CAM, cognitive testing and routine monitoring
Commentary. These are the author’s interpretations of the cited evidence. Disclosures.
The method used at the bedside matters. An algorithm alone is not the whole assessment.
The original Confusion Assessment Method (CAM) is an episodic assessment method. Its feature ratings follow an interview that includes cognitive testing, together with observation and history. The developers’ current guidance explicitly states that the short CAM should be scored on the basis of cognitive screening tests. It should not be reduced to an observation-only checklist.
I use “CAM-Lite” to describe unofficial use of the CAM features without the preceding cognitive assessment. This is my descriptive term, not an instrument issued by the CAM developers. These altered procedures cannot inherit the original CAM’s validation results. Repeating a properly performed CAM is a different issue. Frequency alone does not make a CAM assessment invalid.
There is direct evidence of poor recognition when ratings rely on routine observation. Inouye et al. (2001) compared nurse and researcher ratings in 797 hospitalised older people, with 2,721 paired observations. Nurses identified delirium in 19% of affected observations and 31% of affected patients. Researcher ratings included formal cognitive assessment. This study examined nurse recognition using CAM ratings; it was not a validation of a single standardised instrument called CAM-Lite.
The large routine-care studies raise further concerns. Rohatgi et al. (2019) described a programme covering 105,455 encounters. After CAM screening was introduced, 98.8% of encounters had screening recorded, but only 2.4% of screened encounters were CAM-positive. An expert-assessed pilot found delirium in 17.3% of 278 patients. The pilot and the routine dataset were different samples. Dividing one percentage by the other would not produce a valid sensitivity estimate. The report also did not establish that every routine assessment omitted cognitive testing.
Corradi et al. (2016) analysed routine CAM and other clinical records. The Penfold review reported positive CAM records in 6,926 of 88,206 encounters, about 7.9%, and frequent unable-to-assess records. The findings raise concern about unrecognised delirium and how the algorithm was interpreted. They are not a direct diagnostic-accuracy comparison with a contemporaneous expert assessment.
My conclusion is that observation-only CAM ratings should not be used as a replacement for an assessment that includes cognitive testing. The evidence shows poor recognition in relevant observational use and concerning results in large routine programmes. Time spent recording scores does not compensate for a method that misses delirium. False reassurance can delay assessment and treatment of underlying illness. That is a clinical risk; these studies do not quantify the harm caused by this procedure.
This criticism concerns the procedure. It is not a judgement about a person, a research group or every CAM-family tool. CAM-ICU, bCAM, 3D-CAM and UB-CAM have specified procedures and their own evidence. The 4AT and other tools also need scrutiny of routine delivery. The appropriate question is whether the method actually used identifies patients who need care.
References and instructions
Corradi, J.P. et al. (2016) Analysis of multi-dimensional contemporaneous EHR data to refine delirium assessments. Computers in Biology and Medicine, 75, pp. 267–274.
Inouye, S.K. et al. (1990) Clarifying confusion: the Confusion Assessment Method. A new method for detection of delirium. Annals of Internal Medicine, 113, pp. 941–948.
Inouye, S.K. et al. (2001) Nurses’ recognition of delirium and its symptoms: comparison of nurse and researcher ratings. Archives of Internal Medicine, 161, pp. 2467–2473.
MacLullich, A. (2025) The CAM-Lite. Delirium Words. Background to the term used here.
Penfold, R.S. et al. (2024) Delirium detection tools show varying completion rates and positive score rates when used at scale in routine practice in general hospital settings: a systematic review. Journal of the American Geriatrics Society, 72, pp. 1508–1524.
Rohatgi, N. et al. (2019) Initiative for prevention and early identification of delirium in medical-surgical units: lessons learned in the past five years. American Journal of Medicine, 132, pp. 1421–1430.e8.
Delirium Central (accessed 4 September 2026) CAM instruments, manuals and recommended assessment procedure.