Raw 3D-CAM severity score — Raw 3D-CAM Delirium Severity Score

What this tool is and how to interpret its evidence

Raw 3D-CAM is an alternative method for grading delirium severity from an assessment already completed with the parent 3D-CAM. It adds the positive findings across the 20 core interview and observer-rated results, producing a total from 0 to 20. The contributing results comprise cognitive-test performance, patient-reported symptom probes and interviewer observations. Higher totals indicate a greater burden of delirium-related findings. The method does not require a second patient interview, and the authors did not propose a diagnostic cutoff: it is a severity measure, not an independent test of whether delirium is present.

Vasunilashorn and colleagues applied the method to the original 3D-CAM validation cohort of 201 older hospitalised adults, whose mean age was 84 years and of whom 28% had dementia. The raw score discriminated delirium determined by an independent clinical reference interview with an area under the curve of 0.950 and offered a substantially wider distribution than the earlier 0-7 3D-CAM-S method. This wider range may be useful for research, tracking change and representing subsyndromal symptoms.

The evidence remains internal because the analysis reused the parent instrument's original cohort. Responsiveness over time and relationships with length of stay, readmission and mortality were not tested. This page should link to the publication and authorised current 3D-CAM manual, while avoiding reproduction of the protected questions, prompts or feature-level operating rules.