An update on postoperative delirium: Clinical features, neuropathogenesis, and perioperative management

Abstract

Purpose of Review

We present a focused review on postoperative delirium for anesthesiologists, encompassing clinical features, neuropathogenesis, and clinical identification and management strategies based on risk factors and current delirium treatments.

Recent Findings

The literature on postoperative delirium is dominated by non-experimental studies. We review delirium phenotypes, diagnostic criteria, and present standard nomenclature based on current literature. Disruption of cortical integration of complex information (CICI) may provide a framework to understand the neuropathogenesis of postoperative delirium, as well as risk factors and clinical modifiers in the perioperative period. We further divide risk factors into patient factors, surgical factors, and medical/pharmacological factors, and present specific considerations for each in the preoperative, intraoperative, and postoperative periods.

Summary

Postoperative delirium is prevalent, poorly understood, and often missed with current screening techniques. Proper identification of risk factors is useful for perioperative interventions and can help tailor patient-specific management strategies.

INTRODUCTION

The development of postoperative delirium has significant consequences for patient outcome and cost to the healthcare system. The incidence of postoperative delirium has been found to be anywhere between 5–70% depending on the population studied with higher incidences in older patients, patients with preexisting cognitive impairment, after cardiac surgery, and after emergency surgery, especially surgery for hip fracture [1–4]. Delirium can also lead to increased risk of morbidity and mortality and loss of independence [5]. For example, delirium can accelerate long-term cognitive decline in patients, especially with pre-existing Alzheimer’s disease [6]. Healthcare costs are also increased with delirium due to increased hospital length of stay and increased likelihood of institutionalization after hospital discharge [5, 7].

Since delirium imposes such a burden on patients and the healthcare system, there is great interest in studying what may predispose patients to delirium or if any precautions may be taken to prevent the occurrence of delirium during the perioperative period. Anesthesiologists will play a unique role as the risk factors are better understood; they may be able to intervene early in the preoperative phase, optimizing patients and potentially mitigating the development of postoperative delirium.

Unfortunately, barriers in research, interpretation, and clinical translation limit our current understanding of delirium in the perioperative period and hinder the anesthesiologist’s ability to apply focused interventions to affected patients. For example, postoperative delirium has a complex pathogenesis with myriad risk factors; to this end, studies are either designed to evaluate the contribution of a single risk factor to the pathogenesis of delirium [8], or propose diverse hypotheses to explain the contribution of multiple risk factors to its pathogenesis [9]. Indeed, there is currently no unifying hypothesis to understand the complexity and phenotypic diversity of delirium. Moreover, the nomenclature regarding postoperative delirium is not standardized, resulting in seemingly contradictory information and difficulty interpreting this information in a clinical context [10]. Lastly, the clinical translation of diagnosis or treatment strategies is limited by our current experimental paradigms: 1) No animal models of delirium exist, resulting in a lack of biomarkers for delirium and more theoretical hypotheses, 2) few non-invasive modalities exist with which to test mechanisms of delirium in humans, and 3) it is difficult to neurologically evaluate patients recovering from surgery. Thus, there is large heterogeneity in standard clinical assessments of postoperative delirium and efficacy of treatments or prevention strategies [11, 12], with no single metric to identify and track postoperative delirium [13, 14].

Here we present a focused review of the current literature on postoperative delirium as it pertains to anesthesiologists. Our objectives are to provide the anesthesiologist the tools needed to identify, manage, and prevent delirium throughout the perioperative period. We first present a brief and general background on delirium, complete with updated clinical definitions and standardized nomenclature. We then present a unifying framework to understand postoperative delirium, addressing several hypothesized mechanisms and clinical modifiers. Lastly, we offer management strategies for patients with or at risk for postoperative delirium based on our findings from the literature.

