Delirium in intensive care patients

BMJ 2012; 344 doi: http://dx.doi.org/10.1136/bmj.e346 (Published 09 February 2012) Cite this as: BMJ 2012;344:e346
  1. Valerie Page, consultant in anaesthesia and critical care
  1. 1Watford General Hospital, Watford WD18 0HB, UK
  1. valerie.page{at}whht.nhs.uk

Debate about assessment tools is overshadowing the importance of delirium

The scientific evidence is irrefutable—delirium in the intensive care unit is an independent predictor of death and acquired dementia.1 2 The linked study by Van den Boogaard and colleagues (doi:10.1136/bmj.e420) is the largest study on delirium in intensive care to date, and it provides a risk model to determine the likelihood of patients in intensive care developing delirium.3 The model (PRE-DELIRIC), which determines 10 risk factors, was developed and validated at the Radboud University Nijmegem Medical Centre in the Netherlands. It was then externally validated at four other Dutch hospitals. The risk model showed a high predictive value, and it was significantly better than the predictions of doctors and nurses.

Reassuringly there are no surprises; risk factors that confer the highest risk are coma with any cause, sedatives, and infection. Notably there were too few patients with alcohol dependency or dementia for these subgroups to be included in the model. However, a prediction model is not needed to confirm that these patients have a high risk of delirium because alcohol dependency and cognitive impairment are significant risk factors in any clinical setting,4 and both are non-modifiable.

Risk prediction models are increasingly influential. Once an effective risk model has been fully appraised and validated in the medical context in which it will be used, the next challenge is to implement it in practice. Having been successfully validated in intensive care, PRE-DELIRIC now faces that formidable task. An essential requirement for the model to be implemented is that critical care units are using computerised data input, but this is not the main barrier. The problem is more fundamental—whether intensivists believe that delirium is important and whether sedated critically ill patients can be diagnosed as delirious. A survey in 2008 showed that only 18% of consultants in intensive care knew that delirium is associated with subsequent persistent cognitive impairment.5 Although some intensivists might consider delirium in a patient with sepsis not important, all would recognise septic encephalopathy as so. Intensive care consultants recognise the importance of septic encephalopathy, but not that of delirium as the presenting symptom.

Intensive care is extreme medicine—consider fever in a medical patient as compared with severe sepsis in intensive care, or a clinic patient with an oxygen saturation of 93% as compared with hypoxia in a patient ventilated in intensive care. Delirium is common in intensive care—affecting 65% of sick ventilated patients in the United Kingdom6—and patients have multiple risk factors.1 Because delirium is mostly hypoactive, where patients are quietly confused and apparently compliant, it remains underdiagnosed and underappreciated. This is infuriating for clinicians who see the devastating impact of delirium—over and above the effects of the illness for which they were admitted—on patients, their relatives, and friends.7 Unless this problem is tackled, the implementation of risk prediction of delirium in intensive care will not progress.

In 1959, after seminal work that linked changes on electroencephalography to the alteration of consciousness in delirium, Engel and Romano bemoaned the fact that clinicians were more concerned with protecting the functional integrity of the heart, liver, and kidneys than that of the brain.8 Why would clinicians neglect brain function? The degree of cognitive impairment is related to the duration of delirium, not to the number of days a patient spends on a ventilator.2 Clinicians who consider delirium to be an epiphenomenon might ask whether renal failure is also an epiphenomenon.9

In 2001, two delirium screening tools were made available for intubated critically ill patients—the confusion assessment method for the intensive care unit (CAM-ICU) and the intensive care delirium screening checklist (ICDSC).10 11 Both tools performed well against gold standard diagnosis using the Diagnostic and Statistical Manual of Mental Disorders, fourth edition. A positive CAM-ICU assessment or ICDSC score of 4 or higher is important because it is associated with increased mortality and cognitive impairment,1 12 and it represents a robust clinical marker that needs attention.

One major difference between the tools is whether sedation confounds the diagnosis of inattention—reduced ability to focus, sustain attention, or shift attention—a core feature of delirium. Patients who are attentive do not have delirium. Intensive care clinicians often debate whether sedated patients can be diagnosed as delirious, but they are missing the point. Although screening tools have a place, it is not whether one is better or even valid; what matters for patients is for clinicians to recognise that they are delirious. Intensive care clinicians with a clinical understanding of the delirious state realise that patients who are thought to be depressed are actually delirious, as are apathetic immobile patients, who were previously thought to be still recovering from sedation.

Without a diagnosis of delirium, the precipitating cause cannot be identified or treated. By failing to identify patients at high risk, the opportunity to prevent delirium by modifying predisposing risk factors, as recommended by the National Institute for Health and Clinical Excellence, is lost.13 Research into delirium in the clinical setting needs to move from observation to intervention to identify drugs that can prevent or modify delirium and improve outcomes. Hopefully, such studies would answer the question of whether high risk patients and hypoactive delirious patients need to be treated with antipsychotics. Finally unless it is known that one type of delirium, such as sedation induced delirium, is less serious than another, such as septic induced delirium, clinicians must look for delirium in every patient using whatever means they have confidence in and decide at the very least “is there anything I can do to treat the cause?”


Cite this as: BMJ 2012;344:e346


  • Research, doi:10.1136/bmj.e420
  • Competing interests: The author has completed the ICMJE uniform disclosure form at www.icmje.org/coi_disclosure.pdf (available on request from the corresponding author) and declares: no support from any organisation for the submitted work; no financial relationships with any organisations that might have an interest in the submitted work in the previous three years; no other relationships or activities that could appear to have influenced the submitted work.

  • Provenance and peer review: Commissioned; not externally peer reviewed.