Delirium is a complex neuropsychiatric disorder that occurs commonly among patients in all health care settings, especially among the elderly and those with preexisting brain lesions or cognitive impairment. It is primarily characterized by generalized impairment of cognition, especially orientation and attention, but also involves a range of noncognitive symptoms, including motor behavior, sleep-wake cycle, thinking, language, perception, and affect. It characteristically has an acute onset (hours to days) and a fluctuating course (waxing and waning symptom severity over a 24-hour period), often worsening at night. It may be preceded by a prodromal phase of 2-3 days of malaise, restlessness, poor concentration, anxiety, irritability, sleep disturbances, and nightmares. It has been called acute organic brain syndrome and acute brain failure because of its breadth of cognitive and behavioral symptoms.
Delirium is an abnormal state of consciousness along a continuum between normal alertness and awareness at one extreme and the reduced wakefulness associated with stupor or coma at the other extreme. Consciousness has two main components — level of wakefulness (alertness) and content of higher mental functions (awareness). Because delirium alters both of these components of consciousness, it impairs the person more broadly than most other psychiatric disorders do. Precise delineation between severe hypoactive delirium and stupor can be difficult. Emergence from coma usually involves a period of delirium before normal consciousness is achieved. A prospective study of intensive care unit (ICU) patients found that 89% of survivors of stupor or coma progressed to delirium, whereas the small number who progressed directly to normal consciousness without delirium tended to have had drug-induced comatose states.
Although delirium is usually characterized by an acute onset replete with many symptoms, it may be preceded by a subclinical delirium with more insidious changes in sleep pattern or cognition. Matsushima et al. prospectively studied 10 critical care unit patients with delirium and 10 nondelirious control subjects with electroencephalography. They found prodromal changes of background slowing on the electroencephalogram (EEG) and sleep disturbance associated with changing consciousness. Cole et al. () prospectively studied “subsyndromal delirium,” defined as the presence of one or more of four core symptoms (clouding of consciousness, inattention, disorientation, and/or perceptual disturbances) but not meeting DSM-III-R (American Psychiatric Association 1987) criteria, in 164 elderly medical patients. The more symptoms present, especially on admission, the worse the prognosis, suggesting that even subclinical manifestations of delirium are significant.
Delirium also may be a transient state, as when a patient emerges from general anesthesia, during concussion following a head injury, or postictally. Football players who sustain a head injury during a game are removed to the sidelines until the disorienting effects of concussion resolve sufficiently. Most deliria are considered reversible, but in the terminally ill, delirium may be progressive and intractable despite measures to treat it. On the other hand, delirium occurring in patients with serious illness frequently resolves, as evidenced in the study by Breitbart et al. (), in which two-thirds of the cases of delirium occurring in hospitalized patients with cancer resolved completely with treatment. Similarly, Ljubisavljevic and Kelly found that 85% of the patients admitted with cancer who developed delirium experienced successful symptom reversal.
Delirium is considered a syndrome and not a unitary disorder because a wide variety of underlying etiologies can cause it. Identification of these etiologies, often multiple or occurring serially over time, is a key part of clinical management. Despite these varied etiologies and physiology, delirium symptoms are characteristic and thus may represent dysfunction of a final common neural pathway that includes perturbations of the various brain regions responsible for the abnormal cognitions, thinking, sleep, and behaviors.
Unlike most other psychiatric disorders, delirium symptoms typically fluctuate in intensity over any 24-hour period. Symptom fluctuation is measurable and is an important indicator of delirium emphasized in diagnostic classifications such as DSM-IV. During this characteristic waxing and waning of symptoms, relative lucid or quiescent periods often occur, which frustrate accurate diagnosis and complicate research severity ratings. In milder cases, such periods involve a significant diminution of delirium symptoms or even a seeming resolution of symptoms, but the latter has not been carefully studied. The underlying reason for this fluctuation in symptom severity is poorly understood — it may relate to shifts between hypoactive and hyperactive periods or fragmentations of the sleep-wake cycle, including daytime rapid eye movement (REM) sleep.
Although not nearly as well studied, the symptom profile of delirium in children appears to be similar to that in adults. In the only study of delirium phenomenology in children and adolescents in which a standardized instrument was used, Turkel et al. retrospectively described 84 consecutively evaluated delirium patients (age 6 months to 19 years) and found scores comparable to those in adults, with the only difference being fewer delusions and hallucinations in younger children. Turkel et al. also compared delirium symptoms across the life cycle and, despite differences in methodologies, considered them to be largely similar. Prugh et al. () noted the importance of educating nursing staff about the difference between visual hallucinations and imaginary friends. Documentation of all delirium symptoms in preverbal children or noncommunicative adults is difficult. In these patients, more reliance on inference and observation of changed or unusual behaviors — for example, inferring hallucinations or recording sleep-wake cycle changes — is needed.
