Acute kidney injury (AKI) requiring renal replacement therapy (RRT) in the ICU is associated with high morbidity and approximately 50% mortality (1–3). Continuous RRT (CRRT) or prolonged intermittent RRT (PIRRT) are the recommended modalities of RRT in patients with hemodynamic instability in the ICU (4–6). Renal replacement therapies may be associated with hypotension with increased need for vasopressor requirements in critically ill patients (7 , 8). Studies in patients undergoing intermittent hemodialysis (IHD) have demonstrated that intradialytic hypotension (IDH) is associated with development of oliguria and anuria, as well as delay in renal recovery (7 , 9 , 10). IDH in patients with end-stage renal disease (ESRD) is also associated with increased risk of myocardial stunning and cardiovascular morbidity and mortality (11 , 12). The frequency of IDH may be related to patient dependent factors such as age, hypovolemia, cardiac dysfunction, intradialytic weight gain, poor nutritional status, and autonomic dysfunction or dialysis dependent factors such as volume and rate of ultrafiltration, changes in plasma osmolality, blood flows, and temperature of dialysate (13 , 14).
Cooler dialysate (35.5–36°C) temperatures, as compared to 37°C (which is considered standard in most ESRD facilities), have been associated with hemodynamic stability during hemodialysis in patients with ESRD (15–18). Cooler dialysate may induce activation of the sympathetic nervous system, vasoconstriction, and preserve central blood volume (13). This vasoconstrictive effect of the cooler dialysate is presumed to be mediated by increased endogenous norepinephrine production and reduced nitric oxide release (19 , 20). Cooler dialysate has also been associated with improved left ventricular contractility as compared to dialysate temperature of 37°C (21).
Studies on the effect of cooler dialysate in patients with AKI undergoing PIRRT or CRRT are sparse (8). One study of 22 patients undergoing continuous venovenous hemofiltration demonstrated that lowering the replacement fluid temperature setting for a short duration was associated with improved mean arterial blood pressure (22). A study by Lima et al (23) compared one group who underwent PIRRT at dialysate temperature of 35.5°C, variable sodium concentrations, and variable ultrafiltration rates with control group with dialysate temperature of 37°C, fixed sodium concentration, and fixed ultrafiltration. The study showed that cooling the dialysate temperature combined with sodium and ultrafiltration profiling was associated with lower episodes of IDH. To our knowledge, no study to date has investigated the sole effects of cooler dialysate on patients with AKI undergoing PIRRT. The objectives of this pilot study is to evaluate the efficacy of lower dialysate temperature in the prevention of IDH in critically ill patients with AKI undergoing PIRRT.
MATERIALS AND METHODS
This was a prospective, randomized cross-over trial to evaluate the effect of a cooler dialysate temperature on the hemodynamic stability of critically ill patients with AKI who required RRT. Inclusion criteria were age 18 years old or older and diagnosis of AKI receiving RRT in the ICU or the heart failure step-down unit. AKI was defined using the kidney disease improving global outcomes definition (5). Exclusion criteria were patients with ESRD or kidney transplant, baseline body temperature less than 36°C, or use of more than two vasopressor agents at time of enrollment. If patients received less than two sessions of PIRRT after enrollment, they were excluded from the analysis.
The study protocol was approved by the Institutional Review Board of Washington University in St. Louis/Barnes Jewish Hospital, with a waiver for informed consent. The trial was registered at ClinicalTrials.gov (NCT03397992). Patients underwent block randomization to two groups based on time of enrollment to the study. Patients assigned to group A started the first PIRRT session with dialysate temperature of 37°C followed by a session with lower dialysate temperature (35°C) and alternating thereafter for a maximum of eight sessions or discontinuation of PIRRT. Patients in group B started with lower dialysate temperature (35°C) for the first session followed by 37°C dialysate temperature and alternating treatments (Fig. 1).
At our facility, PIRRT is used primarily as a transition therapy for patients who are showing hemodynamic improvement on CRRT but not considered completely stable for IHD (24). These patients are usually on a single vasopressor or have a somewhat tenuous hemodynamic status but not requiring a vasopressor (e.g., a patient with mean arterial pressures [MAP] ranging from 55 to 60 mm Hg). PIRRT is performed using the NxStage System One (NxStage Medical, Lawrence, MA) usually at night, with blood flow of 300 mL/min and dialysate flow of 2–8 L/hr. Blood urea nitrogen is obtained before and after each treatment, and urea reduction ratio and single pool Kt/Vurea (K = clearance of dialyzer, t = duration of treatment, and Vurea = volume of distribution of urea) are calculated after each treatment. PIRRT is dosed to achieve a standard weekly Kt/V of 2–4 (24 , 25).