BACKGROUND

Delirium is a syndrome of acute cognitive failure characterized by the presence of defined clinical features, although the expression of these features can be diverse. The diagnostic criteria as defined by the fifth edition of the Diagnostic and Statistical Manual of Mental Disorders (DSM-5) [15, 16] are an acute onset and fluctuating course of impaired attention, reduced awareness and orientation to the environment, and disturbance in cognition — which may variably include changes in perception, memory, language, coherent reasoning, and visuospatial processing. The DSM-5 criteria have thus modified the classification of delirium (previously defined in DSM-4 as primarily a disturbance of consciousness), as “consciousness” is difficult to assess clinically [17]. Disturbances of the sleep-wake cycle and emotional regulation are also typical. Psychomotor dysfunction is a prominent feature that defines the motoric subtypes of delirium: a hyperactive subtype marked by agitation, a hypoactive subtype marked by lethargy and decreased motor activity, and a mixed subtype characterized by fluctuating features of both [18]. The vast majority of delirium is hypoactive or mixed, with pure hyperactive delirium being relatively uncommon, and extremely rare in elderly patients [19]. In clinical settings where screening instruments are not used systematically, the diagnosis is missed in ~60–80% of presentations [20, 21]; this is especially true in the post-anesthesia care unit (PACU) where patients may exhibit lethargy and decreased motor activity simply in recovering from anesthesia. Hypoactive delirium, increased age (> 70 years), and a failure to assess the acuity of mental status changes represent the strongest independent risk factors for missed diagnosis. Detailed clinical assessment is usually able to differentiate delirium from primary psychiatric illness (especially agitated depression), dementia, focal neurological syndromes, and nonconvulsive seizure disorders; electroencephalography, neuroimaging, and lumbar puncture rarely aid in diagnosis, and should be reserved for patients with atypical neurological findings or in whom no underlying cause can be established [22, 23].

Postoperative delirium can be classified as a subset of delirium that is distinct from “emergence delirium”, a misnomer in the literature better described as emergence agitation (Figure 1). The term postoperative delirium has been used to describe delirium from all causes occurring in patients receiving general anesthesia or sedation, with arbitrary time courses ranging from postoperative day 0–1 to 5–30 days postoperatively [11, 8, 24]. Within this classification, delirium can be further described by its clinical setting, such as intensive care unit (ICU) delirium [9] or PACU delirium [11]. In contrast, the term emergence delirium has been used to describe an agitated state upon emergence from anesthesia [13, 12, 25]. Eckenhoff [26] first used the term emergence hyperexcitation in 1961 to describe agitation in children upon emergence from anesthesia following ether, cyclopropane, and ketamine. Since then, the terms emergence agitation and emergence delirium have been used interchangeably [27]. Unlike postoperative delirium, “emergence delirium” occurs during emergence (i.e., with no lucid interval between the anesthetized state and “delirium”), and typically has a short (< 30 minute) and largely self-limited time course. Agitation during emergence can be treated with sedatives and analgesics, and is usually not associated with permanent after effects [13, 12, 27]. The literature is especially confusing because many studies on “emergence delirium” use inclusion criteria that are actually consistent with PACU delirium [28, 29]. Because of these differences, we propose the term emergence agitation to describe this condition and do not further address it in our discussion.

Figure 1. Classification of delirium subtypes.

Postoperative delirium is a subtype of delirium that occurs between postoperative days 0–5. PACU delirium is a further subtype of postoperative delirium that occurs in the PACU. ICU delirium is defined by its identification in the ICU; there may be some overlap depending on when patients are admitted to the ICU. Emergence agitation is seen on emergence from anesthesia and has unique etiologies and treatments. Abbreviations: PACU - post-anesthesia care unit; ICU - intensive care unit.

UNIFYING FRAMEWORK FOR POSTOPERATIVE DELIRIUM

Although there are many hypotheses for the pathogenesis of delirium, the clinical features of delirium can be viewed as a disruption of normal cortical integration of complex information (CICI). In this framework, attention and awareness are made possible by 1) complexity of neural information, defined by the level of global neural activity, and 2) appropriate integration of this information, defined by functional connectivity of brain regions at rest. Originally viewed as a systemically driven dysregulation of neuronal activity [30], it has been hypothesized that delirium becomes manifest when functional connectivity within the brain breaks down [31]. Indeed, in electroencephalographic (EEG) studies, postoperative delirium is associated with a reduction in neural complexity as evidenced by a shift to lower frequency activity [32–34]. Moreover, there is reduced global cortical connectivity and disruption of posterior to anterior information flow [33]. Functional magnetic resonance imaging (fMRI) studies of delirious patients show a loss of functional connectivity between the intralaminar thalamic nuclei and multiple subcortical regions [35]; conversely, patients exhibiting shorter and milder episodes of delirium have relatively increased (less impaired) cortical functional connectivity between the dorsolateral prefrontal cortex and the posterior cingulate [35].

