One in five intensive care unit (ITU) admissions is due to severe sepsis and according to several studies the incidence of sepsis in the general population is increasing. However, the limitations of these studies include variations in the definition of sepsis and the possible inaccuracy of discharge diagnosis codes.

This study prospectively screened all medical and surgical admissions up to hospital discharge for the presence of sepsis. Three hospitals were involved over a 4 month period, each serving a distinct geographic area of known population size. Sepsis, severe sepsis and septic shock were defined as per the American College of Chest Physicians/Society of Critical Care Medicine Consensus Conference definitions.

Of 15 852 adults admitted to hospital, 702 were identified with sepsis, giving a cumulative incidence of 367 cases per 100 000 adults per year. Of these cases, 83% were caused by community-acquired infection; of which 56% were pulmonary in origin. Of the 702 patients, 199 developed severe sepsis; however, only 84 patients (12%) were admitted to the ITU. Overall mortality rates for sepsis, severe sepsis and septic shock were 6.7%, 20.7% and 45.7%, respectively. The mortality for severe sepsis in patients not admitted to the ITU was 26%.

Underutilisation of ITUs may account for the high mortality rate in those with severe sepsis; however, nearly half of the patients not admitted to an ITU had “Do Not Resuscitate” decisions, making the proportion of patients eligible for ITU transfer smaller. Further studies are needed to look the management of patients with sepsis.