UB-CAM: Ultra-Brief Confusion Assessment Method

2020

UB-CAM is an adaptive two-stage delirium assessment. A positive UB-2 screen leads to further 3D-CAM assessment with a skip pattern. Paper and app versions are available.

At a glance

Purpose: Episodic assessment
Population: Adult
Setting: General / acute hospital
Time: about 1–2 min on average in published studies; varies with the pathway and patient
Potential items: 20; the number administered varies
Score or threshold: Adaptive CAM-based classification; use the specified version and instructions.

Format: The version 2 manual describes 20 potential items. Questions and ratings are selected through an adaptive sequence. Six was the mean number administered in a preliminary study, not the fixed length of the instrument.

Strengths: Brief adaptive assessment; completion exceeded 97% in the 2022 two-hospital study. Paper instructions and an app are available.
Limitations: Sensitivity was moderate in the 2022 clinical study. A negative first-stage result ends the assessment. Accuracy depends on the pathway, rater and patient group; the current version 2 materials use inconsistent labels for severely reduced consciousness.

Training and instructions

UB-CAM training manual. Developer resource page with the instrument-specific manual. Check the current version and reuse terms.
UB-CAM form. Version 2. The form, manual introduction and later diagnostic instructions use inconsistent labels for severely reduced consciousness; see the version note on this profile.
UB-CAM manual or instructions. Version 2. The form, manual introduction and later diagnostic instructions use inconsistent labels for severely reduced consciousness; see the version note on this profile.

Start here

Tool and original sources

A source for the tool or its access information is linked below. Check the version and current terms before use, copying or adaptation.

Open the official tool page ↗ Open project website ↗

Original paper

Original paperMarcantonio ER, Fick DM, Jung Y, Inouye SK, Boltz M, Leslie DL, et al. Comparative Implementation of a Brief App-Directed Protocol for Delirium Identification by Hospitalists, Nurses, and Nursing Assistants : A Cohort Study. Annals of internal medicine. 2022;175(1):65-73. doi: 10.7326/M21-1687. PMID: 34748377.
This main prospective cohort evaluated the app-directed two-step UB-CAM in 527 patients across 924 research-standard delirium assessment days. Rounded sensitivity and specificity were 65% and 93% for nurses and 63% and 91% for physicians; the separate UB-2 first-stage results should not be substituted for the complete workflow.
Open PubMed ↗

One DTA study identifiedAssessment workflow

Classification

Episodic assessment

A tool may belong to more than one category. Select a linked category to see other relevant profiles.

Evidence

Diagnostic-accuracy evidenceOne DTA study identified

Reported sensitivity and specificity

Headline metric source

Diagnostic accuracy and implementation study · 2022; pmid:34748377

Denominator: 527 patients
Reference standard: Research-standard delirium assessment

56.3–64.4%

Sensitivity

93.1–93.9%

Specificity

2022 Table 4, two-step protocol with skip pattern (UB-CAM): nurses 64.4% sensitivity (95% CI 41.6–82.3) and 93.9% specificity (86.7–97.2), 461 assessments; physicians 56.3% (35.4–76.1) and 93.1% (85.0–96.8), 453 assessments. These are rater-specific results within one 527-patient study, not independent cohorts. The combined two-step estimates include both skip and no-skip protocols.

Expanded evidence profile

What the broader evidence shows

Interpretation: UB-CAM is an app-directed two-step workflow: the ultra-brief UB-2 first stage is followed by 3D-CAM only when the first stage is positive. The main prospective cohort found high completion and high specificity but only moderate sensitivity in routine clinician use; the evidence base is still small, largely developer-associated and without an independent full-workflow replication or dedicated meta-analysis.

Evidence snapshots

Protocol simulation (2020): Secondary-data simulation suggested that staged UB-2 then 3D-CAM could retain similar sensitivity and specificity while reducing average administration time; this was not a new prospective bedside validation.
Main prospective cohort (2022): Across 924 reference-standard delirium assessments in 527 patients, sensitivity was 65% for nurses and 63% for physicians, while specificity was 93% and 91%, respectively.
UB-2 first stage (2022): The first-stage UB-2 alone had sensitivity of 82% to 88% and specificity of 64% to 70%, so it should not be interpreted as the complete UB-CAM workflow.
Evidence boundary (2024): A secondary analysis examined disagreement rather than independently re-estimating sensitivity and specificity; dementia and subsyndromal delirium contributed to false positives, while mild delirium contributed to false negatives.

These results come from different populations and workflows and should not be averaged or treated as interchangeable.

