Customize your JAMA Network experience by selecting one or more topics from the list below.
The prevailing practice in many hospitals is to give immediate, often broad-spectrum, antibiotics to all hospitalized patients with possible infections. Two interlinked factors drive this practice: (1) clinicians fear that any delay in appropriate antibiotics may increase patients’ risk for worse outcomes including death, and (2) it is uncomfortable to withhold antibiotics from a patient who may have an infection, even if the likelihood of infection is low; doing something feels more responsive, responsible, and patient-centric than doing nothing. The net result, however, is widespread use of antibiotics, much of which is unnecessary. Overprescribing is particularly pronounced in patients with pneumonia. Study after study has documented high rates of pneumonia overdiagnosis, suggesting a rush to treat despite equivocal evidence of disease. Up to half of hospitalized patients treated for pneumonia may not actually have pneumonia.1,2
Adding insult to injury, hospitalized patients with suspected pneumonia are often treated with broad-spectrum antibiotics, a trend catalyzed by the 2005 American Thoracic Society and Infectious Diseases Society of America pneumonia treatment guidelines,3 which helped popularize the concept of health care-associated pneumonia. The guidelines recommended treating all patients with any recent contact with the medical system for potentially drug-resistant pathogens. The effect of this recommendation has been striking. Investigators from the Veterans Health Administration, for example, documented an increase in vancomycin prescribing from 16% of pneumonia hospitalizations in 2006 to 31% in 2010 and a parallel increase in piperacillin-tazobactam prescribing from 16% in 2006 to 27% in 2010.4 In practice, however, only 2% of patients with pneumonia had positive blood or sputum cultures for methicillin-resistant Staphylococcus aureus (MRSA) and 2% had positive cultures for Pseudomonas aeruginosa. Rates of MRSA have been steadily declining during the past 15 years, so the prevalence of MRSA pneumonia is likely even lower now.5
An emerging set of studies has added another layer of complexity to this story by suggesting that overtreatment with antibiotics in general and overuse of broad-spectrum agents in particular may not only be unnecessary but may in fact be deadly. The investigation by Jones and colleagues6 in this issue of JAMA Internal Medicine is the latest study to make this assertion.
Jones and colleagues6 retrospectively evaluated 88 605 hospitalizations for community-onset pneumonia in Veterans Affairs hospitals. More than one-third of patients (38%) were empirically treated for MRSA using vancomycin or linezolid. The investigators compared mortality rates among patients who received empirical anti-MRSA therapy vs those who did not, adjusting for possible differences between the 2 groups using state-of-the-art epidemiologic methods and detailed clinical data. Remarkably, Jones and colleagues6 found that patients treated with empirical anti-MRSA agents were about 36% more likely to die compared with patients treated with standard first-line agents for community-acquired pneumonia (typically a quinolone or the combination of a β-lactam plus a macrolide). This finding was consistent in subgroup analyses restricted to patients in the intensive care unit, patients with risk factors for MRSA infection, and patients with positive nasal polymerase chain reaction test results for MRSA. The only group for whom there was no clear signal of harm was the subset with positive clinical cultures for MRSA (2% of the study population). Jones and colleagues6 did elicit some potential mechanisms of harm: empirical vancomycin was associated with higher rates of acute kidney injury, Clostridium difficile infection, and hospital-acquired vancomycin-resistant Enterococcus or gram-negative bacteremia or bacteriuria.
The findings from this study need to be interpreted with caution. Patients treated with anti-MRSA regimens were sicker than those treated with standard regimens. And while the investigators applied state-of-the-art methods to try to control for confounding by indication, it is possible that there was residual confounding. Indeed, it is difficult to reconcile the investigators’ proposition that a few days of treatment with vancomycin (the anti-MRSA agent used 98% of the time) may increase mortality rates by almost 40% with either clinical intuition or randomized clinical trials that have compared vancomycin with other agents.7,8
Notwithstanding the uncertainties in this analysis, the finding that more aggressive antibiotic strategies may be associated with higher mortality rates compared with more conservative approaches has now been reported multiple times.9-11 Webb and colleagues,9 for example, reported that broad-spectrum antibiotics were associated with a 4-fold higher odds of death among 1995 adults treated for pneumonia in 4 Utah emergency departments. Likewise, Kett and colleagues11 reported that broad-spectrum regimens were associated with a 2-fold increase in mortality among 303 patients with health care-associated pneumonia and risk factors for multidrug-resistant pathogens.
Conversely, a growing number of studies suggest that deferring antibiotics until cultures return may not only be safe but may even be associated with lower mortality rates. Hranjec and colleagues,12 for example, compared outcomes in a surgical intensive care unit during a baseline year in which antibiotics were given immediately whenever infection was suspected vs the following year in which antibiotics were given only after objective confirmation of infection (patients requiring vasopressors could be treated immediately). The conservative strategy was associated with less use of antimicrobial agents, more appropriate antibiotic courses, and lower mortality rates. Likewise, randomized trials have reported that outcomes are similar for patients with suspected pneumonia regardless of whether they are treated as soon as pneumonia is suspected or if treatment is withheld until bronchoscopic gram stain or cultures turn positive.13,14 The findings by Jones and colleagues6 are also concordant with pilot studies suggesting that rapid discontinuation of vancomycin driven by negative MRSA polymerase chain reaction results on nasal swabs or respiratory specimens may be associated with lower mortality rates and shorter hospitalizations.15,16
All told, pneumonia is frequently overdiagnosed, MRSA and Pseudomonas are very rare in community-onset pneumonia, empirical coverage with broad-spectrum agents may be associated with higher mortality rates, and there is often no harm in deferring treatment of clinically stable patients until diagnostic microbiology is available. The messages are clear: empirical coverage with broad-spectrum agents is not indicated in most patients being treated for pneumonia, and if the diagnosis of pneumonia is uncertain, there is no harm and likely some benefit in taking some time to acquire more diagnostic clarity before treating (such as further imaging, laboratory testing, treating concurrent conditions, and/or observation), so long as the patient is clinically stable.
Corresponding Author: Michael Klompas, MD, MPH, Department of Population Medicine, Harvard Medical School and Harvard Pilgrim Health Care Institute, 401 Park Drive, Ste 401 East, Boston, MA 02215, email@example.com
Published Online: February 17, 2020. doi:10.1001/jamainternmed.2019.7251
Conflict of Interest Disclosures: Dr Klompas reported receiving grant funding from the Centers for Disease Control and Prevention and the Massachusetts Department of Public Health, and royalty payments from UpToDate for articles on hospital-acquired pneumonia.
Identify all potential conflicts of interest that might be relevant to your comment.
Conflicts of interest comprise financial interests, activities, and relationships within the past 3 years including but not limited to employment, affiliation, grants or funding, consultancies, honoraria or payment, speaker's bureaus, stock ownership or options, expert testimony, royalties, donation of medical equipment, or patents planned, pending, or issued.
Err on the side of full disclosure.
If you have no conflicts of interest, check "No potential conflicts of interest" in the box below. The information will be posted with your response.
Not all submitted comments are published. Please see our commenting policy for details.
Klompas M. Overuse of Broad-Spectrum Antibiotics for Pneumonia. JAMA Intern Med. Published online February 17, 2020. doi:10.1001/jamainternmed.2019.7251
Coronavirus Resource Center
Create a personal account or sign in to: