Background: Delirium is common, morbid, and costly, yet is greatly under-recognized among hospitalized older adults.
Objective: To identify the best single and pair of mental status test items that predict the presence of delirium.
Design, setting: Diagnostic test evaluation study that enrolled medicine inpatients aged 75 years or older at an academic medical center.
Methods: Patients underwent a clinical reference standard assessment involving a patient interview, medical record review, and interviews with family members and nurses to determine the presence or absence of Diagnostic and Statistical Manual of Mental Disorders, 4th Edition defined delirium. Participants also underwent the three-dimensional Confusion Assessment Method (3D-CAM), a brief, validated assessment for delirium. Individual items and pairs of items from the 3D-CAM were evaluated to determine sensitivity and specificity relative to the reference standard delirium diagnosis.
Results: Of the 201 participants (mean age 84 years, 62% female), 42 (21%) had delirium based on the clinical reference standard. The single item with the best test characteristics was "months of the year backwards" with a sensitivity of 83% (95% confidence interval [CI]: 69%-93%) and specificity of 69% (95% CI: 61%-76%). The best 2-item screen was the combination of "months of the year backwards" and "what is the day of the week?" with a sensitivity of 93% (95% CI: 81%-99%) and specificity of 64% (95% CI: 56%-70%).
Conclusions: We identified a single item with >80% and pair of items with >90% sensitivity for delirium. If validated prospectively, these items will serve as an initial innovative screening step for delirium identification in hospitalized older adults.
© 2015 Society of Hospital Medicine.