Antibiotics for hospital pneumonia can be de-escalated

Nosocomial pneumonia is commonly caused by multi-resistant organisms, such as Pseudomonas aeruginosa and methicillin-resistant Staphylococcus aureus. Guidelines from several organizations recommend empirical coverage of these bacteria, and these antibiotics are often continued for 7-10 days when cultures are not revealing.

Although appropriate antibiotics, given early, are paramount in infection management, the risk of bacterial resistance increases with the use of inadequately broad antimicrobial agents.

The researchers in St. Louis retrospectively analyzed 329 patients from a single institution who were identified with culture-negative nosocomial pneumonia.

Almost all patients received anti-pseudomonal agents and all received anti-MRSA agents. Ninety-two patients had decreased MRSA coverage as of day 4 of treatment; mean disease severity scores were similar in the non-scalp and non-scalp groups.

Twenty-eight-day mortality was similar among patients who received non-scalating and scale-free MRSA therapy.

Patients whose therapy was decreased averaged 3 days less in the intensive care unit and 5 days less in the hospital than patients whose antibiotic therapy was not decreased.


In this retrospective cohort study, scaling down of an empirical anti-MRSA agent in culture-negative nosocomial pneumonia did not affect 28-day mortality. De-scaling of the anti-MRSA agent was also not associated with worse outcomes, such as hospital mortality, total hospital stay and ICU, and treatment failure.

In addition, this study also found that patients who had their empirical anti-MRSA agent de-escalated had a shorter hospital stay at 5 days.

The shorter duration of hospital stay is unlikely to be an indication of disease severity because patient populations adjusted well at baseline.

Instead, the shorter length of stay may be an indicator that patients staying in hospital for additional days to receive intravenous antibiotics, or it could be due to the longer length of stay associated with the increased rate of ARTIs. It is more likely that the shorter hospital stay was associated with the longer duration of intravenous antibiotic administration.

Reducing antimicrobials has been shown in the literature to reduce drug-related adverse effects and antimicrobial costs.

In our study, a secondary safety finding focused on the incidence of ARTIs in the two groups. Patients who were not de-escalated had higher rates of ARTIs.

Many factors that may contribute to AKI in the critically ill population were not considered in this secondary outcome analysis. However, one explanation for this result is potentially associated with the use of vancomycin, which was the most widely used anti-MRSA agent in this study.

Previous studies have cited prolonged duration of vancomycin as a risk factor for nephrotoxicity. In this study, patients who were de-escalated had 5 days less anti-MRSA coverage than patients who continued on their anti-MRSA agent.

Although no formal economic analysis was performed for this study, cost savings could be linked to reduced antimicrobial use, length of ICU and hospital stay, and less adverse management of drug effects.

In conclusion, the progressive reduction of an empirical anti-MRSA agent in culture-negative nosocomial pneumonia did not affect 28-day mortality and may have led to a lower rate of ARTIs.

Other studies may delineate the association between de-escalation and AKI for vancomycin and linezolid.

Although scores for disease severity were similar in the small and non-small groups, clinicians probably responded to unmeasured differences in baseline characteristics (or in clinical course during the first few days) when they decided to stop antibiotics early.

Of course, if doctors were treating on instinct, they were good, as patients did well with less aggressive coverage. Starting antibiotics is the easy part: stopping them is often more difficult. This study provides a degree of comfort in doing the latter.