Determinants of delirium among patients in critical care units at Kenyatta National Hospital

Abstract

Background: Delirium is common in patients admitted to critical care units. It is associated with morbidity and mortality, and high hospital costs. The prevalence has been found to range from 20-83% among critically ill patients. Even though delirium prevalence and factors have been described in the developed world, there is a paucity of data from low-resource settings such as Kenya. The aim of this study therefore is to determine the factors associated with the development of delirium in patients admitted to critical care units. Objective: Assess the prevalence, risk factors, and precipitating factors for delirium among critical care patients admitted in the main or medical ICU at Kenyatta National Hospital (KNH). Methods and Materials: A cross-sectional study of 81 patients admitted at the main or medical ICU of the KNH was done in 2020. The principal investigator administered consent to the next of kin and a study questionnaire was used to record demographic and medical data of participants from hospital records. The Richmond Agitation Sedation Scale (RASS) was used to evaluate sedation and agitation status and the Confusion Assessment Method for the Intensive Care Unit (CAM-ICU) was used to evaluate delirium. The CAM-ICU assessments were done once. Data were uploaded into a Statistical Package for Social Scientists (SPSS) spreadsheet and cleaned. Categorical data were summarized into proportions and continuous data as means with standard deviations. Clinical characteristics were evaluated by using frequencies and percentages and risk factors for delirium were evaluated using a chi-square test and logistic regression. Statistical data analysis was done at a 95% confidence level. Results: Eighty-one patients 50 (61.7%) males and 31 (38.3%) females were recruited with a mean age of 39.67±12.09 years. A majority were self-employed 34 (42.0%), had a tertiary 60 (74.1%) education, and did not drink alcohol 54 (66.7%). A majority did not smoke cigarettes 69 (85.2%), had co-morbidity 47 (58.0%), mostly hypertension 19 (40.4%), while the mean number of medications was 10.22±3.58. Most were sedated 53 (65.4%) mostly continuously 48 (90.6%) with Remifentanyl 19 (35.8%), while 57 (70.4%) were physically restrained. The average length of hospital stay was 13.84±10.82 days. The prevalence of delirium was 34.6% (24.3-45.9%). Delirious patients were significantly older (43.68±12.72 years) than non-delirious patients (37.45±10.82), P=0.03. The odds of delirium was 12.58 fold (95% CI=3.87-32.68) higher among patients who consumed alcohol compared to those who did not, P<0.01. The odds of delirium was 6.14 fold (95% CI = 0.66-19.95) higher with cigarette smoking, P=0.01 and 7.44 fold (95% CI=1.97-25.13) higher under sedation, P<0.01, while the odds of delirium was 9.23 fold (95% CI=2.19-41.84) higher under physical restraint, P<0.01. After multivariable analysis, the adjusted odds for delirium was 60.15 fold (95% CI= 2.89-1248.1), higher among patients who consumed alcohol, P=0.008, and 22.85 fold (95% CI=1.07-487.8) higher among patients who were physically restrained, P=0.045. Metabolic conditions as the commonest precipitating factor 25 (30.9%) in the ICUs followed by infections 19 (23.5%). Conclusion: Delirium is a common health complication at the Kenyatta National medical and main ICUs with approximately 34.6% of patients affected. Increasing age, cigarette smoking, sedation, and physical restraint, were major risk factors for delirium but only alcohol use and physical restraint could be used to predict the occurrence of delirium in the population studied.