Delirium is a complex neuropsychiatric syndrome that is of acute or sub-acute onset that includes deficits in multiple domains of cognition and a range of neuropsychiatric disturbances. It occurs across health-care settings and populations but is especially common in medical–surgical patients, with even higher rates in intensive care units (ICUs) and palliative care services.1, *2 Delirium has been recognised for at least two millennia with the term ‘delirium’ deriving from the Latin ‘lira’ meaning to wander from one's furrow. Prior to Diagnostic and Statistical Manual of Mental Disorders (DSM)-III (1980), acute generalised disturbances of brain function were described by a plethora of labels (acute organic brain syndrome, acute confusional state, brain failure, toxic encephalopathy and intensive care psychosis). However, these terms do not represent distinct scientific entities. Delirium is now the umbrella term in current International Classification of Diseases (ICD) and DSM classification schemes that subsume these multiple synonyms (see DSM-IV criteria in Box 1).3 This simplification has allowed for a more coherent research effort and a consistent approach to detection and management. Moreover, the use of the label delirium has led to greater appreciation of the considerable impact upon outcomes and independent need for treatment as a brain disorder beyond only treating its underlying aetiological precipitants.
In the Emergency Department (ED), the prevalence in elderly patients is 9.6%4 and in medical inpatients the occurrence rate per admission is 11–42%.5 In surgical settings, in orthopaedics the incidence rate is 4–53.3% among hip fracture patients and 3.6–28.3% among elective patients,6 and in cardiac surgery the incidence rate is 3–72%.7 Delirium is an independent risk factor for multiple adverse outcomes. It is associated with an increased length of stay in hospital, new institutionalisation and functional decline.5, *8, 9 Delirium is associated with an increased risk of cognitive decline and acceleration of dementia,10, 11, *12 and it is associated with increased mortality, independent of age, sex, co-morbid illness and baseline dementia.8
Despite these potential adverse consequences and high incidence rates across a range of settings, delirium detection rates remain low. For example, Elie et al. found the sensitivity of the standard clinical assessment by ED physicians to be 35.3%.4 In a study of 710 acute elderly medical admissions, Collins et al. found that 72% of the 110 delirium cases were not detected by the clinical teams.13 There are several contributing factors to why delirium detection is so poor, including clinician, patient and systems issues. Poor detection reflects a lack of knowledge and training in clinical staff, and a lack of confidence in practitioners.14 Lower detection occurs in patients who are older, have co-morbid dementia, present with hypoactive features and are receiving care in surgical rather than medical settings.15 Rapid transfer of patients and busy clinical settings may also make proper delirium diagnosis and documentation challenging.13 There is also no consensus over optimal assessment of delirium.
Although there is compelling evidence that a variety of interventions can successfully prevent delirium,2 there are very few studies which have tested treatments of delirium, and a high quality randomised trial is lacking. Placebo-controlled studies are crucial to estimating the impact of treatment interventions in delirium because of its fluctuating nature and potential for recovery with appropriate management of its underlying causes. Thus, some practitioners underestimate the importance of formal detection of delirium. However, this situation reflects a lack of studies rather than positive evidence that delirium detection and management is ineffective. In fact, delirium treatment is manifestly effective in many situations, such as in opioid toxicity, modifying communication methods so as to detect pain, detecting and managing distress caused by paranoia and hallucinations, and so on. For example, some studies have found that surgical patients with delirium receive less analgesia than those with normal cognition16; this is important because pain itself is a cause of delirium. Moreover, there are almost 30 prospective studies of delirium pharmacotherapy that provide evidence that delirium can be effectively managed without serious adverse effects in the majority of patients.17 Delirium detection is also essential because it often signifies an acute and serious disorder or physiological disturbance such as hypoxia, hypercapnia or sepsis.18 Therefore, despite the lack of randomised controlled trials of delirium treatments, detection is clearly of critical importance in routine clinical practice and can be aligned to prudent use of a range of therapeutic interventions.
In this article, we begin by reviewing the features of delirium in its many presentations, provide a summary of the more commonly applied screening tools, then go on to make recommendations about pragmatic detection and monitoring in the clinical settings outside the ICU where an anaesthetist is likely to encounter delirium.
Although delirium has long been accepted as the formal term for all acute or subacute disorders of generalised brain function, including in DSM-IV, multiple formal or semi-formal terms are still used to describe delirium. Examples include ‘acute confusional state’, ‘organic brain syndrome’, ‘ICU psychosis’, ‘postoperative psychosis’, ‘metabolic encephalopathy’, ‘septic encephalopathy’ and even ‘subacute befuddlement’. Informal terms used by clinicians which often signify delirium include
Delirium is a complex syndrome that can include a wide range of neuropsychiatric and cognitive features. Whilst acute onset and impaired attention are invariably present and are thus mandatory for the diagnosis, other less consistent features can be helpful in indicating possible delirium. No single feature is pathognomonic or indeed specific to delirium; however, the pattern of symptoms that indicates widespread disturbance of brain functions along with the highly characteristic context (i.e.,
Delirium detection remains a major obstacle to improved delirium care in everyday practice with the majority of cases missed or diagnosed late,13 including many cases that have poor outcomes.54 Heightened awareness of delirium along with systematic assessment for core features of delirium such as inattention and altered consciousness can allow for more consistent recognition. Approaches to delirium detection need to be tailored to the characteristics of the setting where they are to be applied
Delirium is a day-to-day part of practice in the acute hospital. It has major implications for patient outcomes, and yet unlike many other mainstream acute syndromes it is greatly under-detected, and poorly managed. The most important obstacle to delirium care is its status as a condition that is not seen as a priority by many practitioners,90 along with a lack of knowledge of its features and thus the best methods for detection.14 Compelling epidemiological evidence demonstrating the
Delirium is a neuropsychiatric syndrome of acute onset and fluctuating course, with inattention, altered conscious level and usually an underlying physical cause. It has a wide range of cognitive and neuropsychiatric features. It is very common across a range of hospital settings and is associated with serious adverse outcomes; despite this, it is under-detected and mostly poorly managed. There is no overall consensus on optimal assessment of delirium. Particularly in the older hospital
AMJM holds patents on objective computerised assessments of attention. AMJM has received honoraria from Lundbeck, Shire, Novartis and Napp. Delirium is common across all settings and often indicates the presence of other serious physiological derangements. Approach each acutely unwell patient with the thought “are they delirious?” Seek acute change or fluctuation in mental status from nursing staff or carer. Use all available sources of information. Observe the patient, and conversePractice points
This work was supported by grants from Research into Ageing and the British Geriatrics Society (RJH), the Medical Research Council (AMJM) and the Health Research Board (DM). RJH and AMJM are members of The University of Edinburgh Centre for Cognitive Ageing and Cognitive Epidemiology, part of the cross council Lifelong Health and Wellbeing Initiative. DM is a member of the Cognitive Disorders Research Group at the University of Limerick, Ireland. Funding from the BBSRC, EPSRC, ESRC, HRB and MRC
