to be associated with increased risk of delirium. This is likely related to drug accumulation due to the normal physiologic changes that occur with aging, such as renal and hepatic insufficiency [59]. Benzodiazepine infusions to provide sedation for mechanically ventilated patients are strong contributors to delirium compared with other sedative regimens [60–62]. Deep levels of sedation also carry a higher risk of delirium [60]. General anesthetics have been hypothesized to contribute to both postoperative cognitive dysfunction and dementia, but the evidence is inconsistent, and clear causal association has yet to be established [63–66]. Furthermore, most recent studies have not demonstrated a link between general anesthesia and cognitive impairment [34,67–70].

Other medications hypothesized to precipitate delirium through altered neurotransmission include those with anticholinergic, serotonergic, or dopaminergic properties (commonly used hospital medications include classic antihistamines, cyclobenzaprine, meperidine, famotidine, scopolamine, benztropine, oxybutynin, and tricyclic antidepressants). Treatment with corticosteroids has also been linked to delirium, although it has not been fully elucidated as to whether this is due to the severity of illness and ensuing increased neuroinflammatory response or the known psychological side effects of steroid use [71].

Pathophysiology

The pathophysiology of delirium and its associated long-term cognitive impairment is presumed to be multifactorial due to the complex interactions between the aging brain, baseline comorbid conditions, acute insult, and decreased cognitive reserve. In general, elderly individuals are thought to have less physiologic reserve than their younger counterparts and therefore have less ability to maintain homeostasis to stressors, making them more susceptible to an exaggerated response [72]. Neuroinflammation, endothelial dysfunction leading to cerebral perfusion abnormalities and increased permeability of the blood-brain barrier, cholinergic deficiency and neurotransmitter imbalances, cerebral atrophy and global brain disorders, and modifiable clinical risk factors such as certain pharmacologic therapies have all been shown to play a role in the development of delirium. The majority of current studies investigating the pathophysiology of delirium have focused on the hypothesis that systemic insults lead to inflammatory signaling and increased permeability of the blood-brain barrier, resulting in neuroinflammation and neuronal injury or death [73–76].

The inflammatory response, resulting from a systemic insult such as major illness or major surgical procedure, leads to the widespread release of cytokines and mediators. These cytokines, which are a physiologic response to illness or surgery, then cross into other organs that are not inflamed such as the brain. In the brain, they likely cause an abnormal response such as microglial activation, neuronal apoptosis, and altered synaptic plasticity and long-term potentiation [77]. High baseline levels of interleukin 6 (IL-6) and tumor necrosis factor alpha (TNF-α), high oxidative stress, mitochondrial dysfunction, high free-radical production, cellular senescence, and dysregulation of the hypothalamic-pituitary-adrenal axis and sympathetic nervous system result in a deviant stress responses that leads to dysregulation of neuronal activity [78,79].

Endothelial cells lining the cerebral microvessels form a selective barrier known as the blood-brain barrier [80]. Blood-brain barrier permeability increases as a result of normal aging [1]. Infectious and inflammatory processes and pain stimulate the production of IL-6 and cause endothelial dysfunction, further increasing blood-brain barrier permeability [3]. Elevated plasma biomarkers of endothelial dysfunction such as plasminogen activator inhibitor-1, E-selectin, and S100B (which is a marker of disruption of the blood-brain barrier) have been shown to be elevated in critical illness, and higher levels have been associated with prolonged duration of delirium [73].

A significant number of disturbances in neurotransmitter function have been described as potential contributors to delirium. Normal aging is associated with a decrease in the synthesis of major brain neurotransmitters such as acetylcholine, dopamine, serotonin, and glutamate [81]. Central cholinergic deficiency has also been hypothesized to play a major role in delirium [14]. This hypothesis originated in the early 1980s when it was noted that delirium occurred with consumption of toxins and drugs that impair cholinergic function [82]. This has been further supported by neuroimaging showing areas of brain where cholinergic projections overlap with lesions associated with acute brain dysfunction [77]. Additionally, use of anticholinergic drugs in the perioperative period has been associated with an increased incidence of delirium in elderly patients. Human trials of cholinesterase inhibitors, however, have not demonstrated a benefit in prevention or management of delirium [82,83]. Dopamine, serotonin, and norepinephrine are integral to arousal and sleep-wake cycles and are mediated by cholinergic pathways. Increased levels of dopamine and serotonin have been associated with hyperactive delirium. Additionally, elevated CNS IL-6 levels are associated with delirium development and with degeneration of gamma-aminobutyric acid (GABA)ergic interneurons [84,85]. GABA is an inhibitory neurotransmitter, and GABA agonists such as benzodiazepines have been associated with increased incidence of delirium, which supports the theory that GABAergic dysregulation is a contributing factor for delirium [86]. It is the imbalance of these neurotransmitters that is believed to play a role in delirium, although the exact mechanism has been elusive [82].

