Characteristics and outcomes of patients with culture negative septic shock compared with patients with culture positive septic shock: a retrospective cohort study

Background

Previous studies have identified that nearly 30% of patients with severe sepsis and septic shock lack a definitive microbial etiology. The characteristics and outcomes of culture negative septic shock are not well defined despite large epidemiologic studies on septic shock

Materials and methods

Retrospective nested cohort study of 2,651 patients with culture-negative septic shock and 6,019 culture-positive septic shock patients derived from a trinational, 8,760-patient database of patients with septic shock between 1989 and 2008.

Results

In total, 30.6% of cases of septic shock cases were identified as culture-negative within the database. Patients with culture-negative septic shock (CNSS) experienced similar ICU mortality as did those with culture-positive septic shock (CPSS) (41.7% vs. 40.5% P = 0.276) and identical overall hospital mortality (51.9% vs. 51.9% P = 0.976). Severity of illness was similar between CNSS and CPSS (median APACHE II 25 (IQR 6 to 54) vs. 25 (IQR 4 to 70) respectively). Initial and 6-hour lactate levels were also similar among CNSS and CPSS patients (mean 4.4 vs. 4.1, P = 0.237 and mean 4.0 vs. 4.1, P = 0.221 respectively). Interestingly CNSS patients were significantly more likely to be hypothermic than CPSS patients (temperature <36°C 18.9% vs. 15.3%, P < 0.0001). CNSS patients presented significantly more often from the community (63.3% vs. 58.0%, P < 0.0001), where patients with CPSS presented more often with nosocomial infections (36.7% vs. 42.0%, P < 0.0001). Gastrointestinal and respiratory tract infections were the predominant sources of infection in both groups. However, CNSS patients with respiratory tract infections experienced lower mortality than their CPSS counterparts (49.6% vs. 56.3%, P = 0.008) but similar mortality rates with gastrointestinal infections (61.0% vs. 58.2%, P = 0.289) (Tables 1 and 2).

Table 1 Comparison of variables of culture-positive and culture-negative septic shock

Table 2 Comparison of major sites of infection

Similar to our previous findings, we identified by the second hour after onset of persistent/recurrent hypotension that the in-hospital mortality rate was significantly increased relative to receiving therapy within the first hour (odds ratio, 1.62; 95% CI, 1.21 to 2.15; P < 0.001) in the CPSS group. Following increasing delays in the administration of appropriate antimicrobial therapy over the first 6 hours after the onset of hypotension, patients in both groups experienced nearly congruent, significant increases in hospital mortality; at 6 hours the CNSS group (odds ratio, 2.87; 95% CI, 1.72 to 4.78; P < 0.0001) and the CPSS group (odds ratio, 3.44; 95% CI, 2.17 to 5.48; P < 0.0001) (Figure 1). Survival differences between these time intervals are not significantly different in patients with CNSS and CPSS.

Figure 1

Odds ratio of death by antibiotic delay in culture-positive and culture-negative septic shock.

Conclusions

Patients with CNSS behave similarly to CPSS patients in nearly all respects. As with bacterial septic shock, early appropriate antimicrobial therapy appears to improve mortality. Earlier recognition of infection is the most obvious effective strategy to improve hospital survival. Optimal duration of therapy is not well defined among patients with CNSS. In addition to early, appropriate antimicrobial therapy, use of de-escalation strategies such as serial procalcitonin levels may be useful to determine the length of empiric broad-spectrum antimicrobial use in this population.