Incidence and Risk Factors for Delirium in Severely Injured Patients

This study was a prospective cohort study that reviewed the medical records of 179 trauma patients who were admitted to Asan Medical Center from January 01, 2013 to December 31, 2014. The investigators included all patients who had an Injury Severity Score (ISS) of more than 15 and excluded patients younger than 15 years old. This study was approved by the Asan Medical Center Institutional Review Board (2014-0344).

Patient delirium was checked using the Confusion Assessment Method in Intensive Care Unit (CAM-ICU) three times every day. The CAM-ICU includes four features that assess acute changes in or the fluctuating course of mental status, inattention, altered level of consciousness, and disorganized thinking. If a patient has a Richmond Agitation Sedation Scale (RASS) score of -3 or more with acute changes in mental status, inattention, and disorganized thinking or altered level of consciousness, a diagnosis of delirium can be made. The investigators considered patients to be delirious when CAM-ICU was positive for more than 24 h.

The investigators reviewed clinical information for the study patients including age, sex, and underlying disease. Trauma factors reviewed were cause of trauma, ISS, Glasgow Coma Scale (GCS) score, shock, initial lactate level, massive transfusion within 24 h, cardiopulmonary resuscitation rescue, and time taken to the hospital after trauma. Factors related to treatment included use of restraints, mechanical ventilation, and surgery. Clinical outcomes included mortality, ICU stay, mechanical ventilation duration, and hospital stay.

Statistical analysis were done using both univariate (chi-square test and t-test) and multivariate (logistic regression) procedures. Results are reported as mean ± standard deviation. Significance set at a P value of less than .05. For the multivariable logistic regression, the primary outcome variable of interest was the development of delirium after trauma that was classified as either present or absent. Variables were entered for analysis based on the strength of their univariate association with the presence/absence of delirium.