Meta-Analysis of Sodium Bicarbonate Therapy for Prevention of Cardiac Surgery-Associated Acute Kidney Injury Jun Hyun Kim, MD,* Hyun Jung Kim, MPH, PhD,† Ji Yeon Kim, MD, PhD,* Hyeong sik Ahn, MD, PhD,† Il Min Ahn,†‡ Won Joo Choe, MD, PhD,* and Choon-Hak Lim, MD, PhD§ Objective: The aim of this study was to determine whether or not perioperative administration of sodium bicarbonate had a preventive effect on cardiac surgery-associated acute kidney injury (CSA-AKI) as shown in randomized controlled trials. Design: The authors conducted a systematic review and meta-analysis using MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), and KoreaMed. Setting: The authors searched MEDLINE, EMBASE, CENTRAL, and KoreaMed without language and date restrictions. They used both MeSH and free-text terms to identify relevant studies. Electronic searches were undertaken on July 31, 2014. Participants: Five randomized controlled studies included in this review. Measurements and Main Results: There were no differences in the development of CSA-AKI among patients in the sodium bicarbonate group compared with those in the control group (5 trials, 1,092 patients; n ¼ 233 of 547 in sodium bicarbonate (SB) group versus 225 of 545 in control
group (SC); risk ratio (RR), 0.95; 95% confidence interval (CI), 0.74-1.22. Also, there were no statistical differences in inhospital mortality (3 trials, 573 patients; n ¼ 21 of 288 in SB versus 14 of 285 in SC; RR, 1.44; 95% CI, 0.76-2.72), need for renal replacement therapy (4 trials, 1,000 patients; n ¼ 21 of 503 in SB versus 23 of 497 in SC; RR, 0.90; 95% CI, 0.50-1.60), length of stay in the intensive care unit (ICU) (hours) (4 trials, n ¼ 969 patients, weighted men difference (WMD), 2.17; 95% CI, -1.15-5.49), and length of ventilation (hours) (4 trials, 969 patients; WMD, 0.34; 95% CI, 0.80-1.48). Conclusions: Perioperative administration of sodium bicarbonate did not reduce the rate of CSA-AKI in randomized controlled trials. Therefore, use of perioperative administration of sodium bicarbonate for the prevention of CSAAKI is questionable. & 2015 Elsevier Inc. All rights reserved.
C
as increased vascular resistance, altered coagulation activity, platelet dysfunction, and renal tubular injury.1 The exact mechanism of how sodium bicarbonate might be helpful to reduce AKI after CPB has not been revealed clearly, but there was a report that sodium bicarbonate reduces acidification of the urine and the renal medulla and, thus, decreases free radical production to protect the kidneys from injury.5 Therefore, in this meta-analysis, the authors’ primary objective was to evaluate the effect of sodium bicarbonate in the prevention of CSA-AKI. Authors decided to include only randomized controlled trials (RCTs).
ARDIAC SURGERY-ASSOCIATED acute kidney injury (CSA-AKI) is a common and serious postoperative complication following exposure to cardiopulmonary bypass (CPB). It has been reported to occur in 8.9% to 39% of cardiac surgery patients who undergo surgery with CPB.1 This is related to increased mortality, morbidity, and hospital cost.1 In addition, CSA-AKI is the second most common cause of acute kidney injury (AKI) in the intensive care unit (ICU).2 The pathogenetic features of CSA-AKI are complex and multifactorial. There are several intraoperative elements of CSA-AKI following CPB, including hypoperfusion, ischemiareperfusion injury, decreased oxygen supply by hemodilution, activation of systemic inflammatory pathways, exogenous catecholamines, release of reactive oxygen species, and release of free hemoglobin and iron. These can reduce arterial oxygen content or impair renal oxygen delivery and contribute to AKI.3 Small microemboli formed during CPB can damage renal capillaries directly.4 Also, hemolysis and release of free hemoglobin may result in a variety of serious sequelae such
KEY WORDS: acute kidney injury, cardiac surgery, sodium bicarbonate
MATERIALS AND METHODS
This study was based on the Cochrane Review Methods. The authors report their findings according to the PRISMA (Preferred Reporting Items for Systematic Reviews and MetaAnalyses) statement. Data and Literature Sources
From the Departments of *Anesthesiology and Pain Medicine, Ilsan Paik Hospital, Inje University, Goyang, Korea, †Preventive Medicine, College of Medicine, Korea University, Seoul, Korea, ‡Literary Arts, Brown University, Providence, RI; and §Anesthesiology and Pain Medicine, Korea University Medical Center, Seoul, Korea. Jun Hyun Kim and Hyun Jung Kim contributed equally as first authors. Address reprint requests to Ji Yeon Kim, Department of Anesthesiology and Pain Medicine, Ilsan Paik Hospital, Inje University, 170 JuHwaRo, IlsanSeoGu, Goyang, GyungGiDo, South Korea. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.03.007 1248
The authors searched MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), and KoreaMed without language and date restrictions. They used both MeSH and free-text terms to identify relevant studies. Electronic searches were undertaken on July 31, 2014. The following keywords were searched: “Sodium Bicarbonate,” “Cardiac Surgical Procedures,” or “Cardiopulmonary Bypass and Sodium Chloride”. Search strategies were adapted for other databases based on the MEDLINE strategy (the authors present search terms in the Appendix). After the initial electronic search, they manually searched reference lists of identified studies to look for relevant studies their electronic searches might have missed. They also searched ClinicalTrials. gov to identify ongoing studies.
