Share this article on:

Low- Versus High-Chloride Content Intravenous Solutions for Critically Ill and Perioperative Adult Patients: A Systematic Review and Meta-analysis

Kawano-Dourado, Leticia MD*†; Zampieri, Fernando G. MD*‡; Azevedo, Luciano C. P. MD§‖; Corrêa, Thiago D. MD; Figueiró, Mabel BLS*; Semler, Matthew W. MD#; Kellum, John A. MD**; Cavalcanti, Alexandre B. MD*

doi: 10.1213/ANE.0000000000002641
Critical Care and Resuscitation: Meta-Analysis

BACKGROUND: To assess whether use of low-chloride solutions in unselected critically ill or perioperative adult patients for maintenance or resuscitation reduces mortality and renal replacement therapy (RRT) use when compared to high-chloride fluids.

METHODS: Systematic review and meta-analysis with random-effects inverse variance model. PubMed, Cochrane library, EMBASE, LILACS, and Web of Science were searched from inception to October 2016. Published and unpublished randomized controlled trials in any language that enrolled critically ill and/or perioperative adult patients and compared a low- to a highchloride solution for volume maintenance or resuscitation. The primary outcomes were mortality and RRT use. We conducted trial sequential analyses and assessed risk of bias of individual trials and the overall quality of evidence. Fifteen trials with 4067 patients, most at low risk of bias, were identified. Of those, only 11 and 10 trials had data on mortality and RRT use, respectively. A total of 3710 patients were included in the mortality analysis and 3724 in the RRT analysis.

RESULTS: No statistically significant impact on mortality (odds ratio, 0.90; 95% confidence interval, 0.69–1.17; P = .44; I 2 = 0%) or RRT use (odds ratio, 1.12; 95% confidence interval, 0.80–1.58; P = .52; I 2 = 0%) was found. Overall quality of evidence was low for both primary outcomes. Trial sequential analyses highlighted that the sample size needed was much larger than that available for properly powered outcome assessment.

CONCLUSIONS: The current evidence on low- versus high-chloride solutions for unselected critically ill or perioperative adult patients demonstrates no benefit, but suffers from considerable imprecision. We noted a limited exposure volume for study fluids and a relatively low risk of the populations in each study. Together with the relatively small pooled sample size, these data leave us underpowered to detect potentially important differences. Results from well-conducted, adequately powered randomized controlled trials examining sufficiently large fluid exposure are necessary.

Supplemental Digital Content is available in the text.Published ahead of print November 14, 2017.

From the *Research Institute, Hospital do Coração (HCor), São Paulo, Brazil; Pulmonary Division, Heart Institute (InCor), University of Sao Paulo Medical School, São Paulo, Brazil; Intensive Care Unit, Hospital Alemão Oswaldo Cruz, São Paulo, Brazil; §Intensive Care Unit, Hospital Sirio Libanes, São Paulo, Brazil; Emergency Medicine Discipline, University of São Paulo Medical School, São Paulo, Brazil; Intensive Care Unit, Hospital Israelita Albert Einstein, São Paulo, Brazil; #Division of Allergy, Pulmonary, and Critical Care Medicine, Vanderbilt University Medical Center, Nashville, Tennessee; and **The Center for Critical Care Nephrology, Department of Critical Care Medicine, University of Pittsburgh, Pittsburgh, Pennsylvania.

Published ahead of print November 14, 2017.

Accepted for publication October 6, 2017.

Funding: This meta-analysis was supported by the Brazilian Ministry of Health.

The authors declare no conflicts of interest.

Supplemental digital content is available for this article. Direct URL citations appear in the printed text and are provided in the HTML and PDF versions of this article on the journal’s website.

Reprints will not be available from the authors.

Address correspondence to Leticia Kawano-Dourado, MD, Research Institute - Hospital do Coracao (HCor), Rua Abilio Soares 250, 12o andar, cep: 04005-000, São Paulo-SP, Brazil. Address e-mail to ldourado@hcor.com.br.

KEY POINTS

  • Question: Do low-chloride solutions reduce mortality and renal replacement therapy use when compared to high-chloride fluids?
  • Findings: No impact on mortality or renal replacement therapy use was found. However, overall quality of evidence was low and the small available pooled sample size left this meta-analysis underpowered to detect potentially important differences.
  • Meaning: Larger randomized controlled trials are needed to address the issue.

