When to Consider the Use of Antibiotics in the Treatment of 2009 H1N1 Influenza-Associated Pneumonia
We are now facing a pandemic caused by an epidemiologically distinct, novel virus, the 2009 pandemic influenza A (H1N1) virus (swine flu), against which few persons born since 1970 have antibodies. The severity of illness in the individual varies, and our understanding of the role of bacterial infection in novel 2009 H1N1 infection is still evolving. A current summary of bacterial isolates from 53 fatal pediatric cases of novel H1N1 with adequate sampling of normally sterile sites showed that 17 (32%) had bacterial pathogens of which 8 were Staphylococcus aureus, and of these, 6 were methicillin-resistant S. aureus (MRSA) (http://www.cdc.gov/flu/weekly/). Although data from fatal cases are not representative of the influenza illnesses commonly encountered by community-based practitioners, the fact that secondary bacterial pneumonia is occurring with the novel H1N1 strain highlights a challenge facing clinicians who make decisions about antibiotic treatment of their influenza patients with lower respiratory tract disease. Table 1 qualitatively outlines some of the factors that would make one consider a bacterial component to H1N1 influenza illness.
Three adolescent patients recently hospitalized at the Children’s Hospital at Dartmouth brought this dilemma into focus for us. All three had underlying neuromuscular disease with prior respiratory compromise. One child presented within 24 hours after onset of symptoms, required intubation, and had hematologic and clinical and laboratory findings suggestive of viral pneumonia. The other two had onset of respiratory decline and recrudescence of fever after an influenza-like illness, and they also required intubation and ventilation. Each had novel H1N1 infection confirmed by polymerase chain reaction (PCR). Sputum specimens from the latter two patients grew in one case Moraxella catarrhalis and in the other methicillin-sensitive S. aureus. All three patients were empirically started on broad-spectrum antibiotics (vancomycin and piperacillin–tazobactam) by their pediatric critical care attending physician at the time of hospitalization. The clinical course in each child has been one of slow resolution, with narrowing of the antibiotic spectrum on the basis of the bacteriologic results.
For the child or adult admitted to a hospital intensive care unit in respiratory distress, we believe that empirical initial therapy with broad-spectrum antibiotics to include coverage for MRSA, as well as Streptococcus pneumoniae and other common respiratory pathogens, is appropriate. For the previously healthy child or adult with influenza who requires admission to a community hospital and has features that suggest a secondary pneumonia (Table 1), we would recommend empirical treatment with a drug such as intravenous second- or third-generation cephalosporin, after an effort has been made to prove the association with influenza and to get adequate lower respiratory tract specimens for Gram’s stain and bacterial culture. If the Gram’s stain suggests the presence of staphylococci or if there is a rapidly progressive or necrotizing pneumonia, an additional antimicrobial agent to cover MRSA is appropriate.
For outpatient treatment of most patients who have influenza–associated pneumonia with a suspected secondary bacterial infection, the bacterial component can be treated with appropriate oral antibiotics for age — amoxicillin–clavulanate or a second-generation cephalosporin for both children and adults. There is no evidence for synergistic coinfection of influenza with Mycoplasma pneumoniae or other agents of atypical pneumonia. We do not believe that initial coverage for MRSA is indicated in all patients who are thought to have secondary bacterial pneumonia. Moreover, given emerging epidemiologic and clinical data, we have a strong suspicion that much of the lower respiratory tract illness will turn out to be of viral origin and should not require antibacterial therapy.
Peter F. Wright, M.D.
Kathryn B. Kirkland, M.D.
John F. Modlin, M.D.
Section of Infectious Disease and International Health
Departments of Pediatrics and Medicine
Dartmouth Medical School
No potential conflict of interest relevant to this article was reported.
This article (10.1056/NEJMopv0910749) was published on November 25, 2009, at NEJM.org.
13 Responses »
- Wanneer antibiotica overwegen bij een A/H1N1 griepvirus geässocieerde pneumonie? « Nederlandstalige Journal Club