Time-to-positivity in bloodstream infection is not a prognostic marker for mortality: analysis of a prospective multicentre randomised control trial

Link to article at PubMed

Clin Microbiol Infect. 2021 Jun 7:S1198-743X(21)00291-3. doi: 10.1016/j.cmi.2021.05.043. Online ahead of print.

ABSTRACT

OBJECTIVES: Time to positivity (TTP), calculated automatically in modern blood culture systems, is considered a proxy to microbial load and has been suggested a potential prognostic marker in bloodstream infection. In this large, multi-centre, prospectively collected cohort, our primary analysis aimed to quantify the relationship between TTP of monomicrobial blood cultures and mortality.

METHODS: Data from a multi-centre randomised control trial (RAPIDO) in bloodstream infection was analysed. Bloodstream infections were classified into 13 groups/subgroups. The relationship between mortality and TTP was assessed by logistic regression, adjusted for site, organism, and clinical variables; and linear regression applied to examine the association between clinical variables and TTP. Robustness was assessed by sensitivity analysis.

RESULTS: 4,468 participants were included in RAPIDO. After exclusions, 3,462 were analysed, with the most common organisms being coagulase-negative staphylococci (1,072 patients) and E.coli (861 patients). 785 (22.7%) patients died within 28 days. We find no relationship between TTP and mortality for all groups except for Streptococci (Odds ratio (OR) with each hour 0.98, 95% CI 0.96-1.00) and Candida (OR 1.03, 95%CI 1.00-1.05). There was large variability between organisms and sites in TTP. Fever (Geometric Mean Ratio GMR 0.95; 95% CI 0.92-0.99), age (GMR per ten years 1.01, 95% CI 1.00 - 1.02), and neutrophilia were associated with TTP (GMR 1.03; 95% CI 1.02-1.04).

CONCLUSIONS: Time to positivity is not associated with mortality, except in Candida spp (longer times associated with worse outcomes), and possibly in Streptococci (shorter times associated with worse outcomes). There was large variation between median times across centres, limiting external validity.

PMID:34111588 | DOI:10.1016/j.cmi.2021.05.043

Leave a Reply

Your email address will not be published.