Recent evidence via multiple modalities suggests that disruption of CICI may be an underlying feature of altered conscious states in general. For example, EEG patterns in anesthesia and non-REM sleep are consistent with preserved local neural processing, but with impaired integration across thalamic and cortical areas [36–38] and impaired frontoparietal information transfer [39]. Using EEG, fMRI, and positron emission tomography (PET), Hyder et al. [40] showed that the brain undergoes strong neurometabolic coupling, and both metabolism and neural complexity are linearly reduced in anesthesia, non-REM sleep, and the vegetative state. Fractional anisotropy measurements via diffusion tensor imaging (DTI) in the vegetative state are also consistent with reduced functional connectivity in resting state networks as well as between the thalamus and the posterior cingulate cortex [41], which may be manifest as impaired top-down processing seen in the vegetative state [42]. Interestingly, a return of functional connectivity between the thalamus and the posterior cingulate cortex is a marker of recovery from the vegetative state [43]. Altered functional connectivity has also been demonstrated in other disorders of consciousness and attention including normal aging [44], attention deficit-hyperactive disorder [45], traumatic brain injury [46], and post-traumatic confusional state [47].

By using disruption of CICI as a framework for delirium and altered conscious states, it is possible to understand the contribution of multiple mechanisms about the neuropathogenesis of delirium. There are many theories for the pathogenesis of delirium including network disconnectivity, oxidative stress, neuroinflammation, and neuroendocrine abnormalities (see Maldonado [48] for a detailed and comprehensive review). In the network disconnectivity hypothesis, a person’s baseline functional connectivity may serve as a risk factor for development of delirium, which is further modulated by states that increase inhibitory tone in the brain [31]. Thus, network disconnectivity has clear implications for disruption of CICI as outlined above. The oxidative stress hypothesis states that imbalances in oxygen supply and consumption (most often due to hypoperfusion) lead to oxidative stress and reactive oxygen species, ultimately causing cerebral oxidative damage [48]. Additionally, the failure of the mitochondrial ATPase results in the inability of neurons to maintain ionic gradients, which in turn causes calcium influx. The increased calcium influx further causes release of glutamate, modulating cortical excitability and disrupting CICI. It is worth noting that the posterior cingulate and precuneus have some of the highest metabolic rates in the adult cortex [49] and are functionally connected to resting state networks needed for attention and orientation [50], suggesting that oxidative stress may disrupt CICI through these cortical areas. Accordingly, intraoperative cerebral oxygen desaturation is an independent risk factor for developing postoperative delirium in patients undergoing cardiac or abdominal surgery [51, 52]. The neuroinflammatory hypothesis posits that neuroinflammatory changes cause disruption in the blood-brain barrier (BBB); the loss of BBB integrity further leads to changes in neural excitability and synaptic transmission. Moreover, the increased BBB permeability causes interstitial edema, which can further decrease cerebral perfusion and contribute to oxidative stress [48]. Lastly, the neuroendocrine hypothesis states that glucocorticoid elevations prevent neuronal glucose uptake and render neurons metabolically vulnerable and prone to dysregulation (cf. oxidative stress hypothesis) [53]. Under this hypothesis, the fact that the hippocampus contains the highest concentration of glucocorticoid receptors [54] can help explain the memory disturbances seen in delirium as well as the predisposition of these patients to long-term cognitive decline.