Normal saline (0.9% sodium chloride), the most commonly prescribed intravenous crystalloid solution worldwide, contains supraphysiologic concentrations of chloride (154 mmol/L), with a strong ion difference of zero.1 Balanced or “buffered” solutions, on the other hand, have a lower sodium and chloride content and a positive strong ion difference, with an electrochemical composition that more closely approximates to the extracellular fluid1

In experimental studies, resuscitation with normal saline, but not with low-chloride solutions, leads to hyperchloremic metabolic acidosis, increased inflammatory response, coagulopathy, derangements in renal perfusion, and acute kidney injury (AKI).2–10 Clinically, randomized controlled trials (RCTs) have found that saline leads to hyperchloremic acidosis11–13; however, evidence that high-chloride solutions adversely affect clinical outcomes is mixed. Some studies suggest that, when compared to high-chloride solutions, low-chloride solutions reduce the risk of AKI, use of renal replacement therapy (RRT), coagulopathy, use of blood transfusion, and mortality, while others have not shown benefit.11–27

The objective of this meta-analysis is to synthesize the current medical literature on low- versus high-chloride content solutions for maintenance and fluid resuscitation in critically ill and/or perioperative adult patients and to clarify whether low-chloride content solutions should be preferred over normal saline in critical and perioperative care settings. Our specific aim was to determine whether the use of high- versus low-chloride solutions affects mortality and the use of RRT in unselected critically ill and perioperative adult patients.

Back to Top | Article Outline

METHODS

This systematic review was conducted according to our protocol registered at the international prospective register of systematic reviews (PROSPERO; no. CRD42016039556). The results are reported following Preferred Reporting Items for Systematic Reviews and Meta-Analyses guidelines (Supplemental Digital Content, Table 1 in the Appendix, http://links.lww.com/AA/C135).28

Back to Top | Article Outline

Eligibility Criteria

We considered only RCTs for inclusion. Unpublished trials and trials published in abstract format were also considered for inclusion if adequate information regarding methods and results could be retrieved from authors. Randomized crossover (on the individual level) trials, observational studies, experimental nonrandomized studies, or case series were excluded.

Only trials involving adult critically ill or perioperative patients receiving intravenous high- versus low-chloride content solutions for intravascular volume expansion or maintenance were considered. Trials focusing on cesarean delivery were excluded. Trials with critically ill patients outside the intensive care unit (ICU), for example, emergency department, were included.

To minimize confounding factors, we only considered trials where the difference between the experimental and control arms involved a buffer in the solution (usually lactate, gluconate, and/or acetate), leading to a low-chloride concentration in the balanced solution. We excluded studies that compared crystalloids versus colloids, studies that compared hypertonic saline versus balanced solutions, or studies that compared fluids with different colloid components. Trials with 3 or more arms were included if the criteria were satisfied in at least 2 arms.

Our primary outcomes were all-cause mortality and use of RRT. The longest follow-up period reported was recorded for mortality. Hospital discharge was used to define the window for RRT. Secondary outcomes include the following: AKI, defined by Kidney Disease: Improving Global Outcomes stage ≥2 (or similar per individual study definitions)29; use of RRT at hospital discharge; allogenic blood transfusion; volume of blood products transfused; ICU length of stay (LOS); hospital LOS; number of days on mechanical ventilation and/or days free of mechanical ventilation; and number of days on vasopressors and/or vasopressor-free days. For all secondary outcomes, the longest follow-up period available was considered, unless truncated at hospital discharge.

Back to Top | Article Outline

Search Strategy

We searched the following electronic databases (from inception to October 2016): PubMed/MEDLINE, EMBASE, LILACS, Cochrane Library, and Web of Science. We placed no language restrictions, and we used controlled vocabulary whenever possible (MeSH term for MEDLINE and CENTRAL; EMTREE for EMBASE). We used keywords and their synonyms to sensitize the search, and we applied standard filters for the identification of RCTs (Supplemental Digital Content, Appendix, http://links.lww.com/AA/C135). We adapted our MEDLINE search strategy for use on other electronic databases (EMBASE, LILACS, Cochrane Library, and Web of Science). Additionally, we hand searched the reference lists of the included studies to identify other relevant trials. Finally, we attempted to identify unpublished or ongoing trials by contacting experts in the field and by searching a clinical trial registry (ClinicalTrials.gov).

