Abstract
To compare prognoses and adverse events between bacteraemic patients in the emergency department (ED) who received an early antimicrobial IV-to-PO switch and those treated with late or no IV-to-PO switch, an 8-year multicentre cohort consisting of adults with community-onset bacteraemia was conducted. The clinical characteristics and outcomes were compared in matched cohorts by the closest propensity score calculated based on the independent determinants of 30-day mortality identified by the multivariate regression model. Of the 6664 hospitalised patients who received no or late IV-to-PO switch, 2410 were appropriately matched with 482 patients treated with early IV-to-PO switch and discharged from the ED. There were no significant differences between the two matched groups in their baseline characteristics, including the patient demographics, severity and types of comorbidities, severity and sources of bacteraemia, and the 15- and 30-day mortality rates. Notably, in addition to the shorter lengths of intravenous antimicrobial administration and hospital stay, less phlebitis and lower antimicrobial costs were observed in patients who received an early IV-to-PO switch. Similarity was observed in the clinical failure rates between the two groups. Furthermore, the inappropriate administration of empirical antibiotics and inadequate source control were identified as the only independent determinants of the post-switch 30-day crude mortality in patients who received an early IV-to-PO switch. In conclusion, for less critically ill adults with community-onset bacteraemia who received appropriate empirical antimicrobial therapy and adequate source control, an early IV-to-PO switch might be safe and cost-effective after a short course of intravenous antimicrobial therapy.