Clinical features

Delirium presents in a phenotypically diverse manner ranging from a hypoactive subtype (65% of cases) with a withdrawn patient with decreased motor activity, to an agitated and overexcited patient described as the hyperactive subtype (25%), or a mix of both symptoms (10%) – although the incidence of each can vary with different durations of observation (Meagher Reference Meagher and Trzepacz2000). Extreme cases can include a catatonic variant, which is not to be confused with pure catatonia (Meagher Reference Meagher and Trzepacz2000). ‘Subsyndromal’ delirium is used to describe patients who have begun to fluctuate in awareness and cognition but do not meet all criteria for delirium (van Velthuijsen Reference van Velthuijsen, Zwakhalen and Mulder2018) and evidence shows existence of ‘persistent’ delirium where many patients are discharged with ongoing delirium (Whitby Reference Whitby, Nitchingham and Caplan2022).

The starting point of delirium may be a prodrome in which increasing symptoms of restlessness, anxiety, irritability and sleep disturbances develop. As delirium progresses, impaired attention, cognition and consciousness, as well as perceptual disturbances and behavioural changes, characterise the syndrome. Of note, hallucinations that can occur in delirium are usually non-auditory and are often misinterpretations of environmental or interpersonal cues, especially in visual and tactile cases (Oldham Reference Oldham, Flanagan and Khan2018).

The clinical presentation of delirium (Box 1) may overlap with other diagnoses, such as those reviewed in the final section of this article. Diagnostic testing may help distinguish between diagnoses that cannot easily be differentiated by history and presentation alone.

Diagnosis

The first step in the evaluation of delirium is the history obtained from an informed observer (e.g. family member or staff), followed by a cognitive assessment to screen and confirm delirium. The Single Question in Delirium (SQiD) asks the patient's family/carers whether they feel that the patient has been more confused lately, which may be a feasible screening tool when limited time is available for interview (Sands Reference Sands, Sharma and Carpenter2021). It is crucial to define a baseline level of cognition, consciousness and attention – ideally, in at-risk or prodromal patients, but if necessary also in those who are already in the throes of an acute episode. In addition, given the fluctuating nature of this syndrome, being oriented during one evaluation does not necessarily rule out the diagnosis.

The current standard diagnostic criteria for delirium are outlined in DSM-5-TR (American Psychiatric Association 2022) and ICD-11 (World Health Organization 2022). In terms of preferred screening tools, variability is commonplace. In review of studies to date, work by MacLullich and colleagues (MacLullich Reference MacLullich, Shenkin and Goodacre2019) suggests that the 4 ‘A's Test (4AT) (Table 1) may be the screening tool of choice in the general medical setting. Based on a well-conducted head-to-head trial comparing it with the Confusion Assessment Method (CAM), 4AT showed higher sensitivity and similar specificity (Shenkin Reference Shenkin, Fox and Godfrey2019). A key benefit of the 4AT is that it is quick and simple to administer and a range of health and social care practitioners can carry out the screen without specialised training. Therefore, it is recommended that staff members who have frequent scheduled contact with the patient (e.g. nursing staff) perform the screening and that patients with positive screens are brought to the attention of the healthcare provider for further diagnostic evaluation. Modified versions of the CAM have been identified as effective tools in different settings, such as the ICU (CAM-ICU) with sedated patients or the time-constrained emergency department (Brief CAM) (Box 2).

TABLE 1 Outline of the 4 ‘A's Test (4AT) delirium assessment tool^a

a The full assessment tool and additional information on use are freely accessible on the website developed by the test's author, Alasdair MacLullich (www.the4at.com).

b If patient is asleep attempt to wake with speech or gentle touch on shoulder. Ask patient to state their name and address to assist rating.

c To assist initial understanding one prompt of ‘what is the month before December?’ is permitted.