Back to Top | Article Outline

Study Selection, Data Extraction, and Bias Risk Assessment

Two pairs of authors independently screened all retrieved citations by reviewing their titles and abstracts (L.C.P.A., T.D.C. and L.K.-D., A.B.C.). Then, 2 pairs of reviewers independently evaluated the full-text manuscripts for eligibility using a standardized form (L.C.P.A., T.D.C. and L.K.-D., F.G.Z.).

Two pairs of reviewers independently extracted the relevant data from the full-text manuscripts and assessed the risk of bias using a standardized form (L.C.P.A., T.D.C. and L.K.-D., F.G.Z.). Any disagreement between authors was resolved by a third author (L.K.D. or A.B.C.). We followed specific instructions of the Cochrane Collaboration’s tool for assessing risk of bias in each randomized trial.30

Back to Top | Article Outline

Summary Measures and Synthesis of Results

We performed a meta-analysis of dichotomous outcomes using the inverse variance method with random-effects model.31 Two studies included in this systematic review were cluster randomized trials. For meta-analyses of primary outcomes, odds ratios (ORs) adjusted for cluster effect when applicable were considered, and adjusted pooled OR with 95% confidence interval (95% CI) was reported.21 , 22 Due to the expected small number of trials, a network meta-analysis was not planned; nevertheless, we planned to perform subgroup analysis for the 2 main outcomes according to the balanced solution used (lactated Ringer or Plasma-Lyte). A P value of <.05 was considered significant for all analysis. We have not adjusted the significance criterion for having 2 coprimary end points or the secondary end points

To assess the meta-analysis power and to obtain a reliable estimate of the required sample size to detect the effects of the intervention on our 2 primary outcomes (mortality and the use of RRT) and for AKI, trial sequential analysis (TSA) was performed.32 This analysis was performed using the required sample size to construct sequential monitoring boundaries. The boundaries were established to limit the global type I error to 5% and were calculated with the O’Brien-Fleming function, considering a power of 80% to detect a relative 10% decrease in mortality and in RRT use. The baseline outcome incidence was set as 30% for hospital mortality and 15% for RRT.33 , 34

We used meta-regression with random-effects model to assess the relationship between mean volume of administered fluids and the effect on mortality.

Publication bias was assessed by inspection of funnel plot of included studies.

Back to Top | Article Outline

Quality of Meta-analysis Evidence

The quality of evidence generated by this meta-analysis was classified in accordance with the Grading of Recommendations Assessment, Development and Evaluation (GRADE) system.35

Back to Top | Article Outline

RESULTS

Study Selection

Figure 1

Figure 1

Our search strategy identified 3554 references after excluding duplicates (Figure 1). After screening titles and abstracts, 65 full-text articles were selected. A total of 36 studies were considered for data extraction. We contacted all authors and 3 responded.18 , 36 , 37 Of those, 21 studies did not report any outcome of interest and were excluded. Thus, 15 trials with 4067 patients were included in the systematic review, although only 11 studies (3710 patients in total) had available data on mortality and 10 studies (3724 patients in total) had available data on RRT use.

Back to Top | Article Outline

Characteristics of Included Trials

Table 1

Table 1

The characteristics of the 15 included studies are shown in Table 1. Most of studies were small, ranging from 30 to 67 patients per study, except for 2 studies with 2262 and 974 patients, respectively.21 , 22 Eighty-five percent (3468/4067) of the meta-analysis sample comprised critically ill patients due to the contribution of the 2 largest studies that were performed in the ICU setting.21 , 22 Three studies did not report the volume of fluid infused.13 , 16 , 39 The majority of the studies, including the 2 largest studies included,21 , 22 had a low, 2–3 L study fluid exposure.20–22 , 36–38 , 40 , 41 Volumes of 4 L or more were administered in 4 trials that contributed only 5.2% (215/4067) of the patients in our meta-analysis.11 , 12 , 18 , 19 Volumes around 7 L (or above) were administered to only 3.2% (131/4067) of the patients in our meta-analysis.11 , 19

Back to Top | Article Outline

Risk of Bias

Table 2

Table 2

Figure 2

Figure 2

In general, the majority of the studies were at low risk of bias (Figure 2), except for 3 studies,12 , 36 , 37 with very small impact on pooled estimates, given their small sample sizes (Table 2). A funnel plot for both main outcomes assessed in the meta-analysis is shown in Supplemental Digital Content, Figure 1 in the Appendix, http://links.lww.com/AA/C135. No clear evidence of publication bias was found when inspecting funnel plots.