6 selected PubMed-indexed papers, 2018–2024

2018Original development study
Fick DM, Inouye SK, McDermott C, et al. Pilot Study of a Two-Step Delirium Detection Protocol Administered By Certified Nursing Assistants, Physicians, and Registered Nurses. J Gerontol Nurs. 2018;44(5):18-24. doi: 10.3928/00989134-20180302-01. PMID: 29596707.
This 24-patient pilot tested the feasibility of a two-step protocol delivered by nursing assistants, nurses and physicians; 22% of participants had delirium. Completion was 100% and 91% of assessments met the intended time window, but the sample was far too small for stable diagnostic-accuracy estimates.
PubMed · PMID 29596707 ↗
2020Simulation study
Motyl CM, Ngo L, Zhou W, et al. Comparative Accuracy and Efficiency of Four Delirium Screening Protocols. J Am Geriatr Soc. 2020;68(11):2572-2578. doi: 10.1111/jgs.16711. PMID: 32930409.
This study simulated four screening protocols using existing 3D-CAM data from 201 patients and READI data from 330 patients rather than prospectively administering full UB-CAM. The staged UB-2 then 3D-CAM approach preserved simulated accuracy and reduced estimated average time from 3 minutes 13 seconds for full 3D-CAM to 1 minute 14 seconds.
PubMed · PMID 32930409 ↗
2021Clinical implementation study
Husser EK, Fick DM, Boltz M, et al. Implementing a Rapid, Two-Step Delirium Screening Protocol in Acute Care: Barriers and Facilitators. J Am Geriatr Soc. 2021;69(5):1349-1356. doi: 10.1111/jgs.17026. PMID: 33474729.
This mixed-methods implementation study involved 50 hospitalists, 189 nurses and 83 nursing assistants, with 767 protocol administrations and 231 qualitative interviews. It showed that professional role, local context, perceived responsibility and integration into workflow affected uptake, but it was not an independent diagnostic-accuracy validation.
PubMed · PMID 33474729 ↗
2022Diagnostic-accuracy study
Marcantonio ER, Fick DM, Jung Y, et al. Comparative Implementation of a Brief App-Directed Protocol for Delirium Identification by Hospitalists, Nurses, and Nursing Assistants: A Cohort Study. Ann Intern Med. 2022;175(1):65-73. doi: 10.7326/M21-1687. PMID: 34748377.
The main prospective cohort included 527 patients, 924 research-standard delirium assessments and 399 participating clinicians at two hospitals, with protocol completion above 97%. Sensitivity and specificity were 65% and 93% for nurses and 63% and 91% for physicians, while sensitivity increased to 78% for moderate-to-severe delirium; median completion time was about 104 to 106 seconds.
PubMed · PMID 34748377 ↗
2023Clinical review
Shrestha P, Fick DM. Recognition of Delirium Superimposed on Dementia: Is There an Ideal Tool? Geriatrics (Basel). 2023;8(1):22. doi: 10.3390/geriatrics8010022. PMID: 36826364.
This clinical review considered delirium detection when dementia is present and discussed UB-CAM among the available approaches. It highlighted the limited accuracy evidence in delirium superimposed on dementia and the need to establish baseline cognition rather than treating any single score as definitive.
PubMed · PMID 36826364 ↗
2024Secondary analysis
Dhliwayo R, Trivedi S, Ngo L, et al. Factors associated with disagreement between clinician app-based ultra-brief Confusion Assessment Method and reference standard delirium assessments. J Am Geriatr Soc. 2024;72(3):828-836. doi: 10.1111/jgs.18690. PMID: 38014821.
This secondary analysis reused the 527-patient prospective cohort and examined 1,795 clinician-paired assessments, with disagreement in 12%. Dementia and subsyndromal delirium were associated with false positives, while mild delirium contributed to false negatives, so the study explains errors but is not an independent replication.
PubMed · PMID 38014821 ↗

This is a selected evidence overview, not a systematic review. Six PubMed records were retained because the full UB-CAM literature is small: a feasibility pilot, a secondary-data simulation, an implementation study, the main prospective diagnostic cohort, a relevant clinical review and a follow-up analysis of the same cohort. All six records were checked in PubMed and DOI-bearing papers were checked in Scite on 31 August 2026; no dedicated full UB-CAM systematic review or independent replication was identified.

More evidence and citation details

Full name: Ultra-Brief Confusion Assessment Method
Also known as: UB-CAM, Ultra-Brief CAM, app-directed CAM
Related profiles: 3D-CAM, UB-2
Evidence summary: One prospective diagnostic-accuracy and implementation cohort plus feasibility, simulation and follow-up evidence; no dedicated tool-specific systematic review
Diagnostic-accuracy evaluations: 1
Reference standard: Research reference standard
Reliability: ;
Citation: Marcantonio ER, Fick DM, Jung Y, Inouye SK, Boltz M, Leslie DL, Husser EK, Shrestha P, Moore A, Sulmonte K, Siuta J, Boustani M, Ngo LH Comparative Implementation of a Brief App-Directed Protocol for Delirium Identification by Hospitalists, Nurses, and Nursing Assistants : A Cohort Study Annals of internal medicine. 2022;175(1):65-73. PMID 34748377

Download RIS ↗ Download BibTeX ↗

Primary validation studies

Distinct primary publications from the current diagnostic-accuracy audit. Reviews and repeated evaluations within one paper are not counted again. Overlapping cohorts have not all been reconciled. Read the counting rules and limitations.

Open 1 identified paper

Marcantonio ER, Fick DM, Jung Y, Inouye SK, Boltz M, Leslie DL, Husser EK, Shrestha P, Moore A, Sulmonte K, Siuta J, Boustani M, Ngo LH (2022). Comparative Implementation of a Brief App-Directed Protocol for Delirium Identification by Hospitalists, Nurses, and Nursing Assistants : A Cohort Study. Annals of internal medicine.
Reference category: other comparator or secondary accuracy evidence.

Version: Version 2 available from the official page when checked
Rights holder: Hospital Elder Life Program (CAM copyright); Aging Brain Center (permissions administrator); component notices require rights review
Licence: MIXED-RIGHTS-REVIEW-REQUIRED
Information checked: 2026-07-19

Instrument-specific terms apply. See the access and copyright policy or send a correction.

Base clinical review: 2026-08-31. Instrument description source check: 2026-09-05. Scope and limitations.