The NxStage System One uses the ComfortMate Fluid Warmer (NxStage System One) which is a simple, dry heat device designed to safely and rapidly warm dialysate or replacement fluids during therapy (Supplemental Fig. 1, Supplemental Digital Content 1, http://links.lww.com/CCM/E159). The manual states that, based on a dialysate flow rate of 150 mL/min, the fluid temperature setting can be increased incrementally to a maximum of 37°C. The control knob has nine settings, but the settings are not marked with a specific temperature. To ensure safety and accuracy of temperature settings for our current study, we performed an in vitro validation of the temperature settings in six different NxStage System One machines with different ComfortMate Fluid Warmers, at point 5 (indicator pointed to 12 O’clock position) and point 9 (indicator pointed to 3 O’clock position) (Supplemental Fig. 1, Supplemental Digital Content 1, http://links.lww.com/CCM/E159). Fluid temperature was measured after at least 15 minutes of infusing normal saline into the ComfortMate Fluid Warmer on a set point. The first 20 cc was aspirated and discarded. Then 60 cc was drawn and measured immediately using PolyScience thermometer. The average temperature when gauge was set point 5 (12 O’clock position) was 35.35 ± 0.6°C, and the average temperature when gauge was set at point 9 (3 O’clock position),was 38.05 ± 0.29°C. However, these temperatures were measured with fluid in a relatively static setting, and actual fluid temperatures during PIRRT are expected to be lower, as dialysate or replacement fluid flow rates are typically above 2 L/hr. For this study, patients were either treated with temperature knob at the highest position (3 O’clock; ≈37°C, per manufacturer) or the lower temperature marked by the temperature knob at the 12 O’clock position (≈35°C).
The NxStage machines were set up for the PIRRT treatment by the hemodialysis nursing staff who placed the warmer in the setting as specified by the randomization. Only one of the investigators (F.Y.E.) and the hemodialysis nurse setting up the machine were aware of the temperature setting. The front of the warmer was covered and obscured from the view of the bedside nurse, the treating physicians, patient, and family, who were all blinded to the temperature setting. Primary endpoint was frequency of IDH, which was defined as decrease in systolic blood pressure by greater than or equal to 20 mm Hg or a decrease in MAP by greater than or equal to 10 mm Hg or the need for intervention (such as IV fluids, initiation or increase of vasoactive drugs, decrease of ultrafiltration rate, or termination of dialysis) to maintain a blood pressure target prespecified by the ICU team. The IV fluids that was used included normal saline, lactated ringers, and 5% albumin. Secondary outcomes were changes in body temperature, total ultrafiltration volume, and adequacy of treatment as measured by Kt/Vurea.
Vital signs were monitored and documented hourly by bedside nurse per ICU protocol during PIRRT. Blood pressure was recorded from an arterial catheter when available or from a noninvasive blood pressure cuff if an arterial catheter is not available. When there was a discrepancy between the oscillometric and calculated MAP, the oscillometric reading was used for analysis. Body temperature (oral, esophageal, or bladder) was measured at least every 4 hours per ICU protocol.
Numeric variables such as age, weight, blood pressure, temperature, and blood-chemistry values are presented as means and SDs. Categorical variables such as sex, race, and presence of comorbidities are presented as percentages. Baseline comparisons between low and high temperature dialysate groups were tested with t tests for the numeric variables and with Pearson chi-square tests or Fisher exact tests for the categorical variables. The effects of covariates on the number of low blood pressure events were measured with Poisson regression using Generalized Estimating Equations, a technique appropriate to longitudinal analyses where the outcome is a count variable (26). The natural log of the duration of each SLED session was used as an offset to adjust for differing length of opportunity for hypotensive events. Model building started with creating a preliminary model using the primary covariates of treatment group (A or B) and dialysate temperature arm and their interaction and the identification of the covariate structure based on the Quasi Information Criterion (QIC). Nonsignificant terms were deleted from the preliminary model, leaving a base model. Using the Generalized Estimating Equation on the base model, it was determined that the autoregressive covariance structure best fit the data by virtue of having the lowest QIC value. Treatment group and the treatment by dialysate temperature interaction were not significant and were removed, leaving a base model containing only the dialysate temperature arm. Each additional covariate was tested by adding it and its interaction terms to this base model. Only one additional covariate was tested at a time. The primary endpoint is presented as incidence rate ratio (IRR) with 95% CIs. The effects of group assignment and dialysate temperature arm and their interaction on continuous outcomes were analyzed with mixed-model repeated-measures analysis of variance. The Akaike Information Criterion was used to select the covariance structure. All analyses and computations were performed using SAS v9.4 (SAS Institute, Cary, NC).
