P99 Beyond the pleura: bedside ultrasound evaluation of extravascular lung water in patients undergoing haemodialysis

Background There is growing interest in bedside thoracic ultrasound (TUS) beyond assessment of pleural disease or as an adjunct for interventions. TUS is used by some clinicians as an extension of physical examination, assessing the lung with results used to influence clinical decisions. Despite the publication of training curricula and consensus guidelines,1 there are few objective and robust data to demonstrate the utility of TUS in this area of clinical practice. Methods 30 patients undergoing haemodialysis were prospectively recruited to an observational cohort study (NCT01949402; REC 13/SC/0319). Patients underwent standardised TUS assessment before, during and after haemodialysis; a total lung B-line score was generated, alongside a binary label of whether appearances were consistent with interstitial syndrome or not. TUS video clips were recorded and scored by two blinded expert clinician sonographers asked to follow consensus statement guidance.1 Low-dose non-contrast CT thorax pre- and post-dialysis was used as the ‘gold standard’ radiologic comparison, and completed a questionnaire addressing satisfaction with TUS assessment. Results TUS detected a progressive reduction in B-line score in most patients undergoing haemodialysis, with moderate correlation with the volume of fluid removed once those patients with minimal B-lines pre-dialysis were discounted (figure 1). By contrast, there was no lung parenchymal change evident on CT pre- and post-dialysis in any of the patients studied. Interobserver agreement was good for total B-line score (ICC 0.63, 95% CI 0.52–0.72) and diagnosing interstitial syndrome (κ=0.60, 95% CI 0.47–0.73). TUS assessment was acceptable to patients, with none considering it time-consuming or unwilling to have it again if needed. Conclusion This is the first study to demonstrate, using blinded outcome assessment, that TUS can detect variation in the appearance of the lungs, manifest as a B-line score, caused by changes in fluid status during haemodialysis, and that TUS appears to be more sensitive than CT. Further studies are needed to investigate the utility of TUS as a diagnostic tool in this and similar clinical contexts and how it might impact on patient care and outcomes. Funding Esaote UK; Rosetrees Trust, UK Reference Intensive Care Med 2012;38(4):577–91.