This is a retrospective cohort study of patients with a positive quantitative respiratory culture that were treated for bacterial pneumonia in an SICU over a two year period. Those treated with IV antibiotics exclusively were compared with those initiated on or transitioned to PO antibiotics within 4 days of antibiotic administration. Only 19% of evaluated patients (124 of 647) met inclusion criteria, with only 30 patients being in the PO group. Patients <18 and >88 years of age were excluded, as well as prisoners, pregnant patients, and those with pneumonia secondary to acinetobacter baumannii, pseudomonas, or ESBL producing bacteria, concurrent bacteremia, patients on TPN, and those that had care withdrawn or were transferred out of the SICU during treatment. >90% of patients in both arms were mechanically ventilated.
There was more MSSA and Serratia in the PO group, but more polymicrobial infections in the PO group (43% vs 26%). Duration of antibiotics were similar between the two groups, with a median of 8 days for the PO group vs 9 for the IV group.
There was no difference in clinical improvement (86.7% vs 72.3% PO vs IV), recurrence (10% vs 12.8% PO vs IV), duration of mechanical ventilation, ICU or hospital length of stay, or all-cause mortality. Antibiotic costs were lower in the PO group ($1042 vs $697), as was infection related costs ($20,776 vs $17,381).
The authors hypothesize cost savings from decreased cost for PO antibiotics, prep/administration time, tubing costs, as well as a decrease in repeated imaging to verify lines (PICCs, central lines), which may be able to be reduced if there is not a need for continued IV antibiotics in some cases.
The authors note several limitations of this single center study- the numbers are small, they excluded folks with bugs notorious for multidrug resistance, and they also lump in “PO only” patients with those transitioned from IV to PO within 4 days. IDSA guidelines for non-ICU community acquired pneumonia recommend transition to PO antibiotics within 24 hours of patient improvement, though for ICU patients, they are seemingly vague on the issue (“highly bioavailable agents, such as the quinolones and linezolid, may be easily switched to oral therapy in such patients [Level II]”).
While the critically ill and may have hampered ability to process PO medications, early transition to PO antibiotics is probably reasonable based on this study – and certainly a larger scale study would be warranted may tease out additional significance that this study trended towards but was not powered for- specifically, decreased recurrence of infection in the PO group, better improvement in the PO group, and decreased duration of antibiotic therapy. Likewise, in a larger study, I would not be at all surprised if length of ICU & ward stay was improved in the PO arm.