Thirty-eight patients were included in the study and underwent block randomization to either group A or group B. Ten patients in group A and seven patients in group B were excluded from analyses for various reasons (Fig. 1). A total of 10 patients in group A and 11 patients in group B underwent a combined 78 PIRRT treatments and were included in the analysis. Blood flow was maintained at 300 mL/min, and the dialysate flow was 2–7.5 L/hr. Group A and group B were similar at baseline with regards to comorbidities, body temperature, vasopressor requirement, and laboratory variables (Table 1). Neither group assignment nor the interaction of group assignment and temperature arm were significant for any of the continuous outcomes, so the models were rerun with temperature arm as the independent variable.
The rate of hypotensive events more than doubled during PIRRT sessions with dialysate temperature of 37°C compared with 35°C, with IRR of 2.06 (1.49 ± 1.12 vs 0.72 ± 0.69; p ≤ 0.0001) (Table 2). Treatment sessions with cooler dialysate were more likely to achieve target ultrafiltration goals. The mean prescribed ultrafiltration was 2,487 mL in the 37°C dialysate group and 2,305 mL in the 35°C group. The difference between target and achieved ultrafiltration was 423.8 ± 599.9 mL in the higher temperature arm, compared with 190.7 ± 320.9 mL in the group with dialysate temperature of 35°C (p = 0.03) (Table 3). When calculated as a percentage of prescribed goal, sessions with the lower dialysate temperature achieved 92.5% of the prescribed ultrafiltration goal versus 83.8% with dialysate temperature of 37°C (p = 0.04) (Table 3).
The highest body temperature seen during a PIRRT session was significantly greater in the high temperature sessions, but there was no difference in the lowest body temperature during or after treatment between the two groups. No differences were seen with respect to the other continuous outcomes between the temperature arms. The use of vasopressors at the time of enrollment and at the start of the PIRRT session were significantly associated with increase in number of hypotensive events. Higher mean diastolic blood pressure prior to initiation of therapy, sepsis, and liver disease were associated with decrease in hypotensive events (Fig. 2). However, patients with sepsis and liver disease were more likely to have hypotensive events with dialysate temperature of 37°C, but the difference was not statistically significant.
Hemodynamic instability is a common complication of RRT in critically ill patients (8). Hypotension during RRT leads to increased vasopressor use, inadequate delivery of prescribed therapy, and decreased ultrafiltration during RRT with resultant volume overload. Volume overload is independently associated with adverse short-term survival (1 , 27). Hypotension during RRT is also associated with delay in renal recovery (7). Prevention of hypotension during RRT in patients with AKI is essential in preventing adverse outcomes from the treatment. In this blinded cross-over, feasibility study, we demonstrated that lowering dialysate temperature to 35°C is associated with significant reduction in hypotensive events with minimal adverse effects.
In a previous study, Lima et al (23) compared a profiling group with dialysate temperature 35.5°C, variable sodium concentrations and variable ultrafiltration rates with a control group with dialysate temperature 37°C, fixed sodium concentration, and fixed ultrafiltration rate. The profiling group had significantly less hypotension episodes (23% vs 57%; p = 0.009) and achieved higher ultrafiltration volume (2.23 ± 1.25 vs 1.59 ± 1.03 L; p = 0.04) when compared with the control group. Our study suggests that lowering dialysate temperature might be in of itself sufficient to prevent hemodynamic instability during PIRRT, without need for sodium or ultrafiltration profiling.
Mild hypothermia may result in shivering, loss of fine motor coordination, lethargy, and mild confusion and moderate-to-severe hypothermia at less than 32°C may lead to marked depression of brain and cardiovascular function, electrolyte abnormalities, and fatal arrhythmias (28). It is important to note that our patients did not have hypothermia with the cooler dialysate, and there was no significant difference in the lowest temperature between the two arms. Most of our patients were intubated, but we randomly chose two patients, who had completed eight sessions and were not intubated, to ask about symptoms during the treatment. They reported that they felt cold during each of their eight sessions (four with 37.0°C and four with 35.0°C) and were unable to tell any difference in the dialysate temperature. Both patients completed all eight of their treatments despite feeling cold. It should be noted that sensation of being cold is a frequently reported symptom in the ICU irrespective of whether a patient is undergoing RRT (29 , 30).
