Previous experimental and observational studies have suggested that the use of antibiotics with anaerobic cover for empiric treatment of bacterial sepsis has worse outcomes, compared with anaerobe-sparing antibiotics. A recent clinical trial showed no difference in outcomes at 14 days. 
Rishi Chanderraj and colleagues conducted a retrospective study at a single hospital which included a time period in which there was a shortage in supply of piperacillin-tazobactam (piptaz). Patients that were treated with vancomycin plus either piptaz or cefepime for presumed sepsis were included in the study. 
The primary endpoint was 90-day mortality. Piptaz was associated with greater mortality and duration of organ dysfunction than cefepime. It is suggested that including antibacterials with anaerobic cover for empiric treatment of sepsis could be harmful. One hypothesis for this is that anaerobic gut bacteria may have a protective effect in certain conditions.
Read the full paper on JAMA Network.