Validation of the AWOL Tool: A Delirium Prediction Rule (P6.194)

Objective: To validate a prediction rule administered at hospital admission to identify patients at high risk for delirium in a neuroscience population. Background: Delirium is a frequent and potentially preventable complication of inpatient hospitalization. Prediction rules that identify those patients at highest risk would help target the application of preventive measures. Methods: We previously developed and validated a delirium prediction rule where patients score one point each for Age over 80, inability to spell World backwards, disOrientation, and moderate iLlness severity or greater based on nursing assessment (the AWOL score). This score was incorporated into routine admission assessment on a hospital neurosciences floor for all patients over 50, regardless of the presence or absence of delirium on admission. We conducted a retrospective chart review of these patients (n=172) to identify cases of delirium using validated methodology. Performance of the AWOL score in predicting both incident (developing during hospitalization) and prevalent (present upon admission) delirium was evaluated using Cuzick’s nonparametric test for trend and area under the receiver operating characteristic curves (AUC). Results: The rate of incident delirium (per AWOL score) was 4[percnt] (0), 6[percnt] (1), 42[percnt] (2), and 57[percnt] (3) (p<0.001; AUC 0.78 (95[percnt] CI 0.64-0.93)). The rate of prevalent delirium (per AWOL score) was 4[percnt] (0), 9[percnt] (1), 25[percnt] (2), 59[percnt] (3), and 100[percnt] (4) (p<0.001; AUC 0.84 (95[percnt] CI 0.75-0.93). Rates of combined incident and prevalent delirium (per AWOL score) were 8[percnt] (0), 14[percnt] (1), 56[percnt] (2), 82[percnt] (3), and 100[percnt] (4) (p<0.001; AUC 0.83 (95[percnt] CI 0.75-0.91)). All patients with an AWOL score of 4 had prevalent delirium. Conclusions: The AWOL score can be used by nurses in routine clinical practice to effectively identify patients with prevalent delirium and those at risk for development of delirium in the hospital, allowing targeted application of delirium prevention measures. Disclosure: Dr. Brown has nothing to disclose. Dr. Josephson has received personal compensation in an editorial capacity for Journal Watch Neurology and Continuum. Dr. Douglas has nothing to disclose.