Delirium

Evolution of the Delirium Concept and Diagnostic Criteria

Disturbance of consciousness is a characteristic feature of delirium that distinguishes it from the classical form of Alzheimer disease. It has been the hardest feature to operationalize. Greiner's development of the concept “clouding of consciousness” in delirium can be understood in the context of the nineteenth-century view of impairments in consciousness represented as “organic” brain disease. “Clouding” has come to refer to both a state of lowered arousal between alertness and coma and a denseness of the medium in which psychic events occur. Geschwind's influential views, based mainly on studies of delirium in younger adults, paved the way for operationalization of delirium as an attentional disorder in modern classificatory systems. However, disturbance of attention is not the only cognitive disorder seen in, neither is it the sole preserve of, delirium. So how does the role of consciousness relate to delirium, and how may this assist physicians with diagnosis?

One way of describing consciousness is the state in which all waking thoughts are marshaled and temporally organized and expressed. There is no thought content that, in the study of delirium symptoms, has not in some way been disrupted. Unless there is a recess of consciousness beyond sentinel thought, then it follows that disturbance of consciousness must be axiomatic to a diagnosis of delirium. Consciousness embodies a duality of reflection, subserved through a parallel but binary functional network. Experientially, conscious thought is divided into reflection (default mode) and goal-related (task positive) states, respectively. Functional magnetic resonance imaging (MRI) data would suggest a reversible reduction in resting state connectivity in relation to delirium status to support this premise and reduced task induced deactivation of DMN regions. These perspectives on the relationship between consciousness and delirium provide a conceptual model for research in delirium. Temporally ordered awareness, and consciousness, is a macrostate property whose construct belies definition and prediction according to the component parts. In other words, the condition of consciousness is imbued by central nervous system (CNS) complexity but cannot easily be explained by it. At the event horizon, where consciousness breaks down and the status of delirium is conferred, all such higher function is perturbed with loss of temporal ordering of information being key. Reduced intercortical connectivity is attributed to loss of CNS hierarchy. Lower order, reflexive structures within the primitive brain and limbic system are then left to drive behavior.

Conventionally, attention was related also with consciousness level (arousal) in the International Classification of Diseases (ICD) and earlier iterations of the Diagnostic and Statistical Manual of Mental Disorders (DSM). Both attention and arousal are hierarchically related: it is possible to have full arousal but profound inattention (in hypervigilance), but the converse is not true. It is interesting that the fifth edition of the DSM (DSM-5) has abrogated consciousness in favor of its handmaiden, awareness. There are implicit threats that inclusion of awareness, displacement of low arousal states and consciousness from a diagnosis of delirium may bring. To redress this diagnostic schism, it is important to consider alternative ways in which consciousness, as it relates to delirium, can be operationalized. Arousal still remains an important construct in explaining the spectrum of behavior in delirium.

Clinical Features

Abnormal arousal ranges from the hyperalert-hyperactive state often accompanied by hallucinations, illusions, and delusions (so-called hyperactive delirium) to a state of somnolence and hypoalertness (so-called hypoactive delirium). Whereas the archetypal image of delirium is the hyperactive state with psychosis, older patients are more likely to experience mixed states (lying in bed or fluctuating between states of hyperalertness and somnolence). “Activated” features are overrepresented on screening and diagnostic criteria, leaving hypoactive delirium most vulnerable to underrecognition if placid behavior had not already succeeded in this. The subtype of delirium that is most easily missed may also be associated with the worst outcome. An approach to recognition is to systematizing psychomotor behavior or, more inclusively, arousal. Concentrating on the core features common to all subtypes is a practical solution not without merit. We can posit that inability to engage represents severe inattention and acute low-arousal states equate with inattention and cognitive disruption. These phenomena are as much a delirium that is indistinguishable from delirium diagnosed on the basis of cognitive testing and interview. Not only is this approach conceptually grounded, but it also offers a means of unifying the spectrum of divergent behavior as part of the syndrome.

Cases in which attention is testable offer a portal to consciousness and shine a light on the presence or absence of delirium. Staff involved in daily care of older people often have difficulty in formally assessing inattention and thus may not recognize delirium. Interviewing a patient over a 2-minute period can be a simple way to assess attention, and the Edinburgh Delirium Test Box (and smartphone app) has been developed for this purpose.