Workup

Delirium can be the initial signal of a medical emergency. Consequently, the next steps in evaluating delirium must include the timely evaluation of potential causes and precipitating factors. Table 2 summarises the potential routine investigations and additional methods of workup for specific indications. Of note, the selection of the appropriate workup needs to be context specific and may be varied according to local service provision and practice. Moreover, it is important to note that, according to one study, no attributable cause of delirium may be identified in up to 30% of cases in clinical settings (van Velthuijsen Reference van Velthuijsen, Zwakhalen and Mulder2018).

TABLE 2 Workup for delirium

a Delirium can result in a characteristic pattern of ‘a diffuse slowing with increased theta and delta activity and poor organization of background rhythm’ on the EEG that reflects its severity (Jenssen Reference Jenssen2005).

Sources: Marcantonio (Reference Marcantonio2017); Oh (Reference Oh, Fong and Hshieh2017).

Managing risk factors

Risk factors for delirium among ICU patients have been well studied (Box 3) and several risk factors can apply universally to populations in any clinical setting, such as older age, frailty, functional disabilities and high burden of coexisting conditions (including baseline cognitive impairment). Although many of these risk factors are non-modifiable, identifying high-risk patients can help improve the proactive implementation of preventive strategies such as closer monitoring, identifying the target areas for intervention, and prognostic decision-making.

Non-pharmacological prevention strategies

Multicomponent approaches have the strongest evidence for the prevention of delirium, compared with any single-component interventions (Kim Reference Kim, Kim and Ryu2022). The ‘ABCDEF bundle’ (Box 4) is an example of a multicomponent approach that has been recommended for reduction of delirium risk in the ICU (Devlin Reference Devlin, Skrobik and Gélinas2018). The above-mentioned Hospital Elder Life Program is another widely disseminated and evidence-based multicomponent intervention that is less specific to the ICU setting and can be applied in the general medical setting (Inouye Reference Inouye, Baker and Fugal2006). The protocol uses non-pharmacological approaches to systematically target visual and hearing impairment, immobility, disorientation, sleep deprivation and dehydration.

Individual approaches that have shown efficacy in prevention can be grouped into the following categories: (a) avoiding iatrogenic causes; (b) ensuring smooth bodily functioning, which in turn optimises brain function; (c) regular monitoring and social interactions.

Avoiding iatrogenic causes includes medication reviews and avoidance of medications with high deliriogenic potential (e.g. benzodiazepines, meperidine) or anticholinergic load. This includes avoiding GABAergic agents (e.g. propofol, midazolam) for agitation, particularly in older persons. Table 3 summarises the medications to be reviewed for potential removal or titration based on each patient's clinical circumstances by weighing the benefits and risks associated with the medication and the condition it is treating. Consultations with a clinical pharmacist are helpful in making the relevant clinical decisions to cautiously titrate down any medications that may be contributing to delirium. Medicines reconciliation together with a pharmacist before discharge is also essential to reduce the contributory effects of these agents.

TABLE 3 Medications to review for delirium prevention

a Collectively identified as anticholinergics.

b Individual agents identified as having particularly high anticholinergic and deliriogenic potentials.

c Formally classified as ‘sedative hypnotics’ in the previous version of the AGS Beers Criteria.

d Although these may be required for conditions such as chronic obstructive pulmonary disease exacerbation, should be prescribed in the lowest effective dose for shortest possible duration.

Source: This list was created based on the findings of our literature review for this article and the American Geriatrics Society 2019 AGS Beers Criteria (American Geriatrics Society Beers Criteria^® Update Expert Panel 2019).

Infection prevention during hospital admission is another important step, which can be achieved by avoiding unnecessary urinary catheterisation and implementing appropriate infection control protocols.

Although physical restraints are often employed in clinical settings for hyperactive delirium, these should be avoided as much as possible, as they actually increase the risk of delirium (Wilson Reference Wilson, Mart and Cunningham2020). Physical restraint in ICUs (as opposed to other hospital areas) may be justified by the perceived requirement to prevent the removal of life-sustaining devices such as endotracheal tubes and vascular access catheters; however, there is little evidence that restraint alone effectively achieves this goal (Smithard Reference Smithard and Randhawa2022).

