Validation of the confusion assessment method in detecting postoperative delirium in cardiac surgery patients.

Abstract

Early detection, prevention, and treatment of delirium after cardiac surgery are important for quick postoperative recovery. The Confusion Assessment Method (CAM) may be an easy-to-use instrument for detecting delirium in clinical practice. To compare the congruent validity of the CAM with the results from repeated assessments by using a combination of the Organic Brain Syndrome Scale and the Mini-Mental State Examination according to the Diagnostic and Statistical Manual of Mental Disorders (Fourth Edition, Text Revision) criteria for delirium. Patients aged 70 years or older undergoing cardiac surgery were assessed on postoperative days 1 and 4, and the 2 diagnostic methods were compared. The sensitivity and specificity of the CAM were examined. The reference method allowed categorization of delirium into subgroups of psychomotor activities and psychiatric symptom profiles, which were compared with the CAM results. Postoperative delirium was diagnosed in 78 of 141 patients (55.3%). According to the CAM, 59 patients (41.8%) were categorized as delirious, 53 correctly. Thus, the sensitivity was 68% and the specificity was 90%, indicating false-negative rather than false-positive observations. Patients with psychomotor hyperactivity and mixed psychotic-emotional symptoms were more likely to have delirium detected via the CAM than were patients with less obvious clinical manifestations of delirium. Repetitive cognitive testing and psychogeriatric experience are probably necessary to improve the results obtained with the CAM.