Altered arousal detected through use of tools such as the modified Richmond Agitation Sedation Scale (m-RASS) may be fairly specific to delirium, particularly when attention is difficult to quantify.

Time disorientation is commonly measured best by nursing staff in hospitals ; is a useful guide to the presence and severity of dementia or delirium in older hospital patients ; and, when present at admission, is a risk factor for delirium at follow-up.

What is the impact of these disruptions in conscious flow on an individual level? Many patients experience temporal dislocation (i.e., simultaneously encountered past and present) entwined with the misperception of imagination as reality. Commonly, patients recount visions of a frightening, morbid, or sinister nature. Within psychosis, hallucinations (visual or auditory) and delusions may be elicited and can be disturbing to patients and those around them. Delusions and hallucinations may evoke recall distress in those who remember their delirium, even after recovery.

Acuity of onset, usually over hours to days, in conjunction with fluctuation, is a core feature of, but not peculiar to, delirium. Triangulation of history (from patients, relatives, clinical staff)— to characterize the presentation, recognize the duration of onset, and probe for possible triggers—is mandatory.

Differential Diagnoses

Most patients who arrive at the hospital with acute dementia have, or will develop, delirium (32% to 89%). Because the clinical features may be similar, behavioral changes in an individual with dementia can mistakenly be normalized by the contribution of environmental change or perceived as a natural sequela of dementia (or indeed aging itself). Greater than expected (although what constitutes an “expected” amount is difficult to standardize) deterioration in cognition, along with new onset of hallucinations and delusions, is suggestive of a delirium superimposed on dementia (DSD) that has not been captured by rating scales. The core problem with DSD is that the reference standards for delirium and dementia propose the syndromes as being mutually exclusive. It is clear this is far from common experience. Few of the assessment tools for delirium even included dementia patients, although validation studies for more recently developed scales report the diagnostic accuracy in participants with and without dementia.

Variability in symptoms is also a characteristic of dementia with Lewy bodies (DLB), and the psychopathologic similarities of visual hallucinations and disrupted sleep cycle add another masquerade. Avoidance of typical psychotropic agents, in any patient with Parkinsonian features, helps prevent characteristic neuroleptic sensitivity associated with the disorder. Depression and delirium may share negative symptoms and even a broad range of cognitive deficits. The mood and behavioral changes observed as part of the delirium syndrome, particularly hypoactive delirium, may account for the tendency to misdiagnose the two conditions. Rapidity of onset, relative inattention, and impaired arousal in the acute hospital setting usually reliably guides clinicians to a diagnosis of delirium. Also not to be ignored are the less frequent but potent causes of delirium that arise from primary CNS disorder that can be classified broadly into lesions or infection, particularly when the clinical picture is associated with focal neurologic signs, fever, or both. Acute cognitive decline may be observed also in patients with strategic infarct dementia.

It is worth noting specialized investigation for those disorders where overlap or cause by a CNS disorder is suspected. The electroencephalogram (EEG) offers an adjunct to clinical assessment. Well described and replicated, the pattern of “diffuse slowing” on EEG can help to distinguish a delirium from the following conditions: DSD, temporal lobe epilepsy, depression, and nonconvulsive status epilepticus. The EEG is useful in defining more peculiar CNS causes of delirium, such as limbic encephalitis and Creutzfeldt-Jacob disease. Although the EEG is still valid for use in an older population, a level of patient compliance is required that is not reliably achievable in patients in a delirious state. It therefore remains a diagnostic adjunct to the more challenging differential diagnoses outlined but does not replace the less invasive and more responsive focus of robust, bedside clinical and cognitive assessment.

Behavioral and psychological symptoms of dementia are also difficult to distinguish from DSD, compounded by the fact that neither has a reference-standard definition in the DSM or ICD. A pragmatic approach would maintain that acute change in cognitive, arousal, or perceptual symptoms in existing dementia should prompt the search for a general medical precipitant.