Back to Top | Article Outline

Effects on Outcomes

Pooled data showed no effect of low- versus high-chloride solutions on all-cause mortality (OR, 0.90; 95% CI, 0.69–1.17; P = .44; I 2 = 0%) or on the use of RRT (OR, 1.12; 95% CI, 0.80–1.58; P = .52; I 2 = 0%). Forest plots for both outcomes are shown in Figure 3. We were unable to compare the pooled OR for lactated Ringer versus Plasma-Lyte since only 1 trial with lactated Ringer reported nonzero events. Pooled data also found no effect of chloride on the need for RRT (OR, 0.49; 95% CI, 0.08–2.83; P = .428) for studies that used lactated Ringer and 1.17 (95% CI, 0.82–1.69; P = .383) for Plasma-Lyte studies.12 , 16 , 20–22 , 41

Figure 3

Figure 3

Results for all secondary outcomes are shown in Table 2. AKI risk or allogenic blood transfusion rates did not differ between groups. Data for ICU and hospital LOS, mechanical ventilation, vasopressor use, and RRT use after hospital discharge were not pooled since these outcomes were not reported by enough studies or the reported metrics were not synthesizable. Use of allogenic blood transfusion was a synthesizable metric because most studies reported this outcome in the same metric (number of patients receiving transfusion). Data from Semler et al22 were excluded because that study reported the use of allogenic blood transfusion in milliliters (12 ± 114 vs 40 ± 376 mL, P = .16 for balanced versus saline group, respectively). None of the individual trials showed statistically significant effects of fluid type on any of the secondary outcomes (Table 2).

No statistical heterogeneity was detected in the effects of low- versus high-chloride content solutions on both mortality and RRT use (Figure 3). Nevertheless, clinical heterogeneity was observed in the type of solution, in follow-up period, in purpose of fluid administration (expansion and/or maintenance fluid), in type of patients and duration of exposure to the intervention (perioperative patients; ICU patients), and in total volume administered.

Back to Top | Article Outline

Trial Sequential Analysis

TSA was performed for mortality, use of RRT, and AKI incidence. Neither the curve for mortality nor for RRT or AKI crossed the boundaries of superiority or inferiority. Available data were insufficient to calculate the futility boundary. Optimal sample size was estimated in 9517 patients for mortality and 22,826 for RRT (Supplemental Digital Content, Figures 2 and 3 in the Appendix, http://links.lww.com/AA/C135). For AKI, the information size was over 12,000 patients.

Back to Top | Article Outline

Quality of Evidence (GRADE)

We classified the quality of evidence generated by the meta-analysis for the primary outcomes as low. Our primary reason for downgrading the quality of evidence was imprecision and indirectness. Most studies were small and with few events, except for the trials of Semler et al22 and Young et al21 (Supplemental Digital Content, Table 2 in the Appendix, http://links.lww.com/AA/C135), and in the majority of the studies, a low dose of balanced crystalloid was administered to relatively low-risk patients; therefore, high-risk patients who need a significant amount of fluid were not directly represented in these studies.

Back to Top | Article Outline

DISCUSSION

In this systematic review and meta-analysis, we found no difference between low- versus high-chloride solutions on mortality or RRT in critically ill and perioperative patients. The quality of evidence (GRADE) was considered low due to imprecision and indirectness. Chloride use also did not affect AKI (Kidney Disease: Improving Global Outcomes ≥2) or use of allogenic blood transfusion in our meta-analysis.

Previous meta-analyses have addressed this subject but were unable to draw conclusions regarding clinically meaningful outcomes due to small number of patients available for data synthesis, limited number of studies reporting on clinically meaningful outcomes, and limited exposure to fluids.23–26 , 42 After publication of these meta-analyses, 2 additional RCTs became available for inclusion.21 , 22 The present meta-analysis included mortality and RRT data from these studies, adjusted for their cluster design, but remained inadequately powered to identify clinically meaningful differences. The cumulative number of patients considered in this systematic review confers a power of only 42% and 23% to detect a relative risk reduction of 10% on mortality and RRT, respectively. For instance, with a patient population typical of existing studies (including low fluid exposure level), detecting a 10% decrease in mortality would require 9517 patients (Supplemental Digital Content, Figure 2 in the Appendix, http://links.lww.com/AA/C135).

