Non-invasive measurement of pulmonary blood flow during prone positioning in patients with early acute respiratory distress syndrome

In the daily clinical routine at the bedside, information on effective pulmonary blood flow (PBF) is limited and requires invasive monitoring, including a pulmonary artery catheter, to determine both cardiac output and intrapulmonary shunt. Therefore we evaluated a non-invasive method for the measurement of PBF in a clinical setting, including 12 patients with acute respiratory failure (acute respiratory distress syndrome) undergoing prone positioning. PBF was determined before (baseline), during and after prone positioning, by using a foreign gas rebreathing method with a new photoacoustic gas analyser. Values were compared with the cardiac output corrected for intrapulmonary shunt (COeff). Responders to prone positioning were defined according to the improvement of arterial oxygenation. A total of 84 measurements were performed. PBF values correlated well with COeff (R2=0.96; P<0.0001). Bias and limits of agreement (+/- 2 S.D.) for all measurements were -0.11 +/- 0.76 litre/min. At baseline, responders showed significantly lower PBF levels than non-responders (4.8 +/- 1.0 compared with. 6.4 +/- 1.2 litre/min; P=0.03). During prone positioning, PBF increased continuously in responders and remained high after patients had been returned to the supine position. PBF was unaffected in non-responders. Mean total increase in PBF was 1.2 +/- 0.2 litre/min in responders compared with -0.4 +/- 0.2 litre/min in non-responders (P<0.0001). In conclusion, the investigated rebreathing system allows for a non-invasive determination of PBF at the bedside. The accuracy of the measurements is comparable with the thermodilution method. It is able to reliably reflect changes in PBF induced by prone positioning. Moreover, measuring PBF might be a promising tool to identify responders to prone therapy.