Time to positivity of Corynebacterium in blood culture: Characteristics and diagnostic performance.

ABSTRACT

The presence of Corynebacterium in blood samples can indicate true bacteremia or contamination, thus complicating the diagnosis of true bacteremia. We aimed to evaluate the usefulness of time to positivity (TTP) in diagnosing true bacteremia and contamination in cases where Corynebacterium was isolated from blood samples. We compared the TTP of the true-bacteremia group (n = 77) with that of the contamination group (n = 88). For the true-bacteremia cases that had only one set of positive blood cultures (n = 14), considering clinical and bacteriological data, additional cultures were performed on blood or other specimens. The same Corynebacterium spp. as in blood were isolated from these specimens. Receiver operating characteristic curves were generated, and the sensitivity and specificity of TTP were calculated for diagnosing true bacteremia. The median TTP of the true-bacteremia group (26.8 h) was shorter than that of the contamination group (43.3 h) (P < 0.0001). When considering TTP ≤ 25.0 h as true bacteremia, the sensitivity and specificity were 44.2% and 95.5%, respectively. Moreover, when considering TTP ≤ 69.4 h as true bacteremia, the sensitivity and specificity were 96.1% and 20.5%, respectively. Among the true-bacteremia groups with one set of positive blood cultures (n = 14), no case exhibited a TTP > 69.4 h. Only three cases showed TTP ≤ 25.0 h in the true-bacteremia group with one set of positive blood cultures. TTP > 69.4 h is likely to indicate contamination and may be useful to exclude true bacteremia in cases with one set of positive blood cultures. Meanwhile, diagnosing true bacteremia using the threshold of TTP 25.0 h would be difficult. Therefore, the clinical and bacteriological data are important for diagnosing bacteremia, especially in cases with TTP ≤ 69.4 h.