1257. Epidemiology of Healthcare-Associated Pneumonia (HCAP) as Assessed by Blood Cultures versus Respiratory Cultures
Session: Poster Abstract Session: Clinical Infectious Diseases: Respiratory Infections
Friday, October 28, 2016
Room: Poster Hall
Background: Epidemiological studies of HCAP are usually based on respiratory cultures. The epidemiology of HCAP in relation to culture source is limited. We compared the microbiology of blood and respiratory specimens in a large US database.

Methods: This retrospective cohort study assessed adult patients admitted with HCAP from 2010- 2015 to 177 US hospitals participating in Premier and providing administrative and microbacteriological data. HCAP was defined as having prior hospitalization within 90 d, hemodialysis, admission from skilled nursing facility, or immune suppression. Patients with identical gram negative organisms in blood and urine were excluded. Cultures were obtained in the first 5 hospital days.

Results: Of 48,440 patients hospitalized with HCAP, 36.0% were admitted to ICU. Mean age was 69± 15 (range 18 - 89) and 49% were male. Almost all patients (98%) had at least one culture, including blood (96%) and respiratory (64%) specimens. Of the respiratory cultures, 6554 (22%) were positive, whereas only 3,688 (8%) of blood cultures were positive. Gram-negative bacteria accounted for 55% of respiratory and 36% of blood pathogens. The most common respiratory pathogens were S. aureus (37.5%), P. aeruginosa (20.6%), E. coli (6.2%) and S. pneumoniae (5.9%). The most common blood pathogens were S. aureus (25.8%), S. pneumoniae (18.4%), E.coli (13.7%) and P. aeruginosa (5.3%). Positive cultures were more common in ICU than non-ICU patients (12.0% v. 5.6% for blood, 30.4% v. 15.1% for respiratory), but the relative frequency of pathogens was similar in both settings. Respiratory pathogens had lower mortality than blood pathogens for S. aureus (18.4% v. 26.5%; p<0.002) and P. aeruginosa (13.9% v. 30.4%; p<0.002) but not for S. pneumoniae (13.8% v. 11.4%, p=0.21) and E. coli (20.6% v. 19.2%, p=0.56). Patients with positive blood cultures had the highest mortality, followed by positive respiratory cultures and no growth (16.9% v. 14.8 v. 9.0, p<0.0001).

Conclusion: In a large US inpatient sample, microbial etiology of HCAP pathogens differed by specimen type. S. aureus was the most common respiratory and blood pathogen, P.aeruginosa was more common in respiratory cultures and S. pneumoniae in blood. Lower mortality for some respiratory cultures may indicate colonization rather than infection.

Abhishek Deshpande, MD, PhD1,2, Sarah Haessler, MD3, Kyle Brizendine, MD2, Peter Lindenauer, MD, MSc4, Pei-Chun Yu, MS5, Marya D. Zilberberg, MD, MPH6, Paul Bakaki, MD, PhD7, Thomas Higgins, MD, MBA8 and Michael Rothberg, MD, MPH1, (1)Medicine Institute, Cleveland Clinic, Cleveland, OH, (2)Infectious Disease, Cleveland Clinic, Cleveland, OH, (3)Infectious Diseases, Baystate Medical Center, Springfield, MA, (4)Center for Quality of Care Research, Baystate Medical Center, Springfield, MA, (5)Quantitative Health Sciences, Cleveland Clinic, Cleveland, OH, (6)University of Massachusetts and Evimed Research Group, LLC, Goshen, MA, (7)Epidemiology and Biostatistics, Case Western Reserve University, Cleveland, OH, (8)Medicine, Baystate Medical Center, Greenfield, MA

Disclosures:

A. Deshpande, None

S. Haessler, None

K. Brizendine, None

P. Lindenauer, None

P. C. Yu, None

M. D. Zilberberg, The Medicines Company: Grant Investigator and Investigator , Grant recipient , Research grant and Research support
Tetraphase: Grant Investigator and Investigator , Grant recipient , Research grant and Research support
Merck, Inc.: Consultant , Grant Investigator and Investigator , Grant recipient , Research grant and Research support

P. Bakaki, None

T. Higgins, None

M. Rothberg, None

See more of: Clinical Infectious Diseases: Respiratory Infections
See more of: Poster Abstract Session
<< Previous Abstract | Next Abstract >>

Findings in the abstracts are embargoed until 12:01 a.m. CDT, Wednesday Oct. 26th with the exception of research findings presented at the IDWeek press conferences.