[LUNG ULTRASOUND IN PATIENTS WITH HEMATOLOGICAL MALIGNANCIES AND ACUTE RESPIRATORY FAILURE DUE TO PNEUMONIA.]

To compare the accuracy of bedside lung ultrasound (LUS) and chest computed tomography (CT) for the de- tection of lung lesions in patients with hematological malignancies and acute respiratory failure (ARF).

39 patients with hematological malignancies and ARF were enrolled in prospective study. The investigation of the patients included LUS, chest C, extravascular lung water index (EVLW) by transpulmonary ther- modilution, and bronchoalveolar lavage (BAL).

There was correlation between the total number of B-lines and E VLW index (r = 0,40; p <0,05). The sensitivity, specificity of LUS in the total number of B-lines were 78% and 70%, respectively (and A UC 0,7). There were correla- tions between A-lines and volume of hyperaerated lung regions (r = 0,40; p <0,05) and normally ventilated (r = 0,60; p = 0,001) regions, between A-lines and the total lung volume (r = 0,50; p = 0,001), between A-lines and volume of poorlyventilated lung regions (r = -0,40; p = 0,001), A-lines and weight of normally ventilated lung regions (r = 0,50; p = 0,001), A-lines and weight of poorly ventilated regions (r = -0,35; p <0,05), total count of B-lines and volume of poorly ventilated lung regions (r = 0,4; p = 0,001), between total count of B-lines and weight poorly ventilated lung regions (r = 0,4; p = 0,001). There were associations between USfeathers and etiology ofpneumonia. A-lines were often detected in patients with gram-negative bacterial pneumonia and fungal pneumonia more than in patients with pneu- mocystis pneumonia. B-lines were detected often in patients with Pneumocystis pneumonia. Sensitivity ofLUS pleural effusion assessment was 95%, specificity was 90%.

LUS is high sensitivity and specificity method to detect lung lesions in patients with ARF.