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 5 (October), 2015: pp 1248–1256
1249
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
Study Selection Two reviewers (J.Y.K. and J.H.K.) independently screened titles and abstracts for eligible studies. If eligibility could not be determined, the authors screened the full text. Inclusion criteria for studies were as follows: Studies were randomized controlled trials for prevention of CSA-AKI; conducted during cardiac surgery the intervention was perioperative infusion with sodium bicarbonate versus sodium chloride. Final inclusion of studies was based on the agreement of both reviewers. Data Collection
Idenficaon
From eligible randomized controlled trials, 2 reviewers independently collected information on study characteristics using a predesigned data extraction form. Any discrepancy unresolved by discussion was put under the review of a third author. The following variables were extracted from studies: AKI, mortality, need for renal replacement therapy, length of ventilation, length of stay in the ICU, and length of stay in the hospital. Dichotomous data (AKI, mortality, need for renal replacement therapy) were analyzed using the risk ratio (RR) measure and its 95% confidence interval (CI), and continuous variables (length of ventilation, length of stay in the ICU,
Records idenfied through database searching ( n = 1047 )
length of stay in the hospital) were analyzed using the weighted mean difference and its 95% CI. Assessment of Methodologic Quality The authors used the Cochrane Collaboration’s tool (sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, other sources of bias) to assess the risk of bias of randomized controlled trials. Two blinded reviewers independently assessed the risk of bias for each study. Disagreements were resolved by discussion between the reviewers. The authors did not assess publication bias because of the low statistical power when the number of included studies was fewer than 10. Statistical Analysis The primary outcome of the authors’ review was the difference in the presence of CSA-AKI between patients treated with sodium bicarbonate and those treated with sodium chloride. Secondary outcomes were mortality, the length of stay in the ICU, length of stay in the hospital, length of time on artificial ventilation, and whether or not postoperative renal replacement therapy was required. The presence of AKI, mortality, and postoperative renal replacement therapy were
Addional records idenfied through other sources (n=0)
Eligibility
Screening
Records aer duplicates removed ( n = 935 )
Records screened ( n = 13 )
Records excluded ( n = 922 ) Not relevant
Full-text arcles assessed for eligibility (n=6)
Full-text arcles excluded, with reasons (n=7) Not RCTs, no numeric data, could not be retrieved
Included
Studies included in qualitave synthesis (n=5)
Studies included in quantave synthesis (meta-analysis) (n=5) Fig 1.
Study flowchart.
64
Abbreviations: SB, sodium bicarbonate; SC, sodium chloride; AKI, acute kidney injury; Cr, creatinine; NGAL, neutrophil gelatinase-associated lipocalin; Pl, plasma; Ur, urinary; ICU, intensive care unit; RRT, renal replacement therapy.
22 3
24 5.1
Turner KR et al9 2014
59 Z0.3 mg/dL↑ or 1.5 fold↑ Cr in 48 h or oliguria of o0.5 mL/kg/h for 46 h
7 1.35
48 Z0.3 mg/dL↑ or 1.5 fold↑ Cr in 48 h or oliguria of o0.5 mL/kg/h for 46 h McGuinness SP et al8 2013 215 212 Z25%↑ or 0.5 mg/dL↑ Cr in 5 days 44
Kristeller JL et al10 2013
24 5.1 174 176 Z25%↑ or 0.5 mg/dL↑ Cr in 5 days Haase M et al7 2013
Urine pH
Unknown 24 4 50 Haase M et al6 2009
50 Z25%↑ Cr in 5 days
(mmol/Kg) SC
size
SB Study (reference no.)
