Early versus delayed IV-to-oral antibiotic transition: clinical deterioration after oral switch and variation across regimen pathways
Résumé fourni par la source
This retrospective cohort study aims to assess associations of delayed Intravenous (IV)-to-Oral switch after 72-hours versus early switch within 72-hours with switch failure within 7-days and with adjusted post switch clinical outcomes in hospitalized adults. Adult admissions with an IV-antibiotic regimen followed by an oral regimen in the same hospitalization were included. Early-switch occurred within 72-hours of admission and delayed-switch occurred after 72-hours. Switch failure within 7-days was defined as IV restart/antibiotic escalation within 7-days after the switch. Adjusted models estimated associations with post switch length of stay (LOS), in-hospital mortality, ICU transfer and vasopressor initiation within 7-days. The cohort included 3898 admissions with 2002 delayed-switches and 1896 early switches. Switch failure within 7-days occurred in 14.7% after delayed switching and 6.4% after early-switching. Delayed-switching was associated with longer post switch LOS with adjusted ratio 1.94 (1.70-to-2.22) and higher odds of in-hospital mortality with aOR = 2.29 (1.21-to-4.34). Vasopressor initiation within 7-days was higher with delayed switching with aOR = 9.68 (2.28-to-41.04). ICU transfer within 7-days didn’t show a clear adjusted difference. Delayed-switching was also associated with higher odds of IV restart within 7-days with aOR = 2.24 (1.79-to-2.81). Delayed switching after 72 hours was associated with more 7-day events and worse adjusted outcomes. However, switch timing is a clinician-, policy-, and resource-dependent surrogate for underlying clinical status rather than a direct measure of disease severity or a cause of deterioration. Clinical interpretation should therefore rely on direct measures of physiologic stability, infection severity, source control, microbiology, and oral-treatment feasibility.
Ce résumé expose les affirmations des auteurs. BNTIC ne l’interprète pas comme une validation indépendante des résultats.