In our study, sepsis and liver disease was associated with fewer hypotensive events overall. It is possible that patients with sepsis and liver disease already had tenuous blood pressures before PIRRT, and therefore lower target ultrafiltration volumes were prescribed, thereby preventing significant hemodynamic fluctuations during PIRRT. Interestingly, these patients were somewhat immune to the effect of the cooler dialysate in our study. One hypothesis to explain this lack of effect could be that patients with septic shock and liver failure are in a profound vasodilatory state especially involving the hepatosplanchnic circulation and might be more resistant to the mediators of vasoconstriction that is induced by lowering the dialysate temperature (31). However, in a small study evaluating the effects of using cooler substitution fluids during CRRT in nine patients with sepsis, the authors noted a significant reduction in core body temperature (37.9°C to 36.8°C), as well as increase in mean arterial blood pressure (85 –91 mm Hg) (32). In our study, patient’s body temperature did not change significantly during treatments with cooler dialysate and therefore may not have induced a change in blood pressure in patients with sepsis or liver disease.
Our study has some limitations. This is a pilot study performed at a single center, using a particular type of equipment. PIRRT can be performed as a convective or a diffusive modality, and in our study all patients were undergoing sustained low efficiency dialysis. Therefore, it cannot be extrapolated to PIRRT using convection. We did not standardize the measurement of blood pressure and temperature but followed the preexisting ICU protocol. This might explain the lack of statistically significant difference in body temperature between both arms. However, the same blood pressure and temperature monitoring method was used during each single treatment for the individual patient. Strengths of the study includes blinding of treating physicians, patients, families, and bedside nurses to the treatment arm and allowing the treating nephrologists to prescribe and the ICU team to adjust ultrafiltration targets.
The use of cooler dialysate is an effective way to decrease the frequency of IDH during PIRRT. Prevention of hemodynamic instability during RRT can help to achieve the desired ultrafiltration goals in patients with AKI. Patients with sepsis and liver failure patients may require even lower dialysate temperature to show an effect. A larger, multi-center study, using standardized body temperature monitoring techniques, is required to confirm our findings and to evaluate whether hemodynamic stability associated with lower dialysate temperature may lead to differences in outcomes, such as more rapid recovery of renal function.
We would like to thank the dedicated nurses of the acute dialysis unit at Barnes Jewish hospital whose contributions were invaluable to the successful completion of this study. We would like to sincerely thank Yan Yan, MD, PhD, Professor of Clinical Epidemiology and Biostatistics, Washington University in St. Louis, and Tingting Li MD, Associate Professor of Medicine, Division of Nephrology, Washington University in St. Louis for their time and statistical expertise in preparation of this article.
1. Doyle JF, Forni LG. Acute kidney injury
: Short-term and long-term effects. Crit Care 2016; 20:188
2. Bellomo R, Cass A, Cole L, et al; RENAL Replacement Therapy Study Investigators: Intensity of continuous renal-replacement therapy in critically ill patients. N Engl J Med 2009; 361:1627–1638
3. Gaudry S, Hajage D, Schortgen F, et al; AKIKI Study Group: Initiation strategies for renal-replacement therapy in the intensive care unit. N Engl J Med 2016; 375:122–133
4. Palevsky PM, Liu KD, Brophy PD, et al. KDOQI US commentary on the 2012 KDIGO clinical practice guideline for acute kidney injury
. Am J Kidney Dis 2013; 61:649–672
5. Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury
Work Group: KDIGO clinical practice guideline for acute kidney injury
. Kidney Int Suppl 2012; 2:1–138
6. Nash DM, Przech S, Wald R, et al. Systematic review and meta-analysis of renal replacement therapy modalities for acute kidney injury
in the intensive care unit. J Crit Care 2017; 41:138–144
7. Vijayan A, Delos Santos RB, Li T, et al. Effect of frequent dialysis on renal recovery: Results from the Acute Renal Failure Trial Network Study. Kidney Int Rep 2018; 3:456–463
