Ultrasound B-lines Do Not Accurately Quantify Mild Pulmonary Edema Induced by Rapid Saline Infusion in Normal...

Background: Thoracic ultrasound quantification of B-line artifacts is thought to represent pulmonary edema; scores counting B-lines have been validated in critically ill patients with severe edema. However, B-lines scores increasingly used in research to assess mild interstitial pulmonary edema in normal participants, are unvalidated. Methods: We created interstitial pulmonary edema using rapid infusion of 20ml/kg normal saline, in 10 healthy volunteers. Blinded, randomized video images of thoracic ultrasound performed on 8 regions of the thorax before, during, after infusion, and recovery were evaluated by four independent reviewers; B-line Scores were averaged across reviewers. The total was calculated as the sum of all 8 areas with a second score calculated for posterior lung (dependent) base. Interstitial edema was quantified by impedance cardiography thoracic fluid content (TFC). Hemoglobin, hematocrit, B-type Natriuretic Peptide (BNP), and N-terminal-pro-BNP (NT-pro-BNP) were measured. Results: Hemoglobin (14.0±1.5g/dL pre,13.2±1.3 post) and hematocrit (41.2±4.7% pre, 38.8±4.0 post) decreased significantly post-infusion (P=0.002, P=0.004, respectively); BNP was increased (P=0.006). TFC increased 12.7±2.9% (35.9±6.0 1/kOhm pre-infusion; 40.4±6.7 post-infusion P<0.0001). There were no significant changes in B-line scores either in the whole lung (8.2±5.4 pre-infusion; 10.3±7.2, post P=0.29) or dependent lung (P=0.12) and no significant relationship between B-line scores and TFC (P=0.13 whole lung; P=0.15 dependent) or the change in TFC pre-post infusion (P>0.8 for both). Conclusions: Thoracic ultrasound scores do not show good validity for early, subclinical, pulmonary edema. Further validation in those with subclinical pulmonary are required before they are used for research in this context.