Description of Practice and Outcomes in Healthcare-associated Pneumonia without a Microbiologic Diagnosis in Non-ventilated Patients

Background: Patients with health care-associated pneumonia (HCAP) are typically started on empiric broad-spectrum (BS) antibiotics.  Limited data exists on how to de-escalate antibiotics if a microbiologic diagnosis is never made. 

Methods: We retrospectively identified hospitalized patients treated for HCAP at the Roudebush VAMC between 1/1/2011 and 12/31/2012.  All cases met the definition of HCAP: 1) a new radiographic infiltrate, 2) signs/symptoms of pneumonia, and 3) at least one risk factor for a multidrug-resistant (MDR) pathogen.  Patients receiving mechanical ventilation at the time of diagnosis were excluded.  BS therapy was defined as a combination of anti-pseudomonal and anti-MRSA antibiotics.  De-escalation involved a switch to or discontinuation of an agent resulting in a narrower spectrum of coverage.

 

Results: A total of 90 patients were included in the analysis.  The mean age was 70 years; 88 (98%) were men.  Blood cultures were obtained from 77 patients (86%); sputum cultures were obtained from 35 (39%), and bronchoalveolar lavage was performed in 2(2%).   The sputum specimen was graded as fair in 71% and good in 29%.  Of patients with a pathogen identified, 7/20 (35%) were MDR.  Overall, 70 patients (78%) had no pathogen identified.

The median duration of total antibiotics was 10 days (IQR 8-12).  Antibiotics were de-escalated in 60/68 (88%) of cases who were started on BS therapy.  The median time to de-escalation was 3 days (IQR 2-4).

De-escalation was performed in 47/54 (87%) of patients without a pathogen identified.  The narrower regimen was parenteral in 8 (17%) and oral in 39 (83%).  Among patients without a microbiologic diagnosis, vancomycin was more commonly stopped when the MRSA nasal-swab was negative instead of positive (71% vs. 17%, p<0.01).   Compared to other patients started on BS antibiotics, patients without a microbiologic diagnosis who were de-escalated had comparable rates of mortality and readmission at one-month: mortality 13% vs. 11%, readmission 20% vs. 21%. 

 

Conclusion:   A microbiologic diagnosis was infrequently made in cases of HCAP.  In the absence of a microbiologic diagnosis, physicians safely de-escalated antibiotic therapy.  The effectiveness of this approach may reflect low rates of antibiotic resistance in the study cohort.