Validation of the Delirium Diagnostic Tool-Provisional (DDT-Pro) in a skilled nursing facility and comparison to the...

To validate the Delirium Diagnostic Tool (DDT-Pro) in a SNF and compare its performance to the 4 A's Test (4AT) in real life conditions.

Prospective cross-sectional study of 262 consecutively admitted adults to a SNF, independently assessed by psychiatrists using DSM-5 delirium criteria and by geriatricians using the DDT-Pro (0-9 points) and 4AT (0-12 points).

141 (53.8%) participants had dementia and 79 (30.1%) had delirium. DDT-Pro and 4AT were moderately correlated (-0.59). Accuracies against DSM-5 diagnosis ranged from 80 to 85% and were comparable between tools regardless of dementia. Recommended delirium cutoff for the DDT-Pro (≤6) had 77.2% sensitivity, 84% specificity, and NPV = 89.5% and 4AT (≥4) had 54.4% sensitivity and 92.9% specificity, with lower specificity in the dementia subsample. DDT-Pro sensitivity increased (84.8%) at ≤7cutoff. Sensitivity and specificity of all DDT-Pro and 2/4 4AT items displayed gradients along severity levels, but two dichotomously rated 4AT items had low positivity. The tools had low concordance (p < 0.05) for delirium positivity.

DDT-Pro is valid to detect delirium in SNF population where simple, structured tools with high sensitivity are needed. DDT-Pro items assess the three core domains of delirium as a continuous measure and may have advantages over the 4AT.