The effect of fluid resuscitation on the effective circulating volume in patients undergoing liver surgery: a post-hoc analysis of a randomized controlled trial.

To assess the significance of an analogue of the mean systemic filling pressure (Pmsa) and its derived variables, in providing a physiology based discrimination between responders and non-responders to fluid resuscitation during liver surgery. A post-hoc analysis of data from 30 patients undergoing major hepatic surgery was performed. Patients received 15 ml kg-1 fluid in 30 min. Fluid responsiveness (FR) was defined as an increase of 20% or greater in cardiac index, measured by FloTrac-Vigileo®. Dynamic preload variables (pulse pressure variation and stroke volume variation: PPV, SVV) were recorded additionally. Pvr, the driving pressure for venous return (=Pmsa-central venous pressure) and heart performance (EH; Pvr/Pmsa) were calculated according to standard formula. Pmsa increased following fluid administration in responders (n = 18; from 13 ± 3 to 17 ± 4 mmHg, p < 0.01) and in non-responders (n = 12; from 14 ± 4 to 17 ± 4 mmHg, p < 0.01). Pvr, which was lower in responders before fluid administration (6 ± 1 vs. 7 ± 1 mmHg; p = 0.02), increased after fluid administration only in responders (from 6 ± 1 to 8 ± 1 mmHg; p < 0.01). EH only decreased in non-responders (from 0.56 ± 0.17 to 0.45 ± 0.12; p < 0.05). The area under the receiver operating characteristics curve of Pvr, PPV and SVV for predicting FR was 0.75, 0.73 and 0.72, respectively. Changes in Pmsa, Pvr and EH reflect changes in effective circulating volume and heart performance following fluid resuscitation, providing a physiologic discrimination between responders and non-responders. Also, Pvr predicts FR equivalently compared to PPV and SVV, and might therefore aid in predicting FR in case dynamic preload variables cannot be used.