1036. Is Gram Stain of Bronchoalveolar Lavage a Reliable Clinical Indicator?
Session: Poster Abstract Session: Diagnostic Procedures in Clinical Practice
Saturday, October 22, 2011
Room: Poster Hall B1
Background: Ventilator associated pneumonia (VAP) is common in the surgical intensive care unit (SICU).  Patients with VAP are treated empirically with broad-spectrum antibiotics prior to culture results.  Early de-escalation of antimicrobials based upon Gram stain of BAL samples prior to final culture results promises the potential to decrease costs and drug related complications without compromising patient outcome. 

Methods: We conducted a retrospective cohort study of adult inpatients admitted to the SICU at a tertiary care medical center who underwent BAL over a 5 month period.  Data was collected regarding initial Gram stain and final culture results from BAL samples. 

Results: Among 135 patients, 255 BAL samples were submitted for Gram stain and culture. Of all samples, 63/255 (24.7%) had Gram negative bacilli (GNB) on Gram stain; 57 grew GNB and six did not; 192/255 (75.3%) had no GNB on Gram stain; 31 grew GNB and 161 did not. Of 192 Gram stain negative samples, 10 grew Pseudomonas aeruginosa.  Gram stain for GNB had sensitivity of 64.7%, specificity of 96.4%, positive predictive value of 90.5% and negative predictive of 83.9%.  For P. aeruginosa, negative predictive value of Gram stain was 94.8%.  Of all samples, 77/255 had Gram positive cocci (GPC) on Gram stain; 57 grew GPC and 20 did not; 178/255 had no GPC on Gram stain; 13 grew GPC and 165 did not.  Of 178 Gram stain negative samples, four grew pathogenic GPC (three MRSA, one MSSA).  Gram stain for GPC had sensitivity of 81.4%, specificity of 89.2%, positive predictive value of 74.0% and negative predictive value of 92.7%.  For pathogenic GPC the negative predictive value of a Gram stain was 97.8%.  Using the first sample per patient revealed the growth of two pathogenic GPC (MRSA in one, MSSA in one) in Gram stain negative samples with a negative predictive value was 97.6%.

Conclusion: Gram stain of BAL samples appears to be a reliable predictor for the absence of pathogenic GPC.   Early de-escalation of antimicrobials for resistant GPC based upon Gram stain may decrease costs and drug related complications without compromising patient outcome.   Gram stain of BAL samples appears not to be a reliable predictor for the presence or absence of GNB and should not influence anti-GNB therapy prior to the finalization of cultures.


Subject Category: J. Clinical practice issues

Douglas Haas, MD, Internal Medicine, The Ohio State University Medical Center, Columbus, OH, Jeremy Taylor, PharmD, BCPS, Pharmacy, The Ohio State University Medical Center, Columbus, OH and Kurt Stevenson, MD, MPH, The Ohio State University Medical Center, Columbus, OH

Disclosures:

D. Haas, None

J. Taylor, None

K. Stevenson, None

Findings in the abstracts are embargoed until 12:01 a.m. EST Thursday, Oct. 20 with the exception of research findings presented at IDSA press conferences.