Incidence and mortality of post-operative sepsis in New South Wales, Australia, 2002-2009

Objective: To describe the incidence and mortality of postoperative sepsis in New South Wales, Australia. Design, setting and participants: A retrospective study of adult elective surgical admissions (n = 229 918) in 82 public acute care hospitals in NSW, 2002-2009. Main outcome measures: Changes in the incidence rate of post-operative sepsis and sepsis-related mortality. Results: Although the mortality rate among patients with sepsis decreased from 26.9% in 2002 to 20.2% in 2009 (P = 0.006 for adjusted trend), the incidence rate of sepsis increased from 12.7 to 15.8 per 1000 admissions (adjusted rate ratio [RR], 1.23; 95% CI, 1.06-1.42). Thus, the incidence rate of sepsis-related deaths remained unchanged (3.4 v 3.2 per 1000 admissions; adjusted RR, 0.90; 95% CI, 0.67-1.22), as did deaths from sepsis as a proportion of all elective surgical deaths (P = 0.96 for adjusted trend). The incidence rate of infections without a specified organism identified increased; was twice the rate of gram-positive infections (8.5 v 4.1 per 1000 admissions, P < 0.001); and was three times the rate of gram-negative infections (8.5 v 2.7 per 1000 admissions, P < 0.001). Also, compared with patients with gram-positive infections, patients with an unspecified infection were more likely to die (adjusted RR, 1.33; 95% CI, 1.13-1.57), but patients with gram-negative infections and mixed infections had a similar likelihood of death from their infection. Conclusion: Over 8 years, the mortality from postoperative sepsis decreased, but its incidence rate increased, resulting in a lack of improvement in the incidence rate of sepsis-related deaths. The increasing incidence of postoperative sepsis and the poor record of identification of causative organisms remain a significant public health challenge.