Developing local treatment guidelines for healthcare-associated pneumonia.

TO THE EDITOR—We read with interest the meta-analysis by Chalmers et al [1], which demonstrates that the healthcareassociated pneumonia (HCAP) definition poorly predicts the presence of resistant pathogens. Based on these findings, the authors encourage treatment for HCAP to be guided by the local prevalence of multidrug-resistant pathogens. In hopes of constructing a local syndromic antibiogram specific to HCAP, we retrospectively identified inpatients treated for pneumonia at our facility, the Richard L. Roudebush VA Medical Center, between 1 January 2011 and 31 December 2012. The Roudebush center is a tertiary care facility that provides complete medical care for 85 000 adults in Indianapolis, Indiana. Potential cases were identified by the following International Classification of Diseases, Ninth Revision codes: 480.0–480.9, 481, 482.0– 482.9, 483.0–483.8, 484.1–484.8, 485, 486, and 487. All medical records were reviewed, and only cases that met criteria for HCAP were selected for further analysis [2]. A total of 113 cases of HCAP were identified; 98% of patients were men, and the mean age was 71 years. Blood cultures were obtained in 103 patients (91%), sputum cultures in 47 (42%), and bronchoalveolar lavage specimens in 2 (2%). The sputum specimen was graded as good in 15 (32%), fair in 29 (62%), and poor in 3 (6%). At least 1 microbiologic pathogen was identified by either blood or respiratory samples in only 26 cases (23%). Enterobacteriaceae were isolated in 10 cases (38%), methicillinsusceptible Staphylococcus aureus in 5 (19%), Pseudomonas aeruginosa in 4 (15%), Streptococcus pneumoniae in 4 (15%), and methicillin-resistant S. aureus (MRSA) in 3 (12%). Although we collected 2 years of data, our sample size of culture-positive cases (n = 26) was small. The culture-positive rate of 23% is similar to that in several other HCAP studies [3–7]. This low culture-positive rate may reflect both the difficulty of collecting sputum samples in nonventilated patients and the poor quality of the samples that were collected [8]. To augment our limited microbiologic data, we have also monitored clinical outcomes inpatientswhohadnomicrobiologic