Study of Predictors of Delirium Incidence in Hospitalized Patients In Intensive Care Units

Delirium is the most common neurological diagnosis among patients in intensive care units. The prevalence of delirium in the ICU patients is high and this is associated with many complications. Thus, by assessment and identifying predictive factors of delirium, its incidence can largely be prevented in intensive care units.
This study aims to determine predictive factors of delirium incidence in patients hospitalized in intensive care units.
This study is a descriptive-analytic study which included all patients admitted to intensive care units (neurology, general and trauma) in one of the training centers, Rasht for three months in 2013 without any symptoms and signs of delirium before hospitalization. There was no age limit for selection of samples. At least 24 hours after admission to ICU, with and without mechanical ventilation, ability to see and hear, Persian language, conscious or semi-conscious, not receiving neuromuscular blockers, coma, history of severe nerve damage (such as acute stroke, dementia, aphasia), chronic renal failure, alcohol abuse and drug abuse were the study inclusion criteria. Patients who used painkiller and sedatives during the study were excluded. Thus, samples were selected gradually considering the inclusion criteria. Based on the preliminary results with 20 samples, the study sample size was estimated 81. The tools used in this study were Richmond Agitation Sedation Scale (RASS) and Confusion Assessment Method for the Intensive Care Unit (CAM-ICU). RASS is the standard tool to assess the level of restlessness and sedation in ICU patients. CAM-ICU examines four main characteristics of delirium; acute change or fluctuations in mental status, lack of concentration, disturbance of consciousness and unorganized thinking.
The first part consisted of demographic characteristics including age, sex, Charleson comorbidity index, Acute Physiology Score (APS) in three ranges of 0-9, 10-14-and ≥15, white blood cell count, serum total bilirubin, days of mechanical ventilation, and days of ICU stay and hospitalization. Charlson comorbidity index had 19 conditions in which the patient is scored based on its potential impact on mortality rate. APS is the biggest part of APATCH (Acute Physiology and Chronic Health Evaluation), obtained from 13 clinical evaluations performed 24 hours after ICU admission in which higher scores is indicative of worsening patient's physical condition. The second part consisted of examining the incidence of delirium in which patients were assessed by RASS in terms of level of consciousness. This tool consisted of 10 items, each representing one level of consciousness (of Combative to Unarousable). To determine RASS, without any interaction, the patient was only observed and if conscious, she was scored 0 to . If the patient was unconscious, his/her name was called loudly and asked to look at the researcher. If the patient responded to call, appropriate score (1-3) was assigned. If there was no response, the patient's shoulder was shaken. If no response was observed, his/her sternum was strongly squeezed and appropriate score (4-5) was given. In case of no consciousness level disorder based on RASS, the subjects were assessed by CAM-ICU for examining delirium which took about 3-5 minutes.
The collected data were analyzed using descriptive and analytical statistics (Fisher and chi-square). For multivariate analysis of the associated factors with delirium, Backward Logistic Regression model was used. P 64.2% of the subjects were male and most of them (48.1%) were hospitalized in the general ward. The majority of samples (49.4%) scored 0-9 in terms of APS. Charlson comorbidity index also showed most cases (29.6%) with 1-2 scores. The mean and SD of patients were 50.95 ± 21.33 years, the maximum tracheal intubation days was 146 days. The maximum duration of ICU stay was 147 days and 150 days in the hospital. Furthermore, the maximum total bilirubin was 4.2 mg dl and the highest number of white blood cells was 29.2000 mg dl. Delirium was observed in 27.2% of samples using CAM-ICU. Delirium distribution was not significant based on qualitative variables; sex, ward and Charlson score whereas APS score which was significant (p=0.048). Distribution of delirium was significant in terms of age (P=0.06). However, it was not statistically significant in terms of the number of mechanical ventilation days, hospitalization and ICU stay, total bilirubin and white blood cell count.
Based on logistic regression model, age, sex, days of tracheal intubation, ICU and hospitalization days, total bilirubin, white blood cell count, and Charlson index were not predictors of delirium. Only APS (considering the range of 0-9 scores as the reference) in two ranges of 10-14 (P Patients with APS points higher than those with lower scores are more likely to develop delirium.