NATURE REVIEWS | NEPHROLOGY VOLUME 8 | JANUARY 2012 | 57 Introduction Until recently, dialysis dose was thought to have a pivotal role in improving outcomes in critically ill patients who require contin- uous renal replacement therapy (CRRT). The relationship between treatment dose and outcome in patients with acute kidney injury was first investigated prospectively in a single-center study in which patients were randomly assigned to postdilution continu- ous venovenous hemofiltration (CVVH) at 45 ml/kg/h, 35 ml/kg/h or 20 ml/kg/h. Patients receiving a higher dose had better survival than those randomized to 20 ml/ kg/h. 1 Subsequently, conflicting results from small randomized clinical trials of both intermittent and continuous modali- ties of RRT have highlighted the need for larger multicenter trials. 2–5 Two large multicenter, randomized, con- trolled clinical trials did not find any benefit of an intensive dialysis dose over a standard dose. 6,7 These trials have dampened the enthusiasm of clinicians for considering dialysis dose as a modifiable factor influenc- ing outcomes. The results have contributed to the perception that increased dialysis dose might not be of benefit in improving outcomes and have also increased the ten- dency to not assess dialysis dose. 8 Although the evidence of the lack of benefit of increas- ing the dialysis dose above the levels that were tested in these trials is undeniable, it is instructive to evaluate the assumptions and findings in these studies. Assessing delivered dose Following the study by Ronco et al. 1 in patients on CVVH, all subsequent studies of CRRT have considered the delivered dose as equivalent to the total effluent volume, expressed as a weight-adjusted hourly efflu- ent rate. This assumption is based on the principle that the operational characteris- tics of a low ultrafiltration (in patients on CVVH) and dialysate flow (in patients on continuous venovenous hemodialysis [CVVHD] or continuous venovenous hemodiafiltration [CVVHDF]) result in complete saturation of small solutes in the effluent. Consequently, clearance is equal to the total effluent volume, which is cal- culated as the sum of net fluid removal rate (Qnet), the spent dialysate (Qd) and replacement fluid (Qr) in combined thera- pies such as CVVHDF, and the sum of Qnet and Qr in CVVH (Box 1). Therefore, efflu- ent volume is perceived to represent clear- ance in CRRT and has been widely utilized to prescribe and measure the dose in these modalities. A key supposition for this state- ment is that filter permeability remains constant over time (effluent fluid urea nitrogen [FUN]/blood urea nitrogen [BUN] ratio = 1). Unfortunately, this assumption is not correct. Three important treatment-related factors could reduce the actual delivered dose in patients receiving CRRT. First, con- centration polarization resulting from an increased concentration of rejected solvents on the membrane surface as a function of transmembrane flow, and protein fouling owing to the adsorption or deposition of matter on and in the separation layer of the membrane, lead to a concentrated layer immediately adjacent to the membrane and a decrease in diffusive transport (Figure 1). 9 Both concentration polarization and/or membrane fouling lead to the need for an increased transmembrane pressure in order to maintain an adequate ultrafiltration rate and also lower the concentration of poten- tially important solutes in the effluent. 10 Second, filter clotting progressively causes a decline in the sieving coefficient of the membrane and reduces filter permeability. The measurement of effluent volume is driven by the settings on the CRRT machine pump and does not reflect changes in filter permeability. Additionally, a third issue presents when predilution is applied. The prefilter infusion of replacement solution reduces the concentration of solutes in the plasma and decreases solute clearance. Brunet et al. 11 showed that this decrease in urea clearance could be as high as 15%. The importance of these three factors is not limited to their effect on the delivered dose of dialysis. Filter clotting, membrane fouling, placement of the replacement solu- tions and other treatment-related factors can also affect drug clearance. Periodic moni- toring of therapeutic drug concentration in the serum and effluent of patients on CRRT would improve drug delivery and maintain its therapeutic efficacy. However, future research is needed in this area to better characterize the effect of technique-specific factors on drug clearance. The sieving properties of the mem- brane can be assessed at any given time by measuring the effluent (FUN and BUN). OPINION Effluent volume and dialysis dose in CRRT: time for reappraisal Etienne Macedo, Rolando Claure-Del Granado and Ravindra L. Mehta Abstract | The results of several studies assessing dialysis dose have dampened the enthusiasm of clinicians for considering dialysis dose as a modifiable factor influencing outcomes in patients with acute kidney injury. Powerful evidence from two large, multicenter trials indicates that increasing the dialysis dose, measured as hourly effluent volume, has no benefit in continuous renal replacement therapy (CRRT). However, some important operational characteristics that affect delivered dose were not evaluated. Effluent volume does not correspond to the actual delivered dose, as a decline in filter efficacy reduces solute removal during therapy. We believe that providing accurate parameters of delivered dose could improve the delivery of a prescribed dose and refine the assessment of the effect of dose on outcomes in critically ill patients treated with CRRT. Macedo, E. et al. Nat. Rev. Nephrol. 8, 57–60 (2012); published online 1 November 2011; doi:10.1038/nrneph.2011.172 Competing interests The authors declare no competing interests. PERSPECTIVES © 2011 Macmillan Publishers Limited. All rights reserved