Although our study found no effect of chloride, previous observational studies have observed that use with high-chloride solutions, mainly normal saline, has been associated with a higher incidence of AKI, increased use of RRT, and/or increased mortality.14 , 43 Several mechanisms may explain the lack of concordance between RCTs and observational studies suggested by our data.14 , 43 First, study populations may differ in their susceptibility to nephrotoxicity and coagulopathy. Second, unlike observational cohorts, clinical trials may exclude the most severely ill patients and thus falsely reduce effect size. Third, it is difficult to control all fluids used in RCTs. Some studies controlled most fluid (both resuscitation and maintenance) but still could not control for fluids given outside the study period. Fourth, the fluid exposure in RCTs was low (median volume of study fluid was 2 L) and thus less likely to produce any clinically meaningful effects. One 2016 single-center study suggested that toxic clinical effects of high-chloride solutions may not be significant until approximately 7–8 L.22 , 27 In the pragmatic 2016 SALT randomized trial, mortality in the unbalanced (high chloride) group increased linearly for volumes infused higher than 6 L.22 Because 94.7% (3852/4067) of our meta-analysis cohort received between 2 and 3 L of study fluid, we were unable to identify any definitive relationship between dose of fluid exposure and effect size.44 We also observe that the rate of infusion may have played a role, although our data were insufficient to clarify that point.

The strengths of this meta-analysis include a comprehensive literature search according to a registered protocol and well-validated methods for literature search, trial eligibility, data extraction, and assessment of risk of bias. We focused on primary outcomes (mortality and use of RRT) that are clinically meaningful and directly related to the biological rationale of hyperchloremic toxicity. Additionally, we conducted a structured assessment of the quality of evidence according to the GRADE system and used a TSA to limit random error. Our TSA yielded conservative estimates: if mortality or use of RRT baseline incidence was set at lower levels (lower than 30% and 15%, respectively), that would have yielded an even higher optimal sample size. In fact, if the studies included in the meta-analysis were pooled as a single study, its power to detect the observed effect sizes for mortality and RRT would be low (11.3% for mortality and 8.7% for RRT).

The Balanced Solution Versus Saline in Intensive Care Study (BaSICS trial; NCT02875873), Plasma-Lyte 148 versUs Saline Study trial (NCT02721654), and Isotonic Solutions and Major Adverse Renal Events Trial (NCT02444988, NCT02547779) have been designed to determine the relationship between intravenous fluid chloride content and mortality at low fluid exposure levels (2–3 L).45–48 BaSICS trial is designed to recruit 11,000 patients, while Plasma-Lyte 148 versUs Saline Study trial is designed to recruit 8000 patients and that can provide them with enough power to evaluate whether high-chloride content solutions are toxic at low fluid exposure levels. Furthermore, the relationship between the rate of infusion and mortality in the BaSICS trial will also be investigated, thereby providing information on potential relevant associations for fluid therapy in the critically ill.

Our meta-analysis has limitations. Most of the included studies were small and focused on specific populations. The largest and most recent RCT included was a pilot study for future larger RCT. Our meta-analysis thus provides evidence that all the large ongoing RCTs are indeed necessary (since no clear information can be obtained by the available literature) and that their large estimated sample size is truly needed. Studies also differed considerably regarding fluid delivery protocol, duration of the intervention, and follow-up time on mortality and RRT. The absence of data on time until study start in most trials we reviewed prevents us from assessing the timing of the intervention, which may represent an important source of heterogeneity. It is possible, for example, that the beneficial effects of balanced solutions may be limited to scenarios when the solution is administered early in the disease course. Another important limitation is that death and AKI with need for RRT may affect each other; that is, the longer survival durations may have increased the likelihood of needing RRT. Nevertheless, survival data and time to RRT were unavailable in all included studies, thus making a competing risk meta-analysis impossible. Another important limitation was our inability to identify specific subgroups where the intervention may have had benefit, including strata of illness severity. Such questions should be addressed in future RCTs.

Back to Top | Article Outline

CONCLUSIONS

Our meta-analysis assessing low- versus high-chloride content solutions on critically ill and perioperative patients did not identify any effect on mortality and on RRT use. We also noted a limited exposure volume for study fluids and a relatively low risk of the populations in each study. Together with the relatively small pooled sample size, these data leave us underpowered to detect potentially important differences. Larger RCTs including more critically ill patients and larger study fluid volumes are needed to address the issue.

