This site is intended for health professionals only

Dealing with delirium: The elephant in the room


Mark Borthwick
Consultant Pharmacist,
Critical Care Oxford Radcliffe Hospitals NHS Trust

In life, there are subjects, topics and problems that are so large, so ubiquitous and so obvious that those that look upon them can utterly fail to notice their existence or significance. Within healthcare, the condition of delirium meets this description.

Delirium is an acute-onset mental disorder, characterised by a reduced ability to focus, sustain or shift attention, in combination with a change in cognition or the development of a perceptual disturbance. The symptoms and severity can fluctuate over hours or days and it can be caused by a wide variety of insults, in isolation or in combination.

The most widely used diagnostic criteria are taken from the Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision (DSM-IVTR), 1 where the first three criteria are diagnostic, while the fourth describes aetiology (see Table 1).

Data from clinical trials describe the frequency of delirium in a wide range of healthcare settings. In general, the more acutely unwell the patient population tested, the greater the incidence of delirium.

The incidence of delirium is:

  • In long-term care institutes in the community – 16%
  • On hospital general medical wards – 22%
  • Hip fracture patients – about 28%
  • Vascular surgical patients – 31%, Surgical ICU 44%
  • Medical ICUs – as high as 80%.

Despite these high incidences, a recent analysis by the National Institute for Health and Clinical Excellence (NICE) in the UK finds that the condition is not well recognised. For instance, coding data for the NHS reports an incidence of 0.02 to 0.6% from hospital episode statistics, a rate that stands in stark contrast to the baseline rates found in clinical trials.2

Delirium is often classified into subtypes according to motor activity: agitated patients are classed as hyperactive, while withdrawn, stuporose patients are termed hypoactive and those that display both at different times are classed as having mixed delirium. The majority of delirious patients do not have agitation and so are quiet and quiescent, not drawing attention to themselves. The underdiagnosis of delirium may partly reflect a belief that delirious patients are all agitated.

The effects of delirium
The consequences of delirium are potentially severe. Subjectively for the patient, it can be an utterly terrifying ordeal (see Figure 1).

Studies in cancer patients reveal that the experience can be distressing regardless of motor subtype, although not all patients remember their delirium. Mildly affected patients are more likely to remember the episode.3

A strategy that focuses interventions on the minority of patients with hyperactive behaviours would leave the majority of patients untreated. A recent study shows that only about 20% of delirious patients display any hyperactive features.4

Objectively, delirium is associated with a threefold increase in mortality among critically ill patients, 2.9 times increased mortality in hip fracture patients and a 1.3-1.4 times mortality among medical patients, all at six months.

Delirious patients have an increased hospital length of stay and are three times more likely to end up as a new admission to a long-term care institute on hospital discharge. Elderly patients who experience delirium are nearly six times more likely to have dementia three years after the episode than those who did not have delirium.2 It might be argued that the presence of delirium is a marker for poorer outcome rather than the cause and this may well be true. However, some studies that alter delirium rates through the use of proactive delirium prevention measures and studies that use guidelines to treat delirium find that there is also an associated reduction in length of stay and mortality.5-6

Larger studies are required to confirm these findings, but they provide some evidence that the management of delirium through both preventative and treatment strategies alters the adverse outcomes associated with the condition.

Preventing delirium
Good nursing and medical care are the keystones of delirium prevention. The early and appropriate management of pain, infection, constipation and dehydration are all thought to be important.

Simple measures such as ensuring the correct use of glasses, hearing aids, encouraging active cognitive stimulation (e.g. through the use of games or reminiscence) and the use of good sleep hygiene are all thought to be important.

The care environment should support orientation by ensuring the patient can maintain the appropriate day-night cycle, controlling noise, light, temperature and ensuring that visual clues that aid orientation are present (such as clocks, calendar and clear signage).2

A thorough review of the medication schedule can identify agents that are known to be deliriogenic and such agents should be considered for removal or dose reduction wherever possible, recognising that this may not, of course, be possible.

The list of potential medicines is large and it may not be at all obvious to pharmacists where the danger lies. Drugs that possess anticholinergic activity are good suspects, but some medications are not widely known to possess such activity and can slip through the screening process (e.g. ranitidine, digoxin or prednisolone).

Other neurotransmitter systems can be implicated such as serotonergic or GABA-minergic systems and so medications such as SSRIs or benzodiazepines need close attention.7 This is complicated by the possibility of withdrawal reactions which have been described for these agents, so length of time of exposure and half-life of the medication in question may influence the decision on whether to stop, start or restart the agent.

Although withdrawal reactions have not been considered by the NICE delirium guideline, healthcare workers must consider common offenders such as alcohol or nicotine.

Some studies have been conducted into the use of medicines for prophylaxis to prevent delirium in medical, surgical and critical care environments.8 The use of cholinergic agents has generally been found to be ineffective. However, the use of antipsychotics has been found to reduce delirium severity and improve outcomes (such as length of stay), but their use has remained controversial, primarily due to concerns over side-effects and known adverse effects in dementia patients. For it must be remembered that dementia is a major risk factor for the development of delirium.

Delirium treatment
Non-pharmacologic delirium prevention strategies appear to reduce the new occurrence of delirium (incident delirium) by about half, meaning that significant numbers of patients still become delirious or present with established delirium (prevalent delirium). A further review and reduction of drivers is a prudent step, followed by an assessment of drug therapy.

