CAM-ICU adapts the Confusion Assessment Method for use in critically ill adults, including mechanically ventilated and non-verbal patients, by structuring assessment of four features: acute onset or fluctuating mental status, inattention, altered level of consciousness and disorganised thinking.

Evidence source	Population / setting	Outcome	Key findings
Ely et al., Crit Care Med 2001	38 adult medical and coronary ICU patients; 293 paired daily assessments	Delirium by DSM-IV expert review vs CAM-ICU ratings by nurses and an intensivist	Delirium in 87% of patients; CAM-ICU sensitivity 95-100% and specificity 89-93; inter-rater reliability kappa 0.79-0.95.
Ely et al., JAMA 2001	96 mechanically ventilated adults in medical and coronary ICUs; 471 paired assessments	Delirium by DSM-IV expert review vs nurse-rated CAM-ICU	Delirium in 83% of patients; sensitivity 93-100%, specificity 98-100; kappa 0.96; mean administration time about 2 minutes.
Gusmao-Flores et al. 2012	9 CAM-ICU studies; 969 critically ill adults	Meta-analysis of CAM-ICU accuracy vs DSM criteria	Pooled sensitivity 80% and specificity 95.9%; area under the summary ROC curve 0.97.
van Eijk et al. 2011	10 mixed ICUs; 282 patients (181 assessable)	Routine bedside nurse CAM-ICU vs delirium expert teams	Sensitivity 47% and specificity 98% in daily practice, emphasising the impact of staff training and implementation.

Use and limitations

CAM-ICU is designed for adult ICU patients, including those who are intubated or otherwise unable to speak, and can be applied by nurses and doctors after structured training.
Delirium in critically ill adults is common and associated with increased mortality, longer duration of mechanical ventilation and prolonged ICU and hospital stay; routine CAM-ICU screening supports earlier recognition and management.
The 2018 PADIS guideline recommends regular delirium monitoring in adult ICUs using validated tools such as CAM-ICU or the ICDSC, embedded within pain, sedation and mobilisation bundles.
Accuracy is reduced in deeply sedated, neurologically devastated, sensory-impaired or non-English-speaking patients if adaptations and interpreters are not used, and hypoactive delirium is particularly easy to miss; repeat assessments and ongoing staff training are important.
CAM-ICU is validated in adult ICUs; paediatric and neonatal units should use dedicated tools (for example pCAM-ICU or psCAM-ICU) and specialist assessment.

Confusion Assessment Method for the ICU (CAM-ICU)