Prevalence

Delirium afflicts most acutely hospitalized, frail, older patients. In-hospital reported rates of delirium vary widely, typically 20% to 30%, and are a propensity of the magnitude of the frailty and illness severity burden. The emergency department mixes increasing presentations of older patients with acute illness and extrapolates to a prevalence of delirium of between 7% and 20%. Clinicians who are more removed from the distal consequences of delirium in the emergency department setting often fail to undertake cognitive assessment. Even when cognitive assessment is performed, it does not always register with a diagnosis of delirium or affect management. System pressures within an emergency department, through the prism of patient flow, do not readily enhance the landscape of care for patients with delirium. Overbearing management as a substitute for deft clinical problem solving does little to aid those “butterflies at the wheel” at risk of delirium. Optimization of care aligned to the needs of the older patient in the emergency department is not insurmountable; an education program with geriatrician support has been advocated as a simple way of improving care delivery, and recent work suggests delirium is finally emerging as a priority in the emergency department.

The orthogeriatric population is host to a 4% to 53% postoperative incidence of delirium following hip fracture. Similarly, the intensive care unit is a potent conflagration of illness severity as to be inevitably bound with delirium and a prevalence of 80% in older patients.

Delirium also transgresses beyond the acute sector, most commonly as a continuous episode through discharge from hospital. Even at conservative estimates, delirium and subsyndromal delirium still represent a significant and largely unexplored health burden in postacute facilities.

Many, if not most, individuals dying from chronic illness, regardless of location, experience confusion. An inability to separate those who die with delirium, as opposed to those who die from delirium, possibly conflates the bleak survival figures. As yet, nothing within the delirium syndrome offers the ability to distinguish a temporarily recoverable process from a “terminal drop.” However, in late stages of dementia, a phase of chronic delirium often may not resolve even after due diligence to cause, earning the epithet “delentia.”

Multicomponent intervention strategies have had little impact on outcomes in the palliative care setting, and contemporary studies are needed. An interdisciplinary and consensus-driven focus is advised with goals of treatment not adding to the burden of the dying process.

Experimental Models

It is convenient to consider two mechanisms for brain injury in the context of acute illness: direct and indirect. Direct brain insults include hypoxemia, stroke, trauma, or medications. In such conditions, brain dysfunction evidently arises as a direct consequence of the pathologic process.

Usually, acute pathology arises in the periphery (e.g., infection/inflammation, pain). A unifying idea is that aberrant stress responses have an impact on brain and brain function. There are a number of routes through which systemic processes in the periphery can have an effect on the brain. Inflammatory mediators can interact directly with neurons in areas where the blood-brain barrier is deficient, neurohumoral connections communicate directly through the vagus nerve, and endothelial glial cells can transmit cytokine signals into the brain parenchyma.

Neuroendocrine axes responsible for managing the normal stress response can become pathologically disrupted such that delirium is precipitated, sustained, or both. For example, glucocorticoid regulation through the hypothalamic-pituitary-adrenal (HPA) axis is vital, because sustained high levels may lead to chronic activation of low-affinity receptors, and this, in itself, is cytotoxic (e.g., Cushing disease). Reciprocally, it is known that chronic neurodegeneration in the limbic system leads to dysregulation of the HPA axis so the higher order control of the cortisol response can become exaggerated. Together, these situate neuroinflammatory and neuroendocrine mechanisms as “aberrant stress mediators.”

At a cellular level, it is understood that neurodegenerative conditions such as Alzheimer disease can initiate responses from microglia, the resident monocyte/macrophage system in CNS. Morphologically activated microglia can adopt a wide number of functional phenotypes, determined by a range of conditions. Crucially, microglial responses to neurodegeneration are on a spectrum from M1 (classical macrophage activity) to M2 (growth-repair functions). Thus, these immunologic phenotypes may be deleterious (enhancing neurodegeneration) or beneficial (clearing amyloid deposits). In animal models, microglia have been shown to migrate to new amyloid plaques. In vitro, microglial receptors (e.g., Toll-like receptor 4) can contribute to innate immunity through clearing amyloid plaques. Although the regulatory mechanisms are not completely understood, it appears that the predominant response to amyloid is not overly aggressive and, indeed, may be antiinflammatory in part; taken together, these result in microglial priming. Microglial priming is a key concept and represents a state whereby glia are morphologically activated but not proinflammatory. However, this primed state can result in phenotypic switching in response to an inflammatory challenge. One such murine model of delirium based upon a primed neurodegenerative state exhibits a delirium-like syndrome only in conjunction with an exaggerated proinflammatory cytokine response to peripheral inflammatory challenge. Furthermore, microglia express cyclooxygenase-1, and prostaglandin-mediated cognitive dysfunction can be blocked by nonsteroidal antiinflammatory drugs, suggesting therapeutic opportunities.