Neuroimaging has consistently shown anatomic changes that occur as part of normal aging. Elderly persons experience a prominent loss of volume and thickness in the prefrontal cortex [87,88]. This area plays a prominent role in attention and executive function and is consistent with previous psychological experiments showing decreases in performance on tests of attention and executive function with aging [89,90]. Presumably these neuroanatomic changes that occur with normal aging predispose elderly individuals to the physiologic alterations that result from a systemic insult; application of before and after delirium multimodal neuroimaging, however, should be employed to further characterize this relationship. To date, neuroimaging studies have shown delirium to be associated with brain atrophy and decreased white matter integrity. A study of critically ill patients has shown that increasing duration of delirium in the hospital is associated with increased brain atrophy several months after discharge, including reductions in hippocampal and superior frontal lobe volume [91]. Additionally, increasing duration of delirium in the hospital is associated with increased white matter disruption several months after discharge, in particular in the anterior limb of the internal capsule and genu of the corpus callosum [92].

Nonpharmacologic Interventions

First and foremost, delirium education programs for healthcare professionals have been found to consistently reduce hospital delirium rates [93,94]. These programs have focused on recognition of delirium, screening for delirium, risk factors for delirium, and approaches for prevention and management [94]. Despite initial costs and time, these programs have been shown to be cost-effective [95]. Reorientation, sleep enhancement using a nonpharmacologic sleep protocol and sleep hygiene, early mobility, adaptations for visual and hearing impairment, nutrition and fluid repletion, pain management, and prevention of constipation have all been strongly recommended by the American Geriatric Society as interventions for the prevention of delirium [96]. Inouye et al.’s Hospital Elder Life Program is an innovative model of care for the prevention of functional and cognitive decline of elderly hospitalized individuals. It employs trained staff and volunteers to help with nonpharmacologic prevention techniques such as those listed earlier, including reorientation, cognitive stimulation multiple times each day, assistance with ambulation or range-of-motion activities, visual and hearing aids when needed, feeding assistance, and sleep protocols, and is a potentially cost-effective model to improve outcomes in elderly hospitalized patients [97] (Table 5.3).

A coordinated approach to systematically lightening sedation, liberation from mechanical ventilation, avoidance of benzodiazepine sedation, routine delirium monitoring, and early mobility have consistently been shown to reduce delirium rates in ICU patients. The Awakening and Breathing Coordination, Delirium Monitoring/Management, and Early Exercise/Mobility (ABCDE) bundle was originally published in 2011 [98] and has been shown repeatedly to decrease delirium rates, mechanical ventilation duration, and hospital length of stay [99,100]. Additionally, widespread adoption of this type of bundle has been shown to improve survival and increase the number of days alive without delirium or coma [101].

“Prehabilitation” (rehabilitation-like activities prior to the acute stressor), combining nutritional, physical, and cognitive support, may be a helpful preventive measure for elderly patients with planned surgical interventions who are at risk for delirium [48]. Physical exercise has been shown in experimental studies in rats to be associated with reduced inflammation of the microglia [102], and several human studies have shown the benefit of early mobility in critically ill patients, with interventions ranging from range-of-motion exercises to ambulation, in decreasing delirium [103,104].

Pharmacologic Interventions

In general, pharmacologic measures should be considered only after nonpharmacologic strategies have failed. As expected, pharmacologic strategies for delirium management have been targeted based on hypotheses of delirium pathogenesis. Avoidance of opioids, benzodiazepines, and anticholinergics is a mainstay of delirium management [105]. While adequate pain control is paramount to delirium prevention, opioid-sparing techniques using multimodal analgesia pain regimens and regional anesthesia techniques should be employed to achieve adequate pain control while minimizing opioid use. Of the opioid options, meperidine has been shown to have a higher risk of delirium. A systematic review of postoperative analgesia in elderly patients found meperidine use to consistently be associated with higher rates of delirium [106]. The increased risk of delirium with meperidine use in the elderly is presumably related to its anticholinergic effects and the serotonergic effects of its metabolite normeperidine.

Benzodiazepines, due to their GABAergic effect, have been shown in multiple studies to contribute to delirium in the acute setting, including multiple critically ill and operative cohorts [107–110]. The use of benzodiazepines for sedation has also been shown in large randomized, controlled trials to increase the risk of delirium. A study by Pandharipande et al. compared sedation with dexmedetomidine with lorazepam infusion in mechanically ventilated patients and found that patients receiving dexmedetomidine had 4 more delirium-free days than the lorazepam group and 30 percent less coma [60]. Subsequently, the increased risk of delirium with benzodiazepine infusion in critically ill patients was demonstrated by Riker et al. in a randomized, controlled trial comparing dexmedetomidine with midazolam, with a 20 percent reduction in the number patients developing delirium with the administration of dexmedetomidine [62]. Avoidance of benzodiazepines is therefore prudent when feasible.

