Cardiac output and CVP monitoring… to guide fluid removal

We read with interest the recently published position papers on central venous pressure (CVP) [1] and cardiac output (CO) [2] monitoring in critically ill patients and wish to further comment on their potential benefit. Haemodynamic monitoring is usually considered for haemodynamic support guidance during the early phase of shock. We would like, however, to emphasize the role of CVP and CO monitoring in another phase of critical illness, i.e. during volume depletion. Positive fluid balance and venous congestion have been associated with poor outcome in critically ill patients and in organ dysfunction (i.e. lung, kidney, liver, and gut) [3]. Volume depletion (i.e. using ultrafiltration or diuretics) is therefore a key part of the treatment of organ congestion and fluid overload after the initial phase of shock resuscitation. In these patients, the goal of volume depletion is to limit interstitial oedema and compartmental pressure [4]. Owing to the Starling forces, the intra-vascular hydrostatic pressure, oncotic pressure, and vascular permeability are the three factors driving trans-capillary filtration and therefore oedema generation. Volume depletion can limit trans-capillary filtration mostly through decreasing intra-venous and capillary hydrostatic pressure and possibly increasing intra-vascular oncotic pressure (Fig. 1). This strategy may, however, compromise venous return and therefore cardiac output, which may impact organ perfusion and organ function recovery. Initiation of volume depletion without haemodynamic monitoring may lead to either under- or over-treatment with a risk of hypoperfusion. The best balance would be to reach efficacy (i.e. decreasing intra-vascular hydrostatic pressure) without decreasing venous return and therefore stroke volume or cardiac output. Since venous return depends on the gradient between mean systemic pressure (Pms) and CVP, the ideal fluid depletion rate should probably aim at lowering Pms and CVP to the same extent (Fig. 2b), and therefore not comprising venous return. On the other hand, excessive or inappropriate fluid removal may induce a larger decrease in Pms than CVP and lead to a drop in venous return and stroke volume (Figs. 1 and 2b) [5]. As rightly pointed out by Monnet and Teboul, monitoring arterial pressure does not allow detection of all episodes of a decrease in cardiac output [2]. Guiding the rate of fluid volume depletion using haemodynamic monitoring allows monitoring both efficacy (i.e. lowering CVP) and tolerance (i.e. constant stroke volume). We strongly believe this approach may hasten organ function recovery, allow faster weaning from organ support, and therefore decrease the use of ICU resources.