Delirium definition influences prediction of functional survival in patients one-year postcardiac surgery

Abstract

BackgroundDelirium after cardiac surgery is associated with prolonged intensive care unit (ICU) and hospital length of stay and elevated rates of mortality. The Society of Thoracic Surgery National Database (STS-ND) includes delirium in routine data collection but restricts its definition to hyperactive symptoms. The objective is to determine whether the Confusion Assessment Method for ICU (CAM-ICU), which includes hypo- and hyperactive symptoms, is associated with improved prediction of poor 1-year functional survival following cardiac surgery. MethodsClinical and administrative databases were used to determine the influence of postoperative delirium on 1-year poor functional survival, defined as being institutionalized or deceased at 1 year. Patients experiencing postoperative delirium using the STS-ND definition (2007-2009) were compared with patients with delirium identified by the CAM-ICU (2010-2012). A propensity score match was undertaken, and multivariable Cox proportional hazards regression models were generated to determine risk of poor 1-year functional survival. ResultsThere were 2756 and 2236 patients in the STS-ND and CAM-ICU cohorts, respectively. Propensity matching resulted in a cohort of 1835 patients (82.1% matched). The overall rate of delirium in the matched study population was 7.6% in the STS-ND cohort and 13.0% in the CAM-ICU cohort (P < .001). Delirium in the CAM-ICU cohort was independently associated with poor 1-year functional survival (hazard ratio, 2.58; 95% confidence interval, 1.20-5.54; P = .02); delirium in the STS-ND cohort was not associated with poor 1-year functional survival (hazard ratio, 0.92; 95% confidence interval, 0.49-1.71; P = .79). ConclusionsA systematic screening tool identifies postoperative delirium with improved prediction of poor 1-year functional survival following cardiac surgery.