Back to Top | Article Outline

ACKNOWLEDGMENTS

The research team would like to thank Takil et al, Volta et al, and Song et al for providing us with additional information on mortality and renal replacement therapy from their studies.

Back to Top | Article Outline

DISCLOSURES

Name: Leticia Kawano-Dourado, MD.

Contribution: This author helped conceive and design the study; analyze and interpret the data; and write the manuscript. This author was also actively involved in the original research project.

Name: Fernando G. Zampieri, MD.

Contribution: This author helped conceive and design the study; analyze and interpret the data; and write the manuscript. This author was also actively involved in the original research project.

Name: Luciano C. P. Azevedo, MD.

Contribution: This author helped analyze and interpret the data; and critically revise the manuscript. This author was also actively involved in the original research project.

Name: Thiago D. Corrêa, MD.

Contribution: This author helped analyze and interpret the data; and critically revise the manuscript. This author was also actively involved in the original research project.

Name: Mabel Figueiró, BLS.

Contribution: This author helped acquire and analyze the data; and critically revise the manuscript.

Name: Matthew W. Semler, MD.

Contribution: This author helped analyze and interpret the data; and critically revise the manuscript.

Name: John A. Kellum, MD.

Contribution: This author helped analyze and interpret the data; and critically revise the manuscript.

Name: Alexandre B. Cavalcanti, MD.

Contribution: This author helped conceive and design the study; analyze and interpret the data; and write the manuscript. This author was also actively involved in the original research project.

This manuscript was handled by: Avery Tung, MD, FCCM.