Disruption of CICI also provides a framework for understanding the relationship of various patient and surgical factors as modifiers of postoperative delirium. For example, there are many neural derangements in aging that predispose elderly populations for developing postoperative delirium, including reduced cortical volume [55], lowered baseline neural activity, decreased cortical functional connectivity [56], decreased acetylcholine levels, and decreased cerebral oxidative metabolism [48]. Similarly, the neuronal frailty associated with several pathologic states (Alzheimer’s disease, Parkinson’s disease, traumatic brain injury) can all contribute to a change in baseline network connectivity, rendering these populations vulnerable to postoperative delirium [31]. Patients with anxiety and depression are at higher risk for developing postoperative delirium [57]; this may be in part due to derangements of the hypothalamic-pituitary-adrenal (HPA) axis and increased circulating glucocorticoids [58]. Additionally, surgical and anesthetic factors may predispose a patient to delirium postoperatively. Cerebral hypoperfusion is common and can be caused by volatile anesthetics, hypovolemia, blood loss, and surgical clamping. Similarly, surgery and anesthetics can disrupt the BBB, causing oxidative damage via neuroinflammation [59]. The perioperative experience is associated with disruptions of the HPA axis in myriad ways: glucocorticoids can be endogenously increased by preoperative anxiety and stress, as well as by surgery itself [60], or by exogenous administration of glucocorticoids commonly administered for postoperative nausea or airway edema [61, 62].

PERIOPERATIVE MANAGEMENT STRATEGIES FOR POSTOPERATIVE DELIRIUM

In this section, we present a variety of strategies for identifying modifiers of delirium and managing and/or preventing postoperative delirium throughout the perioperative period. See Table 1 for a summary of discussed strategies.

Table 1.

Identification and management strategies for mitigating delirium throughout the perioperative period.

		Preoperative period	Intraoperative period	Postoperative period
Patient Factors	Identification Of Delirium Modifiers	Anesthesia consultation/interview for: Age, co-morbidities, labs Screening for cognitive frailty Psychiatric history Post-anesthesia history Sleep history	Low alpha power on EEG Delayed emergence	Vigilant screening (CAM ICU, Nu-DESC, others)
	Recommendations For Delirium Management and Prevention	Cognitive screening during anesthesia consultation Enrollment in clinical pathways / ERAS initiatives Potential involvement of other consultants Cardiologists Geriatricians / geriatric psychiatrists Palliative Care Physical Therapists	Optimize cerebral hemodynamics	Restore functional aids (glasses, hearing aids) Re-orientation Reunite family members early
Surgical Factors	Identification Of Delirium Modifiers	Surgery type and expected outcomes Anticipated blood loss Anticipated ICU stay Anticipated risk of confusion/pain	Head and neck surgery Awareness of high stimulation procedures, treatment with pre-emptive analgesia	Consider time of day, recent sleep patterns/habits
	Recommendations For Delirium Management and Prevention	Discuss likelihood of post operative delirium with patient and family members Discuss likelihood of success	Avoid burst suppression Aim for sleep-like EEG patterns	Efficient removal of chest tubes, oral/nasal tubes, masks
Medical / Pharmacologic Factors	Identification Of Delirium Modifiers	Optimization of medical therapy / substance use Assessment of polypharmacy Consider alternative pain management strategies	Hypotension, hemodynamic variability Low anesthetic requirement Slow circulation time	Review home medications
	Recommendations For Delirium Management and Prevention	Avoid benzodiazepine premedication in high-risk patients Avoid anti-cholinergic medications Consider regional techniques for pain control	Use short acting medications Consider alternative anesthetic plans (e.g., regional, TIVA)	Avoid benzodiazepines for agitation Consider Haldol for agitation consider caffeine and other medications for hypoactive delirium

Abbreviations: EEG – electroencephalography; CAM-ICU – confusion assessment method for the intensive care unit; Nu-DESC – nursing delirium screen; ERAS – enhanced recovery after surgery; TIVA – total intravenous anesthesia.

Preoperative management

Because the exact mechanisms underlying delirium are unknown, our understanding of predisposing risk factors is largely based on association studies; such identified risk factors are heterogeneous depending on the surgical procedure and patient population. However, the most commonly identified risk factor across studies is increasing age, especially over age 70 [1, 5]. Other recurring risk factors include pre-existing cognitive impairment, preoperative visual and hearing impairment, and preoperative use of benzodiazepines, opioids, and alcohol [63, 64, 5]. Of note, prolonged preoperative fluid fasting and longer-acting intraoperative opioids have also been identified as modifiable risk factors for early postoperative delirium [65].

