Explore objective clinical variables for detecting delirium in ICU patients: a prospective case-control study / 中华危重病急救医学

ABSTRACT

Objective The aim of this case-control study is to explore clinical objective variables for diagnosing delirium of intensive care unit (ICU) patients.Methods According to the method of prospective case-control study, critical adult postoperative patients who were transferred to ICU of Peking University People's Hospital from October 2015 to May 2016 and needed mechanical ventilation were included. After evaluating the Richmond agitation sedation scale score (RASS), the patients whose score were -2 or greater were sorted into two groups, delirium and non-delirium, according to the confusion assessment method for the ICU (CAM-ICU). Then these patients were observed by domestic multifunctional detector for electroencephalographic (EEG) variables such as brain lateralization, brain introvert, brain activity, brain energy consumption, focus inward, focus outward, cerebral inhibition, fatigue, sleep severity, sedation index, pain index, anxiety index, fidgety index, stress index and the cerebral blood flow (CBF) index which was named of perfusion index. Other variables including indexes of ICU blood gas analysis, which was consisted of variables of blood gas analysis, routine blood test and biochemistry, previous history and prognostic outcome was recorded. Binary logistic regression was used for multivariate analysis.Results Forty-three postoperative patients, who needed intensive care, were included. Eighteen were in delirium group and twenty-five in control group. Excluding the trauma, variables like gender, age, temperature, heart rate, respiratory rate, mean arterial pressure, acute physiology and chronic health evaluationⅡ(APACHEⅡ) score, organ failure, dementia and emergency surgery didn't show any statistical significance between two groups. The trauma in delirious patients increased obviously compared with the control group (33.3% vs. 4.0%,P = 0.031). Except for the brain activity [122.47 (88.62, 154.21) vs. 89.40 (86.27, 115.97),P = 0.034], there were no statistical differences in any other EEG and CBF variables. In ICU blood gas analysis, only pH value (7.43±0.42 vs. 7.47±0.31,P =0.003), chloride concentration [Cl- (mmol/L) 114.66±4.32 vs. 111.90±3.08,P = 0.019], magnesium concentration [Mg2+ (mmol/L) 0.60±0.10 vs. 0.54±0.06,P = 0.035] and blood osmolality [mmol/L 290.10 (284.15, 306.35) vs. 282.70 (280.20, 286.75),P = 0.014] were statistically significant. Compared with control group, the prognostic variables in delirium group such as duration of mechanical ventilation [days 125.0 (49.0, 293.0) vs. 149.5 (32.0, 251.3)], length of stay in ICU [days 216.5 (50.5, 360.8) vs. 190.0 (72.0, 330.5)] and mortality rate (22.2% vs. 24.0%) didn't appear to be statistically significant either (allP > 0.05). It was shown by multivariate logistic regression analysis that pH [odds ratio (OR) = 1.446, 95% confidence interval (95%CI) = 1.116-1.875,P = 0.005] and Cl- (OR= 0.708, 95%CI =0.531-0.945,P = 0.019) were potential risk factors of delirium.Conclusions The brain activity of HXD_Ⅰ may contribute to the clinical diagnose of delirium, but it still remained to be proved further. The pH and Cl- are potential risk factors of delirium.