There are few studies to guide medication choice. The largest study randomly allocated patients to receive either haloperidol, olanzapine or to enter a control arm (no placebo).9 This study found that haloperidol and olanzapine resolved delirium when compared with the control arm. The average dose of haloperidol was 7mg daily (IM) compared with 4.5mg olanzapine (PO/SL), with both medications reducing delirium scores to the cut-off value after about three to four days.

Olanzapine appeared to be faster to take effect when compared with haloperidol, and other small studies suggest that atypical antipsychotics could be more effective at treating hypoactive delirium when compared with hyperactive delirium. This observation may account for the difference in onset of effect between haloperidol and olanzapine in this study, since older patients are more likely to suffer hypoactive delirium.

The small studies are hypothesisgenerating and so confirmatory studies are required before it can be definitively said that hyperactive delirium should be treated with haloperidol and hypoactive delirium treated with an atypical antipsychotic.

NICE guidance advises that only distressed patients be treated with antipsychotics. The Hu study9 did not recruit patients based on levels of distress and so this aspect of the advice from NICE is built on the concern that the widespread use of antipsychotics may be detrimental to a patient group where dementia is a major risk for delirium and where the use of antipsychotics in dementia is known to be hazardous.

Extreme care must be taken in ensuring that patients are adequately assessed for distress to avoid the situation where only hyperactive patients are treated and hypoactive patients are not.

The Hu study treated patients for seven days, and NICE recommends that a short course of seven days be given. Practically speaking, therapy may well be used for longer than seven days; however, extreme caution should be employed to ensure that the medication is actively withdrawn when the delirium episode has passed.

In patients with persistent delirium, careful consideration should be given to the course length to avoid the possibility of inappropriately treating a more permanent reduction in cognitive function – e.g. dementia or a permanent hypoxic brain injury.

There are no established protocols regarding how to discontinue antipsychotics therapy. Delirium is a fluctuant condition and withholding doses because the patient has a clear spell may not be appropriate. Clinical judgement guides the withdrawal, and should take into account parameters such as severity of delirium, the improvement or otherwise in the potential drivers, sleep patterns, etc.

A common strategy with haloperidol is to reduce the frequency every day or two, retaining the nocturnal dose as the last dose to be discontinued. With olanzapine, it is common to halve the dose (which is usually given at night), before halving again and then stopping completely.

Delirium is an overlooked condition that is not benign and often untreated even when recognised. Pharmacists are well placed to recognise behaviours characteristic of delirium in patients whom they come into contact with or from the descriptions given by healthcare staff or carers. They are able to identify contributory factors, particularly those caused by medications, and make recommendations to reduce the burden that drivers may contribute.

Pharmacists are able to recommend suitable drug therapy to manage the condition and ensure that the therapy is withdrawn at an appropriate time and in an appropriate way. Pharmacists are quite capable of making the elephant in the room obvious to all.

Figure 1. Extract from a patients account of delirium

Another episode I remember very clearly happened possibly the next day. The decision had been made to operate on my right wrist, elbow and shoulder to stabilise the fractures I’d suffered in those areas. I had been moved from the ward to a holding area prior to the operation and, obviously hallucinating, I was utterly terrified.


I was 90% convinced that I was to be filmed being killed, chopped up and bits of me barbecued at a party in the hospital grounds. I thought that I was in a mocked-up, pretend hospital and that the staff and other patients were actors in some horrific film I was to unwittingly star in. I felt that there was a 10% chance that I was hallucinating, but I still believed overwhelmingly that if I went into the operating theatre I would die.


I shouted to other patients that I was going to be killed and asked them to remember my friend Rachel’s name, as she was also involved in the plot to kill me; I was certain that I could hear her voice coming from an adjoining room. This feeling was becoming more and more intense. I tried to plan an escape, thinking that if only I could walk, I would use the grey triangular grab handle dangling above me as a weapon and fight my way out of there, but I concluded that, as I couldn’t walk, I was in a totally hopeless position.


My ashen-faced parents suddenly appeared at my bedside (they had been asked to come in as I was behaving so erratically) and I felt an immense sense of relief; maybe I would survive this after all. I asked my father to get a bottle to take a sample of the oxygen I was breathing, grabbed a sandwich from the table next to me and ate it, believing that I couldn’t be operated on immediately after eating. I felt frustrated that my parents seemed to be humouring me and didn’t appear to believe that the hospital wasn’t real and that I was possibly in great danger.

– Forty-six-year-old man, trauma ward, post road-traffic accident 

1. American Psychiatric Association, Diagnostic and Statistical Manual of Mental Disorders, Fourth Edition, Text Revision, Washington DC, American Psychiatric Association 2000
2. Clinical Guideline 103, Delirium: diagnosis, prevention and management, NICE, July 2010
3. Breitbart W et al. Psychosomatics 2002;43:183- 194
4. Yang FM et al. Psychosomatics 2009;50:248-25
5. Lundström M et al. J Am Geriatr Soc 2005;53: 622-8
6. Naughton BJ et al. J Am Geriatr Soc 2005;53:18- 23
7. Brown TM. Semin Clin Neuropsychiatry 2000;5: 113-24
8. Bourne RS et al. Journal of Psychosomatic Research 2008;65:273-82
9. Hu H et al. Chin. J. Clin. Rehab. 2006;10:188-190

Latest Issue

Be in the know
Subscribe to Hospital Pharmacy Europe newsletter and magazine