Dexmedetomidine has also been shown to reduce delirium incidence and duration, ICU length of stay, and cost when compared with propofol when used in postoperative cardiac surgery patients [111]. Dexmedetomidine is widely used as an adjunct to antipsychotics in the management of delirium; the cost-benefit ratio of dexmedetomidine, however, has been questioned. Recently, dexmedetomidine has been shown to be a beneficial and cost-effective rescue drug compared with haloperidol in nonintubated critically ill patients [112]. Addition of dexmedetomidine in patients with agitated delirium on the ventilator reduced mechanical ventilation hours at 7 days in a nonelderly population [113]. A few recent trials have shown low doses of dexmedetomidine to prevent delirium compared with either placebo or propofol in subsets of postoperative patients [111,114,115]. The mechanism of dexmedetomidine beyond avoidance of the GABA receptor is unclear. It has been shown to suppress the serum inflammatory mediators IL-6, IL-8, and TNF-α, but further investigations are warranted. Recently, however, it has been recommended by the European Society of Anaesthesiology as an intraoperative adjunct for prevention of delirium in high-risk elderly patients [116].

Despite antipsychotics being a mainstay for delirium management in clinical practice, their administration in elderly patients has been controversial. In previous studies, haloperidol had been shown to worsen outcomes in elderly patients with delirium, specifically mortality, whereas other studies have shown no difference when compared with quetiapine for the treatment of delirium [117]. A more recent meta-analysis of 17 randomized, controlled trials assessing the mortality of conventional antipsychotics, and haloperidol in particular, did not show an increase in mortality risk in elderly patients [118]. Haloperidol has been investigated recently as a pharmacologic delirium prevention method in the perioperative setting and in critically ill, mechanically ventilated patients. Perioperative haloperidol prophylaxis in elderly hip surgery patients did not decrease the incidence of delirium but was associated with a shorter duration of delirium [119]. A low-dose haloperidol bolus and subsequent infusion in elderly critical care patients who underwent noncardiac surgery decreased the incidence of delirium only after intra-abdominal surgeries [120]. A pre-post study of patients deemed at high risk for delirium found haloperidol prophylaxis to be associated with increased days alive and free of delirium and coma [121]. A randomized, double-blind, placebo-controlled trial conducted in medical and surgical ICUs, however, found that haloperidol reduced the hours patients spent with agitation but did not show a reduction in delirium incidence, delirium duration, days on mechanical ventilation, or ICU mortality [122].

With regard to delirium treatment, data on the efficacy of antipsychotic administration are also limited. In a study of critically ill patients at risk for delirium, placebo versus haloperidol versus ziprasidone for delirium treatment demonstrated no difference between the three groups in days alive and freedom from delirium [123]. A study of haloperidol versus olanzapine for delirium in the ICU found no difference in delirium duration, but patients who received haloperidol had more side effects [124]. Quetiapine in addition to haloperidol resulted in a faster resolution of delirium compared with placebo in addition to haloperidol [125].

Although cholinergic depletion is known to play a role in delirium development, acetylcholinesterase inhibitors such as rivastigmine and donezepil, which have been a mainstay in the treatment of Alzheimer’s disease, have all had disappointing results in the prevention of delirium. They have not been established as an effective therapy in delirium management and actually may increase mortality [126–129].

Given the role of neuroinflammation in the pathogenesis of delirium, strategies targeting reduction of the inflammatory cascade seem prudent, but there are currently no studies showing this to be an effective measure. Prophylactic administration of dexamethasone to decrease systemic inflammatory response and neuroinflammation in patients undergoing cardiopulmonary bypass had no effect on either the incidence or the duration of delirium [130]. The pleiotropic anti-inflammatory effects of statin therapy have been hypothesized to play a role in the management of delirium. Continuation of ongoing statin therapy in critically ill patients has been associated with a decreased risk of delirium, whereas withholding statins in chronic users may increase the odds of developing delirium [131–133].

With regard to the role of anesthesia in the development of POD, studies monitoring the electroencephalogram (EEG) during general anesthesia have shown that current recommended doses of anesthetics for elderly patients may be placing them in a profound state of brain inactivation known as burst suppression [134,135]. Titration of anesthetic doses based on real-time EEG monitoring may help to reduce the incidence of postoperative cognitive disorders [134] and has been associated with reduced POD likely from reduced oversedation [136]. Monitoring of anesthetic depth and avoidance of deep anesthesia in the elderly are considered part of delirium prevention strategies [96].