Intracranial cause of delirium: computed tomography yield and predictive factors

Intern Med J. 2012 Apr;42(4):422-7. doi: 10.1111/j.1445-5994.2010.02400.x.


Background: Computed tomography (CT) of the brain in delirium investigation has a low yield of identifiable causes. We sought to identify the best clinical predictors of an intracranial cause of delirium.

Methods: We performed a case-control study of patients admitted to a delirium unit. Clinical factors of patients with positive findings on scans were compared with those without demonstrated causes. The main outcome measure was intracranial abnormalities accountable for the cause of delirium.

Results: During 18 months, there were 300 admissions to the unit. Mean age of patients was 86.6 years. Among 200 patients who proceeded to CT scanning, only 29 demonstrated intracranial pathology accountable for the cause of delirium, with a yield of 14.5%. There were 13 patients with ischaemic stroke, seven with subdural haemorrhage and nine with intracerebral haemorrhage. In multivariate analysis, new neurological deficits (adjusted odds ratio (OR) 18.17, 95% confidence interval (CI) 5.99-55.15), recent falls history (adjusted OR 5.58, 95% CI 1.90-16.42) and decline in conscious level (adjusted OR 4.58, 95% CI 1.33-15.79) were predictors of clinically meaningful radiological findings. Twenty-six of the 29 patients with scans had these three predictors with a sensitivity of 89.7% (95% CI 78.6-100%).

Conclusion: We identified a history of recent fall as a new independent predictor for clinically relevant intracranial pathology in delirious patients, besides new neurological deficits and decline in conscious state. A flow chart incorporating CT head scanning as part of delirium investigation is proposed.

MeSH terms

  • Aged
  • Aged, 80 and over
  • Australia
  • Brain / diagnostic imaging
  • Brain / pathology*
  • Brain Diseases / complications*
  • Case-Control Studies
  • Delirium / diagnostic imaging*
  • Delirium / etiology*
  • Female
  • Humans
  • Male
  • Middle Aged
  • Predictive Value of Tests
  • Risk Factors
  • Sensitivity and Specificity
  • Tomography, X-Ray Computed