Since there is no single treatment known for postoperative delirium, much attention is directed towards prevention of delirium in populations at risk. Screening for risk factors of postoperative delirium is the first step in prevention. Expert opinion favors preoperative screening for cognitive impairment to assess delirium risk [66]. One such method, the Mini-Cog, is a quick screen that can be performed in less than five minutes with high specificity [67]; the results may also help guide patients’ informed consent [68], as patients with unrecognized cognitive impairment and a positive Mini-Cog are more likely to develop postoperative delirium [69]. Similar results may be obtained by a simple clock drawing test alone [70, 71], and the combination of the two tests yields high sensitivity for detecting cognitive impairment [67]. Further, reviewing a patient’s non-invasive imaging if available (such as prior EEG or MRI) may be helpful in assessing their neurological susceptibility to postoperative delirium; specifically, markers of cortical atrophy such as reduced hippocampal thickness and reduced EEG power (particularly in the alpha band) are associated with cognitive impairment [72, 73]. In addition to formal screening for cognitive impairment, screening for sleep disturbances via a sleep history may also be helpful as sleep disorders are common in patients with cognitive impairment [74]. Recently, links between sleep-disordered breathing and postoperative delirium have been suggested, and it is recommended to screen for sleep-disordered breathing with tools such as the STOP-BANG questionnaire to aid in prevention and even treatment of postoperative delirium with continuous positive airway pressure [75]. Another area of emerging interest is the association of frailty with the development of delirium [76, 77]. Frail patients are more likely to have postoperative delirium and therefore screening for frailty would help target prevention strategies to another population at risk for delirium. Preoperative pain and depressive symptoms, both independently and in combination, have been shown to increase the risk of postoperative delirium and screening for both has been recommended [78, 79].

Although formal guidelines are not available to direct which preoperative medications should be used or avoided to prevent postoperative delirium, various studies have found certain medications and polypharmacy may be associated with delirium [80]. As mentioned earlier, preoperative use of benzodiazepines and opioid medication is associated with an increased risk of postoperative delirium. Similarly, anticholinergic medications such as scopolamine can exacerbate postoperative delirium and should be avoided in the perioperative period for at risk patients [81]. Other preoperative medications have not demonstrated consistent effects on modulating postoperative delirium. For example, preoperative beta-blocker administration was associated with postoperative delirium in patients undergoing vascular surgery [82], but not after cardiac surgery [83]. Similarly, preoperative use of statins has been shown to reduce the incidence of postoperative delirium in one study [84] but showed no difference in another [85]. Knowing that certain preoperative medications may contribute to postoperative delirium and that medication review by pharmacists resulted in fewer delirious days for hospitalized patients [86], the effect of preoperative medication review should be an area of future study.

Taken together, patients at high risk for developing postoperative delirium may benefit from a thorough and comprehensive pre-anesthesia consultation, including a detailed cognitive, psychiatric, sleep, and anesthetic history with a review of medications and substance use. Other information such as previous neuroimaging and cognitive screening may be useful and, with proper knowledge of the surgical procedure and expected outcomes, may help provide a tailored anesthetic plan. It may also be helpful to have a discussion with high-risk patients and their families about expected postoperative outcomes, possibly with the help of other consultants (e.g., cardiology, palliative care, geriatrics) if needed. Premedication should be carefully considered in these patients, and they may benefit from a balanced (i.e., minimizing opioids) pain management strategy.