Back to Top | Article Outline

REFERENCES

1. Myburgh JA, Mythen MG. Resuscitation fluids. N Engl J Med. 2013;369:2462–2463.
2. Healey MA, Davis RE, Liu FC, Loomis WH, Hoyt DB. Lactated Ringer’s is superior to normal saline in a model of massive hemorrhage and resuscitation. J Trauma. 1998;45:894–899.
3. Watters JM, Brundage SI, Todd SR. Resuscitation with lactated Ringer’s does not increase inflammatory response in a Swine model of uncontrolled hemorrhagic shock. Shock. 2004;22:283–287.
4. Todd SR, Malinoski D, Muller PJ, Schreiber MA. Lactated Ringer’s is superior to normal saline in the resuscitation of uncontrolled hemorrhagic shock. J Trauma. 2007;62:636–639.
5. Noritomi DT, Pereira AJ, Bugano DD, Rehder PS, Silva E. Impact of Plasma-Lyte pH 7.4 on acid-base status and hemodynamics in a model of controlled hemorrhagic shock. Clinics (Sao Paulo). 2011;66:1969–1974.
6. Rohrig R, Rönn T, Lendemans S, Feldkamp T, de Groot H, Petrat F. Adverse effects of resuscitation with lactated Ringer compared with Ringer solution after severe hemorrhagic shock in rats. Shock. 2012;38:137–145.
7. Aksu U, Bezemer R, Yavuz B, Kandil A, Demirci C, Ince C. Balanced vs unbalanced crystalloid resuscitation in a near-fatal model of hemorrhagic shock and the effects on renal oxygenation, oxidative stress, and inflammation. Resuscitation. 2012;83:767–773.
8. Martini WZ, Cortez DS, Dubick MA. Comparisons of normal saline and lactated Ringer’s resuscitation on hemodynamics, metabolic responses, and coagulation in pigs after severe hemorrhagic shock. Scand J Trauma Resusc Emerg Med. 2013;21:86.
9. Rohrig R, Wegewitz C, Lendemans S, Petrat F, de Groot H. Superiority of acetate compared with lactate in a rodent model of severe hemorrhagic shock. J Surg Res. 2014;186:338–345.
10. Zhou F, Peng ZY, Bishop JV, Cove ME, Singbartl K, Kellum JA. Effects of fluid resuscitation with 0.9% saline versus a balanced electrolyte solution on acute kidney injury in a rat model of sepsis*. Crit Care Med. 2014;42:e270–e278.
11. Young JB, Utter GH, Schermer CR. Saline versus Plasma-Lyte A in initial resuscitation of trauma patients: a randomized trial. Ann Surg. 2014;259:255–262.
12. O’Malley CM, Frumento RJ, Hardy MA. A randomized, double-blind comparison of lactated Ringer’s solution and 0.9% NaCl during renal transplantation. Anesth Analg. 2005;100:1518–1524.
13. Khajavi MR, Etezadi F, Moharari RS. Effects of normal saline vs. lactated Ringer’s during renal transplantation. Ren Fail. 2008;30:535–539.
14. Yunos NM, Bellomo R, Hegarty C, Story D, Ho L, Bailey M. Association between a chloride-liberal vs chloride-restrictive intravenous fluid administration strategy and kidney injury in critically ill adults. JAMA. 2012;308:1566–1572.
15. Smith CA, Gosselin RC, Utter GH. Does saline resuscitation affect mechanisms of coagulopathy in critically ill trauma patients? An exploratory analysis. Blood Coagul Fibrinolysis. 2015;26:250–254.
16. Wu BU, Hwang JQ, Gardner TH. Lactated Ringer’s solution reduces systemic inflammation compared with saline in patients with acute pancreatitis. Clin Gastroenterol Hepatol. 2011;9:710–717.e1.
17. Shaw AD, Schermer CR, Lobo DN. Impact of intravenous fluid composition on outcomes in patients with systemic inflammatory response syndrome. Crit Care. 2015;19:334.
18. Takil A, Eti Z, Irmak P, Yilmaz Göğüş F. Early postoperative respiratory acidosis after large intravascular volume infusion of lactated Ringer’s solution during major spine surgery. Anesth Analg. 2002;95:294–298.
19. Waters JH, Gottlieb A, Schoenwald P, Popovich MJ, Sprung J, Nelson DR. Normal saline versus lactated Ringer’s solution for intraoperative fluid management in patients undergoing abdominal aortic aneurysm repair: an outcome study. Anesth Analg. 2001;93:817–822.
20. Kim SY, Huh KH, Lee JR, Kim SH, Jeong SH, Choi YS. Comparison of the effects of normal saline versus Plasmalyte on acid-base balance during living donor kidney transplantation using the Stewart and base excess methods. Transplant Proc. 2013;45:2191–2196.
21. Young P, Bailey M, Beasley R, et al; SPLIT Investigators; ANZICS CTG. Effect of a buffered crystalloid solution vs saline on acute kidney injury among patients in the intensive care unit: the SPLIT randomized clinical trial. JAMA. 2015;314:1701–1710.
22. Semler MW, Wanderer JP, Ehrenfeld JM, et al; SALT Investigators and the Pragmatic Critical Care Research Group (2016). Balanced crystalloids versus saline in the intensive care unit: the SALT randomized trial. Am J Crit Care Med. 2016;195:1362–1372.
23. Burdett E, Dushianthan A, Bennett-Guerrero E. Perioperative buffered versus non-buffered fluid administration for surgery in adults. Cochrane Database Syst Rev. 2012;12:CD004089.
24. Krajewski ML, Raghunathan K, Paluszkiewicz SM, Schermer CR, Shaw AD. Meta-analysis of high- versus low-chloride content in perioperative and critical care fluid resuscitation. Br J Surg. 2015;102:24–36.
