Sepsis incidence and outcome: contrasting the intensive care unit with the hospital ward.

OBJECTIVE: To describe the outcome of patients with sepsis according to location on a ward or in an intensive care unit.

DESIGN: Prospective multicentered observational study.

SETTING: Three academic hospitals in Madrid, Spain.

PATIENTS: Consecutive patients with sepsis admitted to participating hospitals from March 1 to June 30, 2003.

INTERVENTIONS: None.

MEASUREMENTS AND MAIN RESULTS: During the study period, 15,852 patients >18 yrs of age were admitted. Sepsis was identified in 702 patients, giving an estimated cumulative incidence rate of 367 cases per 100,000 adult area residents per year and a cumulative incidence rate among patients admitted to the hospital of 4.4%. Most septic patients had a community-acquired infection (71%). Severe sepsis developed in 199 patients (incidence rate, 104 cases per 100,000 adult area residents per year), and 59 patients developed septic shock (incidence rate, 31 cases per 100,000 adult area residents per year). Most of the patients met the criteria for severe sepsis or septic shock on the same day that they would have qualified for the septic status one step down the scale. In the other patients, the median time between sepsis and severe sepsis was 2 days (interquartile range, 2-5) and between severe sepsis and septic shock was 3 days (interquartile range, 1-4). Only 32% of severe sepsis patients received intensive care. The hospital mortality for all septic patients was 12.8%; for severe sepsis, 20.7%; and for septic shock, 45.7%.

CONCLUSIONS: This study shows the high incidence of sepsis in a general population of patients admitted to hospital. A significant proportion of patients with severe sepsis are not transferred to the intensive care unit.