2000 OPEN FORUM Abstracts
DIAGNOSIS OF VENTILATOR ASSOCIATED PNEUMONIA (VAP): BRONCHOSCOPIC BAL VS. BLIND, MINI-BAL USING THE COMBICATHtm
Sandra L. Miller MD, Robert S. Campbell RRT, FAARC, Jay A. Johannigman MD, Emily A. Montgomery BS, Fred A. Luchette MD, Joseph S. Solomkin MD, Donald C. Mann MD, Richard D. Branson BA, RRT. Division of Trauma/Critical Care, University of Cincinnati College of Medicine
Introduction: Diagnosis of VAP is difficult. CDC has defined VAP clinically by CXR, WBC, temp, and sputum characteristics. CombiCath? provides a way of blindly obtaining a protected alveolar lavage (PAL) specimen. We determined the Sensitivity and Specificity (SS) of PAL specimens from the CombiCath? to BAL specimens obtained bronchoscopically in pts with CDC defined VAP.
Methods: Forty-five pts were studied. BAL specimens were bronchoscopically obtained from each lung. A trained Respiratory Care Practitioner used the CombiCath? to blindly obtain PAL specimens. FiO2 of 1.0 was used during all procedures. Volume of lavage fluid administered and returned was recorded. Specimens were prepared identically and microorganisms were identified using standard lab methods. Threshold for infection was >104 CFU/ml for both techniques. Data collection included: age, gender, diagnosis and APACHE II, area of infiltration on CXR, ICU days, hosp days, and vent days. Baseline vital signs and O2 sat (SpO2) were recorded and any changes noted during/after each procedure. Change in MAP or HR > 20% or a decrease in SpO2 > 5% was considered a complication. Antibiotic use within 48hrs of sampling was noted.
Results: Table 1 reveals outcome variables based on culture results. Culture results were in agreement in 40/45 pts (32 NEG, 8 POS). In five cases, BAL was POS and PAL NEG. SS of PAL compared to BAL is 73% and 100%. Lavage administered/returned during BAL and PAL was 83/16 ml and 20/2 ml, respectively. Complications were noted during 21 BAL procedures (19 HR, 13 MAP, 5 SpO2) and 6 PAL procedures (5 HR, 3 MAP). Vent settings were altered during BAL on 33 occasions and none during PAL. Fourteen pts received antibiotics within 48 hr of sampling, of which five had POS cultures. CXR involvement was bilateral in 24 pts, R side only in 15, and L side only in 6. Time for each procedure was 16 min for PAL and 52 min for BAL. Discussion: CombiCath? may be useful in identifying pts with VAP. Confirmed VAP was not associated with increased hosp or ICU length of stay, vent days, or mortality. Trained bedside clinicians may perform PAL with a lower complication rate compared to BAL. Conclusion: The CombiCath? provides a safe and accurate alternative to bronchoscopic BAL for diagnosing VAP.
|Table 1. Outcome variables relative to culture results.|
|Variable||Hosp days||ICU days||Vent days||APACHE II||Death|
|Culture + (13)||25.4 ± 8.6||23.6 ± 8.2||15.7 ± 6.1||18.1 ± 7.9||0|
|Culture - (32)||28.1 ± 9.8||21.4 ± 8.7||15.1 ± 8.1||18.4 ± 6.3||5|