8. Douvris A, Malhi G, Hiremath S, et al. Interventions to prevent hemodynamic instability during renal replacement therapy in critically ill patients: A systematic review. Crit Care 2018; 22:41
9. Schiffl H, Lang SM, Fischer R. Daily hemodialysis and the outcome of acute renal failure. N Engl J Med 2002; 346:305–310
10. Mc Causland FR, Asafu-Adjei J, Betensky RA, et al. Comparison of urine output among patients treated with more intensive versus less intensive RRT: Results from the Acute Renal Failure Trial Network Study. Clin J Am Soc Nephrol 2016; 11:1335–1342
11. Stefánsson BV, Brunelli SM, Cabrera C, et al. Intradialytic hypotension
and risk of cardiovascular disease. Clin J Am Soc Nephrol 2014; 9:2124–2132
12. Burton JO, Jefferies HJ, Selby NM, et al. Hemodialysis-induced cardiac injury: Determinants and associated outcomes. Clin J Am Soc Nephrol 2009; 4:914–920
13. Gul A, Miskulin D, Harford A, et al. Intradialytic hypotension
. Curr Opin Nephrol Hypertens 2016; 25:545–550
14. Sharma S, Waikar SS. Intradialytic hypotension
in acute kidney injury
requiring renal replacement therapy. Semin Dial 2017; 30:553–558
15. Ayoub A, Finlayson M. Effect of cool temperature dialysate on the quality and patients’ perception of haemodialysis. Nephrol Dial Transplant 2004; 19:190–194
16. Cruz DN, Mahnensmith RL, Brickel HM, et al. Midodrine and cool dialysate
are effective therapies for symptomatic intradialytic hypotension
. Am J Kidney Dis 1999; 33:920–926
17. Mustafa RA, Bdair F, Akl EA, et al. Effect of lowering the dialysate temperature in chronic hemodialysis: A systematic review and meta-analysis. Clin J Am Soc Nephrol 2016; 11:442–457
18. Maggiore Q, Pizzarelli F, Santoro A, et al; Study Group of Thermal Balance and Vascular Stability: The effects of control of thermal balance on vascular stability in hemodialysis patients: Results of the European randomized clinical trial. Am J Kidney Dis 2002; 40:280–290
19. Beerenhout CH, Noris M, Kooman JP, et al. Nitric oxide synthetic capacity in relation to dialysate temperature. Blood Purif 2004; 22:203–209
20. Daugirdas JT. Pathophysiology of dialysis hypotension: An update. Am J Kidney Dis 2001; 38:S11–S17
21. Levy FL, Grayburn PA, Foulks CJ, et al. Improved left ventricular contractility with cool temperature hemodialysis. Kidney Int 1992; 41:961–965
22. Robert R, Méhaud JE, Timricht N, et al. Benefits of an early cooling phase in continuous renal replacement therapy for ICU patients. Ann Intensive Care 2012; 2:40
23. Lima EQ, Silva RG, Donadi EL, et al. Prevention of intradialytic hypotension
in patients with acute kidney injury
submitted to sustained low-efficiency dialysis
. Ren Fail 2012; 34:1238–1243
24. Edrees F, Li T, Vijayan A. Prolonged intermittent renal replacement therapy
. Adv Chronic Kidney Dis 2016; 23:195–202
25. Diaz-Buxo JA, Loredo JP. Standard Kt/V: Comparison of calculation methods. Artif Organs 2006; 30:178–185
26. Ballinger GA. Using generalized estimating equations for longitudinal data analysis. Organ Res Methods 2004; 7:127–150
27. Xu J, Shen B, Fang Y, et al. Postoperative fluid overload is a useful predictor of the short-term outcome of renal replacement therapy for acute kidney injury
after cardiac surgery. Medicine (Baltimore) 2015; 94:e1360
28. Danzl DF, Pozos RS. Accidental hypothermia. N Engl J Med 1994; 331:1756–1760
29. Meriläinen M, Kyngäs H, Ala-Kokko T. Patients’ interactions in an intensive care unit and their memories of intensive care: A mixed method study. Intensive Crit Care Nurs 2013; 29:78–87
30. Kalfon P, Mimoz O, Loundou A, et al. Reduction of self-perceived discomforts in critically ill patients in French intensive care units: Study protocol for a cluster-randomized controlled trial. Trials 2016; 17:87
31. Prin M, Bakker J, Wagener G. Hepatosplanchnic circulation in cirrhosis and sepsis. World J Gastroenterol 2015; 21:2582–2592
32. Rokyta R Jr, Matejovic M, Krouzecky A, et al. Effects of continuous venovenous haemofiltration-induced cooling on global haemodynamics, splanchnic oxygen and energy balance in critically ill patients. Nephrol Dial Transplant 2004; 19:623–630
acute kidney injury; cool dialysate; intradialytic hypotension; prolonged intermittent renal replacement therapy; sustained low-efficiency dialysis
Supplemental Digital Content
Copyright © by 2019 by the Society of Critical Care Medicine and Wolters Kluwer Health, Inc. All Rights Reserved.