The sepsis syndromes in the emergency department
- N Shapiro1
© BioMed Central Ltd. 2004
Published: 15 March 2004
The ACCP/SCCM established uniform criteria for defining the sepsis syndromes: systemic inflammatory response syndrome (SIRS), sepsis, severe sepsis, and septic shock. Subsequent studies showed increasing mortality with each successive syndrome. The utility of these definitions and mortality rates are not known for an emergency department (ED) population. Furthermore, the mortality rates at 1 year have not been previously reported.
The sepsis syndrome definitions will predict increasing mortality with each syndrome for both 28-day inhospital mortality and 1-year mortality.
A prospective, cohort study of consecutive adult patients (age > 18 years) at risk for infection (indicated by physician ordering a blood culture) in the ED of an urban university hospital. The study period was 1 February 2000–1 February 2001. ED charts were prospectively reviewed. The following definitions were applied: SIRS/sepsis, the presence of two or more of hypothermia (T< 36°C) or hyperthermia (T > 38°C), tachycardia (> 90 beats/min), tachypnnea (> 20 breaths/min) or hypoxia (pox < 90%), and leukocytosis (> 12,000) or bandemia (> 10%); severe sepsis, sepsis plus organ dysfunction; septic shock, sepsis plus hypotension (systolic blood pressure < 90) after a 10–20 cm3/kg fluid challenge. The outcomes were 28-day inhospital mortality and 1-year mortality. The 95% confidence interval [CI] was calculated where appropriate, and Duncan's Multiple Range Test for comparison of mortality rates between groups (alpha = 0.05).
Of 3926 eligible patients, 3763 (97%) were enrolled in the study. The overall 28-day inhospital mortality rate was 4.6%, and the 1-year mortality was 24% for all patients. Of these, 1850 (49%) had no SIRS criteria with 2.6% (95% CI = 1.9–3.3%) 28-day mortality, 715 (19%) patients had SIRS/sepsis with 1.8% (0.8–2.8%) 28-day mortality, 1144 (30%) had severe sepsis with 8.5% (6.9–10%) 28-day mortality, and 54 patients had septic shock with 28% (16–40%) 28-day mortality. The 1-year mortality rates were: no SIRS = 21% (19–23%), SIRS/sepsis = 11% (8–13%), severe sepsis = 36% (33–39%), and septic shock = 57% (44–71%). Duncan's test showed no difference between the no SIRS and SIRS/sepsis groups for either 28-day inhospital mortality or 1-year mortality. The severe sepsis group and septic shock group had a sequential increase in both mortality outcomes that were significantly higher than each group (P < 0.05).
In ED patients at risk for infection, patients with severe sepsis, and septic shock are at increasing risk of short-term and long-term mortality, while patients with SIRS/sepsis did not demonstrate significantly higher mortality rates than the no SIRS group. The mortality rates in this ED population were lower than those reported in previous ICU populations.