Bronchoalveolar lavage fluid dilution in ICU patients: what we should know and what we should do

Abstract

The development of bronchoscopy and bronchoalveolar lavage (BAL) has led to an increase in their use in intensive care units (ICUs), where their applications for differential diagnosis of pulmonary diseases make them indispensable instruments for intensivists [1]. Despite their common use, a few studies have raised concerns about potential impacts on bronchoalveolar lavage fluid (BALF) dilution, which affects mainly the quantitative detection of soluble substances. Urea is a diffusible substance that can easily be detected in capillaries and alveolar spaces. The urea concentration in plasma and that in BALF are approximately equal and their ratio (urea plasma/urea BALF) has previously been applied as an index of BALF dilution. Furthermore, it has been shown that the ratio of high-quality lavage is low in clinical settings [2, 3]. We reviewed all ICU-admitted patients who received BAL from January 2016 to September 2018 in Ren Ji Hospital and analyzed their urea plasma/urea BALF values. Guidelines of the American Thoracic Society were followed during the BAL procedure [3]. (The procedure is described in Additional file 1.) Among 223 patients included, the median level of urea plasma/urea BALF was 4.2 (interquartile range of 3.2–8.6). The patients were categorized into groups A (urea plasma/urea BALF <4.2) and B (urea plasma/urea BALF ≥4.2). The patients in group A were more likely to receive bronchodilators (35.6% versus 15.9%, P <0.001) and a recruitment maneuver (15.5% versus 5.3%, P = 0.013) than those in group B. More invasive pulmonary aspergillosis (IPA) patients with BALF galactomannan of more than 0.5 could be detected in group A than in group B (84.6% versus 33.3%, respectively; P = 0.019) as well as more bacterial pneumonia patients with the quantitative cultures of BALF of more than 10 CFU/mL (90.6% versus 52.7%, respectively; P <0.001). Primary care physicians performed more BAL than residents did (58.3% versus 31.8%, respectively), especially in group A (Table 1). Pulmonary function was associated with the urea plasma/urea BALF ratio. It was found that there was a correlation between urea plasma/urea BALF and partial pressure of arterial oxygen/fractional concentration of inspired oxygen (PaO2/FiO2) (R 2 = 0.196, P <0.001). The less oxygen-deficient the patient was, the lower the urea plasma/urea BALF level was (Fig. 1a,b). Sixty-eight patients with chronic obstructive pulmonary disease (COPD) were enrolled in our study. The forced expiratory volume in the first second (FEV1) was suggested as a measure of bronchial obstruction. FEV1 of less than 50% of the predicted normal value indicated the presence of severe ventilatory impairment, which led to a lower volume of instilled saline flow into the alveoli. In our study, a correlation was also found between FEV1 and urea plasma/ urea BALF (R = 0.299, P <0.001). A lower value of urea plasma/urea BALF was obtained in a group with FEV1 of at least 50% of the predicted value than in that with FEV1 of less than 50% of the predicted value (P <0.05, Fig. 1c, d). Providing appropriate training in BAL skills to intensivists while ensuring patient safety is challenging [4]. Inter-operator variability in the recovery of lavage fluid during a BAL procedure may affect the concentration of soluble substances such as galactomannan and the results of quantitative cultures [5]. More attention should be paid to patients with hypoxia and impaired pulmonary function. Bronchodilators and a recruitment maneuver may improve BALF dilution during the procedure, and residents in ICUs need more practice. * Correspondence: [email protected]; [email protected] Yuetian Yu and Chunyan Liu contributed equally to this work. Department of Rheumatology, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, 145, Middle Shangdong Road, Shanghai 200001, China Department of Critical Care Medicine, Ren Ji Hospital, School of Medicine, Shanghai Jiao Tong University, 145 Middle Shangdong Road, Shanghai 200001, China Full list of author information is available at the end of the article