Risk stratification for the development of a subsequent pneumonia after a nondiagnostic bronchoalveolar lavage.

Background: Broncho-alveolar lavage (BAL) is an invasive bedside procedure to define type and concentration of pathologic organisms causing ventilator associated pneumonia (VAP). We evaluated if the absence of pathogens on final results represented a lavage aspect of the BAL as a therapeutic procedure to eliminate organisms.

Methods: BAL results collected from 2008 to 2009 were stratified as positive (POS) ≥ 100,000 cfu), indeterminate (INT) ≤ 100,000 cfu pathologic organisms, or negative defined as mixed flora (MF) or sterile (STR). The INT, MF, and STR results were assessed by incidence of a subsequent POS sample.

Results: Nine-hundred forty-nine BALs performed on 490 SICU patients were interpreted as POS in 227 patients (46%). 237 non- POS patients needed a subsequent BAL. Any pathogen on the first BAL (INT group) indicates a high likelihood for subsequent BAL which will be POS. Monthly cumulative sum analysis (CUSUM) of yield was unable to identify any specific period in which BAL performance varied from trend.

Conclusion: MF and STR represent adequate sampling of secretions that are clinically benign. Any pathogen, regardless of concentration, should be considered a biomarker for future pneumonia. CUSUM analysis suggest better training in timing and indication may decrease unnecessary procedures yielding negative results.