Nonbronchoscopic Bronchoalveolar Lavage Improves Respiratory Culture Accuracy in Critically Ill Patients

Abstract

Objectives: Diagnosis of pneumonia is challenging in critically ill, intubated patients due to limited diagnostic modalities. Endotracheal aspirate (EA) cultures are standard of care in many ICUs; however, frequent EA contamination leads to unnecessary antibiotic use. Nonbronchoscopic bronchoalveolar lavage (NBBL) obtains sterile, alveolar cultures, avoiding contamination. However, paired NBBL and EA sampling in the setting of a lack of gold standard for airway culture is a novel approach to improve culture accuracy and limit antibiotic use in the critically ill patients.

Design: We designed a pilot study to test respiratory culture accuracy between EA and NBBL. Adult, intubated patients with suspected pneumonia received concurrent EA and NBBL cultures by registered respiratory therapists. Respiratory culture microbiology, cell counts, and antibiotic prescribing practices were examined.

Setting: We performed a prospective pilot study at the Cleveland Clinic Main Campus Medical ICU in Cleveland, Ohio for 22 months from May 2021 through March 2023.

Patients or subjects: Three hundred forty mechanically ventilated patients with suspected pneumonia were screened. Two hundred fifty-seven patients were excluded for severe hypoxia (Fio₂ ≥ 80% or positive end-expiratory pressure ≥ 12 cm H₂O), coagulopathy, platelets less than 50,000, hemodynamic instability as determined by the treating team, and COVID-19 infection to prevent aerosolization of the virus.

Interventions: All 83 eligible patients were enrolled and underwent concurrent EA and NBBL.

Measurements and main results: More EA cultures (42.17%) were positive than concurrent NBBL cultures (26.51%, p = 0.049), indicating EA contamination. The odds of EA contamination increased by eight-fold 24 hours after intubation. EA was also more likely to be contaminated with oral flora when compared with NBBL cultures. There was a trend toward decreased antibiotic use in patients with positive EA cultures if paired with a negative NBBL culture. Alveolar immune cell populations were recovered from NBBL samples, indicating successful alveolar sampling. There were no major complications from NBBL.

Conclusions: NBBL is more accurate than EA for respiratory cultures in critically ill, intubated patients. NBBL provides a safe and effective technique to sample the alveolar space for both clinical and research purposes.