With the release of the 2005 IDSA/ATS pneumonia guidelines healthcare providers were advised to consider numerous patient risk factors for multidrug resistant organisms. Furthermore, these guidelines encouraged empiric treatment with two agents active against pseudomonas, based on the assumption that combination therapy increases treatment success. An evaluation of our institutions respiratory cultures revealed that levofloxacin covers only 2% of pseudomonas isolates resistant to piperacillin-tazobactam, and no isolates resistant to cefepime or meropenem.
Our primary objective was to correlate this evaluations findings with patient outcomes by comparing mortality rates experienced by patients receiving levofloxacin plus an antipseudomonal beta-lactam versus those receiving beta-lactam alone for the empiric treatment of pneumonia. Secondary objectives were to identify between group differences in length of stay, 30-day readmission, duration of mechanical ventilation, and occurrence of Clostridium difficile infection.
This single-center, retrospective chart review was conducted by evaluating records of patients with a discharge diagnosis of pneumonia from January 1, 2014 to September 1, 2016. Patients were included if they received at least 48 hours of empiric pneumonia treatment with an IV anti-MRSA agent plus an antipseudomonal beta-lactam. Patient enrollment is displayed in figure 1.
Of 1897 patient screened, 228 patients were included. There were 146 patients who received monotherapy with an antipseudomonal beta-lactam and 82 patients who additionally received levofloxacin. Baseline characteristics were comparable between groups. The mean age was 68.8 years, 51% were male, 45.6% had a diagnosis of COPD at baseline, and the average qSOFA score upon admission was 0.66. There was no significant difference in mortality (p=0.438), nor any secondary objective. No significant difference in the duration of therapies prior to de-escalation was observed (p=0.395).
Addition of levofloxacin to beta-lactam therapy did not impact clinical outcomes in this population. Further analysis of site-specific data is warranted, as the results of this study may not be generalizable.
Figure 1: Patient enrollment.
C. Pleiman, None
S. Burdette, None