Objective: To investigate the numbers, clinical characteristics, resource use, and outcomes of admissions who met precise clinical and physiologic criteria for severe sepsis (as defined in the PROWESS trial) in the first 24 hrs in the intensive care unit.
Design: Observational cohort study, with retrospective analysis of prospectively collected data.
Setting: Ninety-one adult general intensive care units in England, Wales, and Northern Ireland between 1995 and 2000.
Patients: Patients were 56,673 adult admissions.
Interventions: None.
Measurements and Main Results: We found that 27.1% of adult intensive care unit admissions met severe sepsis criteria in the first 24 hrs in the intensive care unit. Most were nonsurgical (67%), and the most common organ system dysfunctions were seen in the cardiovascular (88%) and respiratory (81%) systems. Modeling the data for England and Wales for 1997 suggested that 51 (95% confidence interval, 46-58) per 100,000 population per year were admitted to intensive care units and met severe sepsis criteria in the first 24 hrs.
Of the intensive care unit admissions who met severe sepsis criteria in the first 24 hrs, 35% died before intensive care unit discharge and 47% died during their hospital stay. Hospital mortality rate ranged from 17% in the 16-19 age group to 64% in those >85 yrs. In England and Wales in 1997, an estimated 24 (95% confidence interval, 21-28) per 100,000 population per year died after intensive care unit admissions with severe sepsis in the first 24 hrs.
For intensive care unit admissions who met severe sepsis criteria in the first 24 hrs, median intensive care unit length of stay was 3.56 days (interquartile range, 1.50-9.32) and median hospital length of stay was 18 days (interquartile range, 8-36 days). These admissions used 45% of the intensive care unit and 33% of the hospital bed days used by all intensive care unit admissions.
Conclusions: Severe sepsis is common and presents a major challenge for clinicians, managers, and healthcare policymakers. Intensive care unit admissions meeting severe sepsis criteria have a high mortality rate and high resource use.
Literature reviews have reported a wide variation in the reported prevalence and incidence of severe sepsis and in its associated mortality rate (1, 2). This may reflect differences in the populations studied and the definitions of severe sepsis used. Patient populations studied have ranged from those in a single intensive care unit (ICU)/hospital (3-6) to large studies including >100 centers (7, 8). Some studies have investigated all hospital admissions (4, 9, 10), whereas others have been restricted to specific hospital areas (3, 5-7, 11).
The definitions used for severe sepsis also have varied. In 1992, the American College of Chest Physicians/Society of Critical Care Medicine (ACCP/SCCM) consensus conference (12) distinguished between the systemic inflammatory response syndrome (SIRS), sepsis, and severe sepsis (Appendix 1). Although the criteria for SIRS were explicitly physiologic, organ system dysfunctions were not defined in terms of physiologic derangement and guidance was provided neither for how to judge whether physiologic abnormalities were a consequence of infection nor how the baseline physiologic state should be defined (13, 14). The time frame within which the observed derangements had to occur also was not specified. Consequently, epidemiologic studies have used a variety of definitions for organ system dysfunctions (3, 4, 6, 7, 15).
The consensus conference definition of sepsis (and hence severe sepsis) depended on a confirmed infectious process being present. Some have interpreted this to mean microbiological confirmation (7, 10), although the frequent lack of microbiological results during the acute phase means that the diagnosis is only possible retrospectively. Others have modified the ACCP/SCCM definitions to make them more clinically applicable by either including clinically suspected infection (4, 6, 7) or using proxies such as the start of antibiotics (5) or the taking of blood cultures (16).
For all these reasons, the interpretation of published epidemiologic studies of severe sepsis presents a major challenge. As there have been no effective interventions specifically for patients with severe sepsis until recently, this lack of rigorous data has had few clinical consequences. However, the PROWESS trial (17) suggests that recombinant human activated protein C is effective in reducing all-cause 28-day mortality rate for patients with severe sepsis. One of the strengths of the trial was that it used precise physiologic definitions for both SIRS and organ system dysfunction. An understanding of the numbers, clinical characteristics, resource use, and outcomes of patients satisfying the definition for severe sepsis used in that trial is now, therefore, of practical importance to all ICU clinicians and policymakers.
The primary objective of this study was to investigate the numbers, clinical characteristics, resource use, and outcomes of adult ICU admissions who met precise clinical and physiologic criteria for severe sepsis in the first 24 hrs following admission to an ICU in England, Wales, and Northern Ireland. These definitions were matched as closely as possible to those used in the PROWESS trial (17).