The second category – ensuring brain optimisation – is achieved by ensuring smooth bodily functioning (Box 5). Regular monitoring of at-risk patients can be achieved through the routine use of delirium screening scales (e.g. by each shift up to post-operative day 5) and diagnostic interviews by properly trained staff. Staff education focused on recognition, screening, risk factors, prevention and management has also been shown to be effective, as it allows for a concerted effort by all members of the treatment team (Rudolph Reference Rudolph, Boustani and Kamholz2011). Assessing for pain, including non-verbal signs in patients with communication difficulties (e.g. people with intellectual disability or dementia, or those unable to speak owing to the presence of an endotracheal or tracheostomy tube) is also crucial to ensure adequate pain management, which is associated with a reduction in delirium risk (Peden Reference Peden, Miller and Deiner2021).

Social interactions are also integral to reducing delirium risk and can be facilitated by family and visitors as well as the hospital staff. Some specific examples include daily awakening protocols for ICU patients (i.e. ‘sedation holidays’), cognitive stimulation (e.g. books, puzzles, games) and reorientation (e.g. introducing self, role and location at every patient encounter; using an orientation board).

Pharmacological prevention strategies

In terms of preventing delirium in critical care patients, use of anaesthetics such as dexmedetomidine (an alpha-2 agonist) as adjuncts to GABAergic agents (e.g. propofol, midazolam) has been supported by evidence (Burry Reference Burry, Hutton and Williamson2019). For pain management, multimodal analgesia should be utilised while seeking to avoid or at least minimise use of opioids (Aldecoa Reference Aldecoa, Bettelli and Bilotta2017).

Acetylcholinesterase inhibitors (e.g. rivastigmine or donepezil) should be continued for patients with dementia who are already on the medication at home (Maldonado Reference Maldonado2017). They should not be newly prescribed peri-operatively for delirium prevention or treatment, however, as clinical trials have shown that agents such as rivastigmine resulted in higher delirium duration and mortality compared with placebo (van Eijk Reference van Eijk, Roes and Honing2010).

Melatonin may reduce the incidence of post-operative delirium by allowing for improved sleep (Maneeton Reference Maneeton, Kongsaengdao and Maneeton2022). Ramelteon (a melatonin receptor agonist) has not shown efficacy in preventing delirium, however, and neither melatonin nor ramelteon individually have shown efficacy in preventing delirium in older persons (Maneeton Reference Maneeton, Kongsaengdao and Maneeton2022). Additional studies are needed for a stronger recommendation and to determine the optimal dosage, duration and formulation of melatonin.

Non-pharmacological management

When encountering a patient experiencing delirium, the first step to take is verbal and non-verbal de-escalation, such as adopting a respectful, non-confrontational and empathetic tone and posture while talking calmly and slowly to the patient (Box 6). It is important to understand how distressing the experience of delirium may be to a patient: one individual describes feeling that they experienced ‘7 years worth of torture’ during 3 weeks with delirium in an ICU (Hudson Reference Hudson2022).

Another immediate step after the identification of delirium should be medical evaluation, appropriate diagnostic tests and clinical consultations to look for identifiable causes. A caveat to note is that in clinical settings, causes of delirium are often multifactorial. In fact, numerous causes of smaller effect can sometimes cumulatively lead to an episode of delirium, which makes it more challenging to identify the causes in vulnerable patients.

As family members can also be distressed when encountering a relative in a state of delirium, family education about the condition and their role in providing interpersonal stimulation and therapeutic presence is vital.

It is also important to note that the management of delirium does not necessarily end on discharge – arranging follow-up services for patients who experienced delirium while in hospital can help manage any complications and ensure rehabilitation to minimise any prolonged negative consequences of delirium.

Pharmacological management