Intraoperative management

Despite the lack of a single mechanism underlying postoperative delirium, intraoperative decision making may influence a patient’s postoperative cognitive course. Although patient characteristics such as age are not modifiable, adjusting the dose of delivered anesthetic to avoid burst suppression seems a prudent strategy for anesthetic maintenance [87]. As susceptibility to burst suppression is associated with increasing age [88] it is reasonable to intraoperatively monitor the neurophysiology of all patients suspected to be at increased risk for delirium. Detection of burst suppression intraoperatively would require the use of EEG or processed EEG devices and possibly recognition of simple EEG patterns. For many processed EEG devices the appearance of an index lower than the recommended range is associated with burst suppression [89, 90]. Such training in waveform interpretation has shown to be feasible for anesthesia providers [91]. Similarly, low alpha power on EEG may be associated with postoperative delirium, and it may be possible to modify EEG signals with pharmacologic choices [92]. While it is unknown if delayed emergence or emergence agitation is associated with postoperative delirium, it stands to reason that careful titration of analgesic medications to prevent over-sedation or high sympathetic activity from pain on emergence may be beneficial in mitigating all three phenomena.

Pre-existing patient factors such as hypertension and diabetes are not easily modifiable by the anesthesiologist, but skill in minimizing the contribution of these diseases to cerebral hemodynamics might be expected to decrease the risk of developing postoperative delirium. For example, patients who have recently had head and neck surgery [93] or procedures with large amounts of noxious stimulation (e.g., thoracotomy [94]) may be at increased risk for postoperative delirium; in these populations, preemptive analgesia, and regional techniques are likely to confer benefits. Other strategies have been suggested as potential improvements for managing a patient at high risk for postoperative delirium, including the use of short-acting medications [95], alternative anesthetic plans (e.g., regional [96], TIVA with dexmedetomidine [97, 98]), and pre-emptive anti-psychotic medication [99]. Unfortunately, none of these can be universally applied to patients at risk for delirium because the etiology of delirium is so variable.

Postoperative management

The sensitivity for clinical detection of delirium without the use of targeted assessment tools is poor (as low as ~20% in specific settings) [20, 21], emphasizing the need for a systematic approach to screening and diagnosis in the postoperative period. More than 20 instruments have been developed to detect delirium [100], although only a selection have been rigorously validated and are in non-research use [101]. The most widely used and studied is the Confusion Assessment Method (CAM) [102], which assesses nine features of delirium and can be administered by a non-specialist clinician in five minutes. The predictive value and reliability of the CAM is highest only when the assessor has been appropriately trained in its use. A version of the CAM has been developed and validated for use in the intensive care setting (CAM-ICU) [103], while another adaptation, the Family Confusion Assessment method (FAM-CAM), has been validated for reporting by family and caregivers [104]. The Intensive Care Delirium Screening Checklist (ICDSC) [105] is yet another screening instrument for the intensive care setting, although its predictive value may be lower than the CAM-ICU [106]. The Delirium Observation Screening Scale (DOSS) [107] and Nursing Delirium Screening Scale (Nu-DESC) [108] are rapid (5 and 1 minute) screening tools that can be administered by nursing staff, and may be superior to the older NEECHAM Confusion Scale [109]. The Memorial Delirium Assessment Scale (MDAS) [110] is well-validated for assessing the severity of delirium across serial measurements, but takes 10 minutes to administer and best reserved for tracking changes once the diagnosis is established. Other tests such as the Delirium Rating Scale-Revised-98 (DRS-R-98) [111] and the Global Attentiveness Rating (GAR) [112] have excellent predictive value, but require higher levels of training. Lastly, the Mini-Mental State Examination (MMSE) [113] assesses cognitive deficits and is sensitive in detecting delirium, but has limited predictive value because of its poor specificity [114]. A summary of the predicative values of a number of instruments is presented in Table 2.

Table 2.

Common assessment tools for the diagnosis of delirium.

Scale		Likelihood Ratio		Comments
		Positive	Negative
Confusion Assessment Method	CAM[102]	9.6	0.16	Requires some training
Confusion Assessment Method-ICU	CAM-ICU[103]	20[126]	0.21	Rapid ICU screen
Delirium Observation Screening Scale	DOSS[107]	5.2	0.10	Rapid nursing screen
Delirium Rating Scale-Revised-98	DRS-R-98[111]	8.0	0.08	Requires training
Family Confusion Assessment Method	FAM-CAM[104]	44[104]	0.12	Family and caregiver questions
Global Attentiveness Rating	GAR[112]	65	0.06	Physician assessor with training
Intensive Care Delirium Screening Checklist	ICDSC[105]	4.1[126]	0.32	Rapid ICU screen
Memorial Delirium Assessment Scale	MDAS[110]	12	0.09	Assesses changes in severity
Mini-Mental State Examination	MMSE[113]	1.6	0.12	Limited specificity
NEECHAM Confusion Scale	NEECHAM [109]	17[127]	0.14	Rapid ICU screen
Nursing Delirium Screening Scale	Nu-DESC[108]	3.1	0.06	Rapid nursing screen

