RESUMO
OBJECTIVE: To assess the cost effectiveness of an initial ICU admissionforpatients with severe sepsis or those in septic shock following the initial resuscitation in the emergency department. MATERIAL AND METHOD: Mortality data was generated through retrospective data obtained from 1,048 adult patients with severe sepsis or in septic shock from one tertiary care and eight community hospitals in Phitsanulok during the period of October 2010 to September 2011. These patients were categorized into two groups; as either admitted from the emergency department directly to the ICU (stated as an immediate ICU admission) or admitted from the emergency department to the general hospital ward due to an unavailability of lCU beds (stated as a delayed ICU admission). The overall direct costs and characteristics were simulated from a second group of 994 adult patients, admitted a year later from selected data by the ICD-10 codes [International Classification of Diseases, 10th edition] with the same conditions of severe sepsis and septic shock (September 2011 through September 2012), as there was no collection of costs and characteristics during the first period (October 2010 through September 2011). A decision tree model and an incremental cost-effectiveness ratio (ICER) were used for the analyses of the cost-effectiveness. RESULTS: There were no significant differences in either the mean ages or lengths of stay between both groups. All-cause mortality rates have shown an incidence of 22.2% for the immediate ICU admission group and an incidence of 46.3% in the delayed ICUadmission group (odds ratio for the immediate ICU admission group was 0. 479 with a 95% confidence interval, 0.376-0.611). Total costs (mean, 95% CI) of the immediate ICUadmission group [37,194 baht (32,389-44,926)] were higher than had been seen in the delayed ICU admission group [26,275 (24,300-27,936)]. Incremental cost was 10,919 baht. ICER for the immediate ICU admission group was 45,307 baht per life saved. CONCLUSION: Immediate ICU admission for patients with severe sepsis or in septic shock following the initial resuscitation in the emergency department has shown a satisfactory cost-effectiveness profile in low-to-middle income countries.