Good-bye CRRT, here comes SLED? ... not so fast!

In the continuing dispute about the superiority of either intermittent or continuous renal replacement therapy for the critically ill, hybrid methods such as sustained low-efficiency dialysis (SLED) combining the advantages of both modalities - that is, excellent hemodynamic stability and low costs - receive growing attention. The study by Schwenger and colleagues is the first randomized trial indicating that there may be no significant difference in survival at 90 days between patients treated with SLED as compared with those treated with continuous veno-venous hemofiltration.

Th e study by Schwenger and colleagues in a previous issue of Critical Care may add some new aspects to the rapid evolution of renal replacement therapies in the critically ill over the last three decades [1]. Th e fi rst major breakthrough occurred in 1977 when continuous renal replacement therapy (CRRT) was created by the appearance of arterio-venous hemofi ltration, with the original intention of fl uid removal in unstable patients with diuretic resistance [2]. Th is approach turned out to also provide detoxifi cation in uremic patients with superior cardiovascular stability compared with conventional inter mittent hemodialysis available at that time [3]. With the evolution of pump-driven continuous veno-venous hemofi ltration (CVVH), higher ultrafi ltration rates and thus doses became achievable, which could be even further augmented by adding diff usion typical for continuous hemodiafi ltration.
Triggered by the issue of costs, which are mainly driven by the requirement for sterilized solution bags for substitution fl uids or dialysates in CRRT [4,5], the concept of extended daily dialysis was developed using a conventional dialysis machine, with treatment times of around 8 hours allowing for slower fl uid and toxin removal [6]. Next, the invention of a single-batch dialysis system with online dialysate production, providing sustained low-effi ciency dialysis (SLED), allowed one to replace expensive and complex dialysis machines requiring operation by dialysis nurses [7], showing similar hemo dynamic stability and effi cacy in terms of urea removal as CVVH [8].
Th e well-designed single-center study by Schwenger and colleagues is the fi rst randomized trial investigating the outcome of critically ill patients treated by SLED compared with those treated by CVVH at a dose of 35 ml/kg/hour [1]. Achieving the same survival at 90 days, SLED also showed a tendency towards shorter ICU stays and less ventilation days at signifi cantly lower costs than CVVH. Th e study has several implications, however. First, in terms of outcome and cardiovascular stability, SLED and CVVH appear to be quite equivalent. Furthermore, since average treatment durations were quite similar for both modalities (14.9 hours vs. 19.9 hours for SLED and CVVH, respectively), the study indicates that convection provided by hemofi ltration possibly does not confer signifi cant benefi t over diff usion provided by dialysis, which may also be concluded from other trials [9]. Finally, the slightly longer days on ventilation as well as longer ICU stay reported for patients treated with CVVH may refl ect a major disadvantage of CRRT requiring patients to be attached to the extracorporeal circuit around the clock, thus limiting early physiotherapy and mobilization. Th e latter fi ndings, however, were barely statistically signifi cant and defi nitely require further substantiation.
So fi nally, we apparently have a well-tolerated, effi cient and potentially cheap modality at hand. In times of emerging cost restraints, should this not become the new standard for renal replacement therapy in the ICU?
A closer look at the study still leaves us with some hesitation. First of all, dialysis using a batch dialysis system requires a central dialysate preparation unit in a separate room with signifi cant investments. Th e proportional costs for a single renal replacement therapy treatment depend on the number of machines in use as well as the number of treatments, and would be much higher for units with lower frequencies. Furthermore,

Abstract
In the continuing dispute about the superiority of either intermittent or continuous renal replacement therapy for the critically ill, hybrid methods such as sustained low-effi ciency dialysis (SLED) combining the advantages of both modalities -that is, excellent hemodynamic stability and low costs -receive growing attention. The study by Schwenger and colleagues is the fi rst randomized trial indicating that there may be no signifi cant diff erence in survival at 90 days between patients treated with SLED as compared with those treated with continuous venovenous hemofi ltration.
insuffi cient purity of water and contamination of the dialysate have been a reported problem for dialysis units [10].
Th e second issue is the quite low average fi lter survival of roughly 20 hours reported for CVVH by Schwenger and colleagues [1] -implicating frequent clotting and fi lter changes, resulting in increased nursing time for setting up new circuits and enhanced blood loss. Th is observation was probably due to the unusually high fi ltration fractions >40% resulting from low average blood fl ows of 102 ml/minute as well as the use of unfractionated heparin for anticoagulation in the CVVH group. Increasing fi lter survival applying lower fi ltration fractions and using low mole cular weight heparin for anti coagulation probably would have reduced the costs for CVVH [11]. Furthermore, over the last few years regional citrate anticoagulation has become available for all modes of CRRT and has been recommended by the KDIGO 2012 guidelines [12], making average fi lter survival >72 hours easily achievable [13,14], reducing blood transfusion requirements [13] as well as providing better biocompatibility and possibly survival [15,16]. Last, but not least, current dose recommendations are considerably lower, with 20 to 25 ml/kg/hour reducing the amount of substi tution fl uid required. A cost calculation regard ing all these aspects would show far less superiority of SLED over CVVH, if any at all.
Th erefore, although SLED holds some promise for becoming the new low-cost carrier for renal replacement therapy, we still need further stringent economic as well clinical evaluation of SLED compared with CRRT performed in a current state-of-the-art way before any general changes to treatment patterns can be recommended on solid grounds.

Competing interests
MJ received speaker's and consulting honoraria from Baxter, Fresenius and Gambro.