Delirium symptoms in adults across the age range are comparable, although the co-occurrence of another cognitive mental disorder is particularly likely in the elderly compared with younger adults and is usually related to degenerative or vascular dementia. How the presence of a comorbid dementia alters the phenomenological presentation of delirium in the elderly is not well studied, but existing data suggest that delirium overshadows the dementia symptoms (). Likewise, diagnosing delirium in mentally retarded patients can be more challenging.
One of the challenges for both clinicians and delirium researchers is the myriad of terms applied to the delirious state. Historically, acute global cognitive disturbances have been labeled according to the setting in which they occurred or the apparent etiology for the confusional state, resulting in the myriad of synonyms that exist in practice and the literature.
TABLE Terms used to denote delirium
|Acute brain failure||Cerebral insufficiency||Organic brain syndrome|
|Acute brain syndrome||Confusional state||Posttraumatic amnesia|
|Acute brain syndrome with psychosis||Dysergastic reaction||Reversible cerebral dysfunction|
|Acute dementia||Encephalopathy||Reversible cognitive dysfunction|
|Acute organic psychosis||Exogenous psychosis||Reversible dementia|
|Acute organic reaction||Infective-exhaustive psychosis||Reversible toxic psychosis|
|Acute organic syndrome||Intensive care unit (ICU) psychosis||Toxic confusion state|
|Acute reversible psychosis||Metabolic encephalopathy||Toxic encephalopathy|
|Acute secondary psychosis||Oneiric state|
Little evidence supports these as separate entities, and, as such, delirium has been adopted as the accepted umbrella term to denote acute disturbances of global cognitive function as defined in both DSM-IV and ICD-10 (World Health Organization 1992) research classification systems. Even though the term delirium has been used since classical Greek medical writings, unfortunately, different terms continue to be used by nonpsychiatric physicians (e.g., ICU psychosis, hepatic encephalopathy, toxic psychosis, posttraumatic amnesia). These terms inappropriately suggest the existence of independent psychiatric disorders for each etiology rather than recognize delirium as a unitary syndrome. Terms such as acute brain failure and acute organic brain syndrome highlight the global nature and acute onset of cerebral cortical deficits in patients with delirium, but they lack specificity in regard to other cognitive mental disorders. The term delirium subsumes these many other terms, and its consistent use will enhance medical communication, diagnosis, and research.
Little work has been done with the use of daily delirium ratings to better understand the temporal course of this syndrome. In a study of 432 medical inpatients 65 years or older, Rudberg et al. found that 15% had delirium, and 69% of those had delirium for only a single day. Mean delirium scores on day 1 were significantly higher (i.e., worse) in those whose delirium occurred for multiple days compared with those whose delirium lasted for 1 day (25.4±3.6 vs. 22.6±4.4), suggesting a relation between severity and duration in delirium episodes.
Delirium continues to be understudied compared with other psychiatric disorders, as well as underrecognized and underdiagnosed. It is commonly misdiagnosed as depression by nonpsychiatrists. Misdiagnosis of delirium is more likely when delirium is hypoactive in presentation and when patients are referred from surgical or intensive care settings. Van Zyl and Davidson reviewed charts of 31 delirious patients who were referred for psychiatric consultation and received standardized de-
lirium assessments. They found that delirium or a synonym was noted in 55% of the structured discharge summaries and in none of the unstructured summaries, for an overall rate of 16%. It was more likely mentioned when it occurred in women, was more severe, or was the main reason for admission. Johnson et al. studied consecutive elderly patients admitted to a general hospital and found that delirium was explicitly documented in 5% of the patients and noted as a synonym in 18%, with a variable but poor recognition of individual delirium symptoms. The missed cases were denoted as dementia (25%), a functional psychiatric disorder (25%), or no diagnosis noted (50%).
Nondetection was associated with poorer outcome, including increased mortality, in a study of detection of delirium in emergency department patients. In contrast, explicit recognition of delirium was associated with better outcome in the form of shorter in-patient stays and lower mortality. Detection can be improved by providing formal educational programs, for example, with house staff. Personal attitudes are important among nursing staff, who often play a key role in identifying and reporting symptoms because the symptoms fluctuate, for example, at night. Detection is a challenge in ICU settings, where the sickest patients are at the highest risk for delirium. Ely et al. () distributed a survey to 912 physicians, nurses, respiratory therapists, and pharmacists attending international critical care meetings and found that 72% thought that ventilated patients experienced delirium, 92% considered delirium a very serious problem, and 78% acknowledged that it was under-diagnosed. Yet only 40% routinely screened for delirium, and only 16% used a specific tool for assessment. Rincon et al. reported that critical care unit staff underdiagnosed delirium (and other psychiatric disorders) and used psychotropic medications without any clear documentation.
ICU populations have delirium prevalence rates ranging from 40% to 87%. ICU delirium isunderstudied and neglected probably because it is “expected” to happen during severe illness, and medical resources are preferentially dedicated to managing the more immediate “life-threatening” problems.
