Say to the patient: 'I am going to read you 10 letters. Every time you hear the letter A, squeeze my hand.' Then read clearly and at 1-letter-per-second pace: S-A-V-E-A-H-A-A-R-T. Errors = failed to squeeze on A (miss) + squeezed on non-A letters (false positive). Total errors >2 = positive Feature 2. If the patient is intubated and cannot squeeze, use an alternative inattention method (eye blinking, nodding) as described in the CAM-ICU training materials at icudelirium.org.

Yes. RASS -2 (light sedation — briefly awakens to voice, eye contact <10 seconds) is assessable. Feature 3 (altered level of consciousness) will automatically be positive because RASS ≠ 0. You still need to assess Features 1 and 2. If Features 1 and 2 are positive and RASS ≠ 0 (Feature 3 positive), CAM-ICU = POSITIVE. RASS -4 and -5 are the only levels where CAM-ICU cannot be performed.

This is clinically challenging. Hypoactive delirium in a patient on minimal or no sedation (RASS -1 to 0) who has positive CAM-ICU Features 1+2+3/4 is true delirium. A patient on propofol with RASS -2 who is CAM-ICU positive may have delirium OR sedation-mimicking delirium — try reducing sedation to RASS 0 and reassess. If CAM-ICU remains positive when fully awake (RASS 0), it is true delirium.

CAM-ICU can still be applied, but Feature 1 (acute/fluctuating change) must be assessed relative to the patient's baseline, not against a 'normal' standard. The key is demonstrating an ACUTE change from the patient's known baseline cognitive state. Obtain collateral history from family or long-term care facility about baseline behavior, orientation, and attention. Without a reliable baseline, delirium is more difficult to diagnose.