Time to blood culture positivity in Staphylococcus aureus bacteraemia to determine risk of infective endocarditis.

OBJECTIVES

Patients with Staphylococcus aureus bacteraemia (SAB) at risk for infective endocarditis (IE) need to be identified because they should undergo echocardiography. We validated previous scoring systems for IE risk determination and evaluated whether time to blood culture positivity (TTP) could improve scoring systems.

METHODS

This retrospective population-based study included adults with SAB in 2016 in a derivation cohort and those from 2017 in a validation cohort. TTP was compared between patients with and without IE. A new score including TTP was constructed using a least absolute shrinkage selection operator. The new POSITIVE score was compared to the previously described PREDICT and VIRSTA scores.

RESULTS

A total of 465 episodes with SAB were included in the derivation cohort, of which 38 (8.2%) represented IE. Median (interquartile range) TTP was significantly shorter in episodes with IE, at 8.7 (7.7-10.6) hours compared to those without, at 13.3 (10.5-16.5) hours. When using a cutoff at 13 hours, TTP had a sensitivity of 100% (95% confidence interval (CI), 91-100) and specificity of 52% (95% CI, 47-57) for IE. The POSITIVE score included TTP, intravenous drug use, embolizations and presence of preexisting heart conditions. It had a sensitivity of 93% (95% CI, 76-99) and a specificity of 70% (95% CI, 66-74) in the validation cohort. The performance of POSITIVE was superior to PREDICT, and the specificity was higher than that of VIRSTA.

CONCLUSIONS

TTP, either by itself or as part of the POSITIVE score, can be used to identify patients with SAB at low risk for IE. Further validation is needed because TTP is sensitive to several external factors.