25. Rochwerg B, Alhazzani W, Sindi A, et al; Fluids in Sepsis and Septic Shock Group. Fluid resuscitation in sepsis: a systematic review and network meta-analysis. Ann Intern Med. 2014;161:347–355.
26. Rochwerg B, Alhazzani W, Gibson A; FISSH Group (Fluids in Sepsis and Septic Shock). Fluid type and the use of renal replacement therapy in sepsis: a systematic review and network meta-analysis. Intensive Care Med. 2015;41:1561–1571.
27. Zampieri FG, Ranzani OT, Azevedo LC, Martins ID, Kellum JA, Libório AB. Lactated Ringer is associated with reduced mortality and less acute kidney injury in critically ill patients: a retrospective cohort analysis. Crit Care Med. 2016;44:2163–2170.
28. Stewart LA, Clarke M, Rovers M; PRISMA-IPD Development Group. Preferred reporting items for systematic review and meta-analyses of individual participant data: the PRISMA-IPD statement. JAMA. 2015;313:1657–1665.
29. Khwaja A. KDIGO clinical practice guidelines for acute kidney injury. Nephron Clin Pract. 2012;120:c179–c184.
30. Higgins JP, Altman DG, Gøtzsche PC; Cochrane Bias Methods Group; Cochrane Statistical Methods Group. The Cochrane Collaboration’s tool for assessing risk of bias in randomised trials. BMJ. 2011;343:d5928.
31. DerSimonian R, Laird N. Meta-analysis in clinical trials. Control Clin Trials. 1986;7:177–188.
32. Thorlund K, Engstrøm J, Wetterslev J, Brok J, Imberger G, Gluud C. User Manual for Trial Sequential Analysis (TSA). Copenhagen, Denmark: Copenhagen Trial Unit, Centre for Clinical Intervention Research; 2011: 1–115. Available at: http://www.ctu.dk/tsa/files/TSA%20manual.pdf. Accessed October 10, 2016.
33. Soares M, Silva UV, Homena WS, et al; ORCHESTRA (ORganizational CHaractEeriSTics in cRitcal cAre) Study Investigators. Family care, visiting policies, ICU performance, and efficiency in resource use: insights from the ORCHESTRA study. Intensive Care Med. 2017;43:590–591.
34. Hoste EA, Bagshaw SM, Bellomo R. Epidemiology of acute kidney injury in critically ill patients: the multinational AKI-EPI study. Intensive Care Med. 2015;41:1411–1423.
35. Guyatt GH, Oxman AD, Schünemann HJ, Tugwell P, Knottnerus A. GRADE guidelines: a new series of articles in the Journal of Clinical Epidemiology. J Clin Epidemiol. 2011;64:380–382.
36. Volta CA, Trentini A, Farabegoli L. Effects of two different strategies of fluid administration on inflammatory mediators, plasma electrolytes and acid/base disorders in patients undergoing major abdominal surgery: a randomized double blind study. J Inflamm (Lond). 2013;10:29.
37. Song JW, Shim JK, Kim NY, Jang J, Kwak YL. The effect of 0.9% saline versus Plasmalyte on coagulation in patients undergoing lumbar spinal surgery; a randomized controlled trial. Int J Surg. 2015;20:128–134.
38. Base EM, Standl T, Lassnigg A. Efficacy and safety of hydroxyethyl starch 6% 130/0.4 in a balanced electrolyte solution (Volulyte) during cardiac surgery. J Cardiothorac Vasc Anesth. 2011;25:407–414.
39. Van Zyl DG, Rheeder P, Delport E. Fluid management in diabetic-acidosis–Ringer’s lactate versus normal saline: a randomized controlled trial. QJM. 2012;105:337–343.
40. Potura E, Lindner G, Biesenbach P. An acetate-buffered balanced crystalloid versus 0.9% saline in patients with end-stage renal disease undergoing cadaveric renal transplantation: a prospective randomized controlled trial. Anesth Analg. 2015;120:123–129.
41. Verma B, Luethi N, Cioccari L. A multicentre randomised controlled pilot study of fluid resuscitation with saline or Plasma-Lyte 148 in critically ill patients. Crit Care Resusc. 2016;18:205–212.
42. Wan S, Roberts MA, Mount P. Normal saline versus lower-chloride solutions for kidney transplantation. Cochrane Database Syst Rev. 2016:CD010741.
43. Shaw AD, Bagshaw SM, Goldstein SL. Major complications, mortality, and resource utilization after open abdominal surgery: 0.9% saline compared to Plasma-Lyte. Ann Surg. 2012;255:821–829.
44. Lambert PC, Sutton AJ, Abrams KR, Jones DR. A comparison of summary patient-level covariates in meta-regression with individual patient data meta-analysis. J Clin Epidemiol. 2002;55:86–94.
45. Cavalcanti A, Zampieri F, Azevedo L, Falavigna M, Brandao N, Berwanger O. Balanced Solution Versus Saline in Intensive Care Study (BaSICS trial). ClinicalTrials.gov identifier: NCT02875873.
46. Finfer S. Plasma-Lyte 148® versUs Saline Study (PLUS trial). ClinicalTrials.govidentifier: NCT02721654.
47. Rice TW, Semler MW. Isotonic Solutions and Major Adverse Renal Events Trial in the Medical Intensive Care Unit (SMART-MED trial). ClinicalTrials.govidentifier: NCT024449882015.
48. Rice TW, Semler MW. Isotonic Solutions and Major Adverse Renal Events Trial in the Non-Medical Intensive Care Unit (SMART-SURG trial). ClinicalTrials.govidentifier: NCT025477792015.
© 2018 International Anesthesia Research Society