Dear Editor,

Pneumonia is among the most common ventilator-associated conditions observed in the intensive care unit [1]. Incorporating the Gram staining results of respiratory tract secretion specimens into a score which can be used to diagnose pneumonia in mechanically ventilated patients [modified Clinical Pulmonary Infection Score (mCPIS)] improves diagnostic accuracy and patient survival [2]. Bronchoalveolar lavage (BAL) is widely used for the microbiological diagnosis of pneumonia with good sensitivity and specificity. Non-bronchoscopic BAL (blindBAL) is a simple, manageable and cost-effective alternative to BAL with flexible bronchoscopy, when BAL is contraindicated [3]. However, some authors have suggested that it may be wise to perform a BAL where the disease is known to be present and that lavage should to be done in the most affected areas of the lung [4]. While this guideline is relatively easy to apply when the procedure is performed during bronchoscopy, to date no comparable guideline has been proposed to guide a blindBAL.

Electrical impedance tomography (EIT) is a novel, promising, non-invasive, radiation-free monitoring tool that allows real-time imaging of lung ventilation. It has been recently proposed that this instrument can be applied for positive end-expiratory pressure (PEEP) titration and lung recruitment monitoring, especially in patients with the acute respiratory distress syndrome [5].

We applied EIT (PulmoVista® 500; Dräger Medical GmbH, Lübeck, Germany) monitoring while a physician with expertise in bronchoscopy performed a bronchoscopic BAL in two different patients with ventilator-associated pneumonia, with BAL performed in the right lung of one patient and in the left lung in the other patient. Subsequently, we applied EIT monitoring once again while a blindBAL was performed in two different patients with pneumonia. All patients or next of kin provided informed consent for data gathering. The patients were sedated and paralyzed during the procedure and were ventilated through an endotracheal tube in a volume-controlled mode [FiO2 (fraction of inspired oxygen) = 1, PEEP = 0, tidal volume and respiratory rate set to allow adequate ventilation and to avoid excessive peak pressure]. Both the BAL and blindBAL procedures were performed through the insufflation of 100 ml of normal saline (5 aliquots). During the procedure we monitored and recorded a 5-min period of EIT signals which were later reviewed by three different physicians with no expertise in EIT who were blind to the procedures performed. The placement of the bronchoscopy and of the blindBAL catheter produced a silent EIT signal, since the absence of tidal ventilation corresponds to no flow due to airway occlusion at the location of bronchoscopy/catheter placement. When 80–100 ml of normal saline was insufflated in four to five aliquots, new EIT signals were recorded in a previously non-aerated area of the lungs, suggesting alveolar lavage in both the right [Fig. 1; Electronic Supplementary Material (ESM) Fig. 3] and left (ESM Figs. 2, 4) lung. Such signals were clearly recognizable during both the bronchoscopic (ESM Figs. 3, 4) and blind procedures (Fig. 1; ESM Fig. 2).

Fig. 1
figure 1

Electrical impedance tomography (EIT) signals acquired during a non-bronchoscopic bronchoalveolar lavage (blindBAL) performed in the dorsal right lung (ROI3). Note: The following description pertains to Fig. 1 shown here, as well as to ESM Fig. 2 [EIT signals during a blindBAL performed in the dorsal left lung (ROI4)], ESM Fig. 3 [EIT signals during a bronchospic BAL performed in the dorsal right lung (ROI3)] and ESM Fig. 4 [EIT signals during a bronchoscopic BAL performed in the dorsal left lung (ROI4)]. Screenshots of the PulmoVista® 500 data analysis. All EIT signals were acquired with a 16-electrode belt placed around the thorax between the 5th or 6th parasternal intercostal space. The signals were recorded and later reviewed offline using a dedicated software program (Dräger Medical GmbH EIT data analysis version 6.1). Upper left Dynamic images which describe the regional distribution of the electrical impedance of lung parenchyma, where the freeze frame shows the regional impedance of the lung at the moment of normal saline insufflation and the yellow circle shows how the freeze frame of the dynamic image appears when normal saline is insufflated in the lung. Horizontal layouts ROI 1–4 describe the variation of impedance over time in four different lung quadrants, as defined in the caption. Top horizontal layout (TV = 100 %) is the average of the four layouts. Layout RO13: yellow square highlights the recordings obtained during the procedure, red circles are the signal spikes which indicate the insufflation of a normal saline aliquot

Full size image

The three physicians who were asked to blindly review the recordings were all able to detect the moment and to identify the EIT signals corresponding to normal saline insufflation in the lung where the BAL was performed.

Further studies are warranted to validate a protocol of EIT-monitoring to guide lung sampling with non-bronchoscopic BAL in mechanically ventilated patients with pneumonia.