Interventions to prevent hemodynamic instability during renal replacement therapy in critically ill patients: a systematic review

Background Hemodynamic instability related to renal replacement therapy (HIRRT) may increase the risk of death and limit renal recovery. Studies in end-stage renal disease populations on maintenance hemodialysis suggest that some renal replacement therapy (RRT)-related interventions (e.g., cool dialysate) may reduce the occurrence of HIRRT, but less is known about interventions to prevent HIRRT in critically ill patients receiving RRT for acute kidney injury (AKI). We sought to evaluate the effectiveness of RRT-related interventions for reducing HIRRT in such patients across RRT modalities. Methods A systematic review of publications was undertaken using MEDLINE, MEDLINE in Process, EMBASE, and Cochrane’s Central Registry for Randomized Controlled Trials (RCTs). Studies that assessed any intervention’s effect on HIRRT (the primary outcome) in critically ill patients with AKI were included. HIRRT was variably defined according to each study’s definition. Two reviewers independently screened abstracts, identified articles for inclusion, extracted data, and evaluated study quality using validated assessment tools. Results Five RCTs and four observational studies were included (n = 9; 623 patients in total). Studies were small, and the quality was mostly low. Interventions included dialysate sodium modeling (n = 3), ultrafiltration profiling (n = 2), blood volume (n = 2) and temperature control (n = 3), duration of RRT (n = 1), and slow blood flow rate at initiation (n = 1). Some studies applied more than one strategy simultaneously (n = 5). Interventions shown to reduce HIRRT from three studies (two RCTs and one observational study) included higher dialysate sodium concentration, lower dialysate temperature, variable ultrafiltration rates, or a combination of strategies. Interventions not found to have an effect included blood volume and temperature control, extended duration of intermittent RRT, and slower blood flow rates during continuous RRT initiation. How HIRRT was defined and its frequency of occurrence varied widely across studies, including those involving the same RRT modality. Pooled analysis was not possible due to study heterogeneity. Conclusions Small clinical studies suggest that higher dialysate sodium, lower temperature, individualized ultrafiltration rates, or a combination of these strategies may reduce the risk of HIRRT. Overall, for all RRT modalities, there is a paucity of high-quality data regarding interventions to reduce the occurrence of HIRRT in critically ill patients. Electronic supplementary material The online version of this article (10.1186/s13054-018-1965-5) contains supplementary material, which is available to authorized users.

One group conducted two studies (first an observational study and then an RCT) investigating blood volume and temperature controlled dialysis. This strategy is based on automatic biofeedback control for both blood volume and temperature changes. In theory, this should prevent temperature rises and rapid decreases in blood volume that depends on the rate of fluid shift from the interstitial compartment during UF. The first by du Cheyron et al. (2010) was a prospective observational study comparing 20 ICU patients with monitoring devices to historical controls [2]. In their initial analysis, there was a trend towards a significant reduction in HIRRT (21.7% vs 28.7%; p=0.09), and greater net UF was achieved in the treatment group (3.0 +/-0.6L vs 2.1 +/-0.6L; p<0.0001). UF discontinuation occurred similarly in both groups, but the treatment group had less sessions discontinued (4% vs 10%, p=0.02). They also employed a matched case-control analysis using treatment sessions as the unit of analysis, which showed a statistically significant reduction in HIRRT, from 28.8% to16.6% (p=0.005) of dialysis sessions. This group subsequently conducted a three-arm RCT (A=standard; B=blood volume monitor; C=blood volume and temperature monitor) with 74 patients which found no significant difference in the rates of HIRRT between groups with blood volume and temperature monitoring alone [3]. Of note, the IHD parameters used for all patients in this study included a low temperature dialysate of 36.0ºC, a high dialysate sodium concentration (145mmol/L) and a high dialysate calcium bath of 1.75mmol/L. The blood flow speed was initially set at 200-250 mL/min, and adjusted on an unspecified basis.
The retrospective observational study by Lynch et al. (2015) assessed the effect of dialysate sodium modeling alone [4]. Sodium modeling did not reduce the incidence of HIRRT, and had no impact on the amount of UF achieved. Of note, both groups were dialyzed with a 1.25mmol/L calcium bath, but the sodium modeling group was more often dialyzed using cool dialysate (12.0% vs 2.3%, p<0.001).
Schortgen et al. (2000) conducted a before-after study evaluating the effectiveness of implementing a set of guidelines to reduce HIRRT [5] in the context of IHD. The guidelines were outlined in the study but the frequency of which each recommendation was implemented was not reported. These guidelines included using modified cellulosic Despite this, the 'after guideline implementation' group had a lower incidence of hemodynamic impairment, defined as a composite of reduction in SBP, intravenous fluid infusion or administration of vasoactive drugs (71% vs 61%, p=0.015). The study also assessed drops in systolic blood pressure according to IHD session timing and found that the 'after guideline implementation' group had fewer drops in SBP both at session onset (33% vs 21%, p=0.002) and during sessions (68% vs 56%, p=0.002). As well, it was found that the observed mortality rate in the 'after guideline implementation' group was less than expected, and their median length of stay was significantly shorter than had been for the 'before guideline implementation' group (11 versus 7 days).

Sustained Low Efficiency Dialysis
We included two studies of different interventions in patients treated with slow low efficiency dialysis (SLED), both of which were RCTs.

Continuous Renal Replacement Therapy
Two studies tested manoeuvres to prevent HIRRT with continuous renal replacement therapy (CRRT) (RCT: n=1). Studies are outlined in Tables 1, 2 and 3 (main paper) and outcomes according to intervention are provided in Table S2. Robert et al. (2012) conducted a prospective crossover randomized study comparing the effect of setting the temperature of a heating device warming blood return from CRRT at 38˚C or 36˚C on hemodynamic parameters during treatment [8]. Body temperature was similar between groups for the duration of the study (36.3+/-0.5˚C and 36.4+/-0.5˚C); however, reducing the device temperature from 38˚C to 36˚C at the beginning of CRRT in group B was shown to significantly increase mean arterial pressure and decrease vasopressor requirements as compared to maintaining the heating temperature at 38˚C. This hemodynamic effect of lower temperature was not seen in the second phase of the study, where temperature was lowered from 38˚C to 36˚C at six hours after initiation of CRRT.  Table 2). They did not record any episodes of hypotension during the start of treatment in either group [9]. This study was focused on hemodynamics at CRRT initiation, and did not record outcomes beyond thirty minutes into the session.