A novel scoring system to predict delirium and its relationship with the clinical course in patients with acute decompensated heart failure

Background: Delirium is known to be a poor prognostic factor in patients with acute decompensated heart failure (ADHF). The purpose of this study was to determine predictors of delirium on admission of ADHF patients, and to establish a scoring formula to identify patients at high risk for delirium.

Methods and results: We recorded the Intensive Care Delirium Screening Checklist (ICDSC) score in 120 ADHF patients during their stay in the coronary care unit (CCU). Patients with a highest ICDSC score of 4 or more were diagnosed with delirium. We examined independent candidate predictors of delirium using multivariate logistic regression analysis and developed the following scoring formula, the delirium prediction score (DPS), using independent predictors of delirium and their regression coefficients: DPS=inferior vena cava diameter+C-reactive protein (and additionally +10 for patients with a history of cerebral infarction). Receiver operating curve analysis indicated that evaluation using this scoring system at the time of admission was able to predict delirium with high accuracy (C-statistic: 0.885). In addition, the calculated scores had significantly positive correlations with duration of CCU stay and overall length of hospital stay.

Conclusions: We established a novel scoring system to predict on admission the likelihood of development of delirium in ADHF patients; this system also predicts prolongation of intensive care and hospital stay.