Distribution and clinical predictors of time to positivity (TTP) of blood cultures (BC): tools for a shorter duration of empirical antimicrobial therapy for suspected sepsis

Background:

Currently, empiric antimicrobial therapy for sepsis is de-escalated if BC drawn before start of treatment have remained negative for 48 to 72 hours. However, with current instrument-based continuous monitoring blood culture systems, TTP has been reduced substantially. The study objective was to determine the distribution of TTP in patients with sepsis and to identify clinical predictors of short (<16h) and long (>24h) TTP.

Methods:

All adults with proven bacteraemia with microorganisms targeted by empiric sepsis therapy in the Leiden University Medical Center during 1 year, were included. BC were incubated using the BACTEC FX, Becton Dickinson B.V, the Netherlands. TTP was defined as the time between sampling and a positive signal. Patient characteristics and clinical data were obtained from the electronic charts. Uni- and multivariate regression analyses were performed to determine predictors of TTP and repeated separately for patients without antibiotic pre-treatment.

Results:

A total of 318 patients, 58.5% male, mean age 62.8 years (range 18-93), were included. The median TTP was 14.7 h (range 4.9-116.1). In 61.3% and 89.1% the TTP was< 16h and <24h, respectively. At the time of BC sampling 26.6% of patients were on antibiotic treatment, which affected TTP (adjusted OR for TTP >24h 2.10, 1.25-3.52, p<0.01, figure 1). Figure 2 depicts the distribution of TTP per microorganism. For patients without pre-treatment, organ transplantation (RR 2.55, 1.04-6.25, p=0.04) and a urinary tract focus (RR 2.24, 1.03-5.67, p=0.038) were associated with a TTP >24h. Noted to be ill on physical examination (RR 1.24, 1.04-1.5, p=0.03) and the presence of neurologic symptoms (RR 1.32,1.09-1.57, p=0.02) were associated with short TTP. In the multivariate analyses, only pre-treatment was independently associated with long TTP.

Conclusion:

In patients with suspected sepsis, there is a very low absolute risk of a positive BC after 24h of incubation. Pre-treatment, transplantation and urinary tract infection were associated with TTP >24h. Considering toxicity of broad-spectrum antibiotics, selection of resistance and costs, further studies that investigate limiting empiric antimicrobial therapy for suspected sepsis to 24h, in selected patients, are warranted.