Related to pressures to reduce acute hospital care costs, elderly patients are discharged, often to nursing homes, before delirium resolves. Kiely et al. studied 2,158 patients from seven Boston, Massachusetts, area skilled nursing facilities and found that 16% had a fullblown delirium. In general, such facilities are even less equipped with health care professionals to diagnose and manage delirium than are acute care settings.
Delirium can have a profound effect on a patient’s morbidity and mortality as well as on his or her caregivers and loved ones. Delirious patients have difficulty comprehending and communicating effectively, consenting to procedures, complying with medical management (e.g., removing intravenous lines, tubes, or catheters), benefiting from many therapies, and maintaining expected levels of self-hygiene and eating. They also are at risk for inadvertent self-harm because of confusion about the environment or in response to hallucinations or paranoid delusions. Delirium-recovered patients were uncomfortable discussing their delirium episodes — even to the extent of denial — because they feared that it meant that they were “senile” or “mad”. Breitbart et al. () prospectively interviewed and rated 101 cancer patients with a resolved delirium episode, their spouses, and their nurses (). About half (43%) of the patients recalled their episode, with recall dependent on delirium severity (100% of patients with mild delirium vs. 16% of patients with severe delirium recalled the episode). Mean distress levels were high for patients and nurses but were highest for spouses. However, among patients with delirium who did not recall the episode, the mean distress level was half that of those who did recall. The experience of the delirium was frightening and stressful for all involved, but for somewhat different reasons — for patients, the presence of delusions; for nurses, the presence of perceptual disturbances or overall severe delirium; and for spouses, the low ability to function was predictive of distress level. Spouses perceived the delirium as indicating a high risk for death and loss of the loved one, contributing to bereavement. Medical complications, including decubitus ulcers, feeding problems, and urinary incontinence, are common in patients with delirium. Effects on hospital length of stay, “persistence” of cognitive impairment, increased rate of institutionalization, and reduced ambulation and activities of daily living (ADL) level have been reported.
The Academy of Psychosomatic Medicine Task Force on Mental Disorders in General Medical Practice reviewed studies finding that comorbid delirium increased hospital length of stay 100% in general medical patients, 114% in elderly patients, 67% in stroke patients, 300% in critical care patients, 27% in cardiac surgery patients, and 200%-250% in hip surgery patients. The Academy of Psychosomatic Medicine task force noted that delirium contributed to increased length of stay via medical and behavioral mechanisms, including the following: decreased motivation to participate in treatment and rehabilitation, medication refusal, disruptive behavior, incontinence and urinary tract infection, falls and fractures, and decubiti.
Significantly increased length of stay associated with delirium has been reported in many studies but not all. A meta-analysis of eight studies supported statistically significant differences in length of stay between delirium and control groups. Ely et al. () found that delirium duration was associated with length of stay in both the medical ICU and the hospital (P<0.001) and was the strongest predictor of length of stay even after adjustment for illness severity, age, gender, and days of opiate and narcotic use (). McCusker et al. () studied elderly medical inpatients and found significantly longer length of stay for those with incident, but not prevalent, delirium. Methodological issues often affect interpretation of such studies.
Franco and colleagues identified the increased costs associated with delirium in a prospective study of 500 elective surgery patients older than 50 years. Delirium occurred in 11.4% of the patients during postoperative days 1-4, and these patients had higher professional, consultation, technical, and routine nursing care costs. Milbrandt and colleagues compared costs associated with having at least one delirium episode in 183 mechanically ventilated medical ICU patients and nondelirious control subjects after controlling for age, comorbidity of illness, degree of organ dysfunction, nosocomial infection, and hospital mortality. Median ICU costs per patient were $22,346 for delirious and $13,332 fornondelirious patients (P<0.001), and total hospital costs were $41,836 and $27,106 (P=0.002), respectively; more severe delirium cases resulted in higher costs than did milder ones.
Decreased independent living status and increased institutionalization during follow-up after a delirium episode were found in many studies, especially in the elderly.
Reduction in ambulation and/or ADL level at follow-up is also commonly reported. Delirium also has an effect in nursing home settings, where incident cases are associated with poor 6-month outcome, including behavioral decline, initiation of physical restraints, greater risk of hospitalization, and increased mortality. Even subsyndromal delirium is reported to increase index admission length of stay and postdischarge dysfunction and mortality after adjustment for age, sex, marital status, previous living arrangement, comorbidity, dementia status, and clinical and physiological severity of illness.
In nursing home patients, better cognitive function at baseline was associated with better outcome from delirium, supporting the notion that impaired brain reserve is an important predelirium factor that needs to be taken into account in any longitudinal outcome assessments. Alternatively, longitudinal postdelirium cognitive assessments of younger adults who are not at risk for dementia could help answer the question of persistent cognitive impairment following delirium.