Likelihood ratios combine sensitivity and specificity values to assess the likelihood of delirium given a positive or negative test result, and do not change with pretest probability. Positive likelihood ratio is sensitivity/(1 – specificity); negative likelihood ratio is (1 – sensitivity)/specificity. Ratios are taken from the report of pooled values by Wong et al[101] unless specifically noted.

The appearance of delirium may signal the initial presentation of an acute and potentially life-threatening emergency. The initial management should consider the possibility of critical causes, such as hypoxia, hypercapnia, hypo- or hyperglycemia, pulmonary embolus, blood loss, arrhythmia, or the emergence of infection. Airway protection and hemodynamic support should be established if required. Laboratory tests can assess for electrolyte derangements, and the possibility of drug or alcohol withdrawal should be considered. Further diagnostic evaluation is targeted, and may variably include neurological or other imaging, electrocardiography, arterial blood gases, and further assessment for metabolic and endocrine abnormalities. The evaluation for reversible etiologies should also include a detailed assessment of all medications and possible interactions, with particular attention to reducing or removing benzodiazepines and other sedatives, opiates, and anticholinergics.

There is no clear evidence to support any singular intervention in the treatment of newly established delirium. The initial approach is the multidisciplinary institution of nonpharmacological strategies supported by evidence of effectiveness in primary prevention trials [115, 116]. This includes reorientation, family presence and interaction, minimizing disturbance in the sleep-wake cycle, addressing sensory impairment with hearing aids and eyeglasses, noise reduction, and mobilization. A 2016 Cochrane Review [116] failed to show supporting evidence for any pharmacological agent in the prevention or treatment of delirium. Small studies provide some support for the efficacy of melatonin [117], the melatonin agonist ramelteon [118], and the atypical antipsychotics quetiapine [119], olanzapine [120], and respiridone [121] — but larger randomized controlled trials are required. The efficacy of cholinesterase inhibitors (e.g. donepezil) and 5-hydroxytryptamine receptor antagonists (e.g. trazodone) are not supported. The most widely used pharmacological treatment is the neuroleptic haloperidol, which may be an appropriate intervention in hyperactive delirium to ensure patient safety and minimize distress; however, haloperidol (and the atypical antipsychotics) may mostly be masking the symptoms of delirium and have no clear benefit on the duration or clinical outcomes. A number of larger randomized controlled trials have addressed the use of dexmedetomidine for sedation in the intensive care setting [122–125], with encouraging but inconsistent demonstration of effectiveness in the reducing the incidence and duration of delirium in this setting. However, there is insufficient evidence to support its use as a treatment response to delirium in other settings.

CONCLUSION

In summary, postoperative delirium is common, especially in patients with pre-existing cognitive impairment and those undergoing major surgery. Vigilant identification and proper management of these patients is critical, as delirium can lead to increased morbidity, mortality, loss of independence, as well as increased hospital stay and increased total healthcare costs. Unfortunately, relatively little is known about postoperative delirium because of its phenotypic heterogeneity and heterogeneity of screening tools, often resulting in underdiagnosis. Recent research suggests that patients with delirium exhibit a disruption of cortical integration of complex information (CICI), a feature that is common to many disorders of consciousness. Anesthesiologists are in a unique position to recognize high-risk patients preoperatively and may be able to prevent or mitigate postoperative delirium with risk identification strategies preoperatively, tailored anesthetic plans, and targeted postoperative interventions. Because postoperative delirium is an emerging area of research, the knowledge gained from ongoing and future studies will significantly impact the landscape of postoperative delirium and will continue to be useful to anesthesiologists for preventing and mitigating the consequences of postoperative delirium.