Sample
Definition of AKI
Total SB Dose
Infusion Time (hours)
(SB vs SC)
Table 1. Characteristics of Individual Studies Included in Meta-analysis
Outcomes
KIM ET AL AKI, Pl Cr & Urea, Ur NGAL, NGAL/Cr, length of ventilation, length of ICU stay, hospital length stay, mortality, need for RRT 6.5 v 6.0 (6h) AKI, Pl Cr & Urea, Ur NGAL, length of ventilation, length of 7.5 v 5.5 (24h) ICU stay, hospital length stay, mortality 5.75 v 5.63 (3-4h) AKI, length of ventilation, length of ICU stay, hospital length stay, mortality Unknown AKI, length of ventilation, length of ICU stay, hospital length stay, mortality, need for RRT Unknown AKI, hospital length stay, mortality, need for RRT
1250
measured in pooled risk ratios via the Mantel-Haenszel method, using the number of events and total from both groups (sodium bicarbonate and sodium chloride groups). Length of stay in the ICU, length of stay in the hospital, and length of time on artificial ventilation were measured in weighted mean difference via the inverse variance method, using the mean and standard deviation from each study. For studies that presented median and interquartile ranges (IQRs), the authors followed the guidelines from the Cochrane Handbook to translate the values to mean and standard deviations. Heterogeneity was measured using the I2 statistic, the ratio of variation arising not from chance but from true difference among studies. I2 values above 50% were considered moderately heterogenous and were analyzed via the random-effect model. In this meta-analysis, the authors used The Cochrane Collaboration’s Review Manager (RevMan) computer program, Version 5.2 (Copenhagen, The Nordic Cochrane Center, 2012). RESULTS
Identification of Studies Searches of the databases resulted in 1,047 articles (154 from MEDLINE, 731 from EMBASE, 23 from Cochrane, 139 from KoreaMed) (Fig 1). After removal of 112 duplicated studies, the authors screened 935 studies and excluded 922 publications as they were clearly not relevant according to the titles and abstracts. The remaining 13 studies were analyzed.
Fig 2.
The Cochrane traffic lights summary of bias assessment.
1251
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
Table 2. Sensitivity Analysis Outcome
Studies
Hospital mortality Author-defined AKI Required renal replace therapy Hospital length stay
All (3) not including All (5) not including All (4) not including All (5) not including
Turner KR 2014 Turner KR 2014 Turner KR 2014 Turner KR 2014
Patients
573 450 1092 969 1000 877 1092 969
Effect Estimate (95% CI)
1.44 3.03 0.95 0.97 0.90 0.94 0.60 0.62
(0.76, 2.72) (0.99, 9.25) (0.74, 1.22) (0.72, 1.29) (0.50, 1.60) (0.49, 1.80) (-0.22, 1.42) (-0.21, 1.45)
I2
49% 0% 58% 66% 0% 0% 0% 0%
Abbreviation: AKI, acute kidney injury.
After review of the full-text articles, 7 publications were excluded from the meta-analysis because they were not randomized trials, reported no numeric data, or could not be retrieved. Therefore, the total number of studies included in the review was 5. Study Characteristics and Patient Populations Patient Populations Four studies6–9 included patients who fulfilled the following characteristics, which are predisposing factors for AKI: age above
Fig 3.
70 years, preexisting renal impairment, New York Heart Association class III/IV or left ventricular ejection fraction o35%, valvular surgery or concomitant valvular and coronary artery bypass graft surgery, redo cardiac surgery and insulin-dependent type 2 diabetes mellitus. Kristeller et al10 included patients with baseline estimated GFR of less than 60 mL/minutes/1.73 m2 using the Modification of Diet in Renal Disease equation. Some authors7,9 also presented results in per-protocol analysis, but the authors analyzed and presented all data on an intention-to-treat basis.
Cardiac surgery-associated acute kidney injury. AKI, acute kidney injury; Scr, serum creatinine.
1252
KIM ET AL
Fig 4.
Study Characteristics Treatment regimen varied among studies. Haase et al6 used an infusion of sodium bicarbonate at a dose of 0.5 mmol/kg body weight (¼bolus) diluted in 250 mL of 5% dextrose and H2O over 1 hour immediately after induction of anesthesia and before the first surgical incision, followed by continuous intravenous infusion of 0.15 mmol/kg/h (¼maintenance) diluted in 1,000 mL 5% dextrose and H2O over 23 hours (total dose of 4 mmol/kg over 24 hours) in 2009. In 2013, Haase et al7 used 0.5 mmol/kg of body weight in 250 mL given over 1 hour, commencing with induction of anesthesia. During the following 23 hours, a continuous intravenous infusion of sodium bicarbonate or sodium chloride at a dose of 0.2 mmol/ kg/h in 1,000 mL was administered. The bolus and continuous infusion achieved a total volume of 1.25 L and a total dose of 5.1 mmol/kg over 24 hours. Kristeller et al10 used 0.9% sodium chloride or sodium bicarbonate, 150 mEq mixed in 850 mL
Fig 5.
Mortality.
of a 5% dextrose solution. The infusion was started at 3 mL/kg/h starting 1 hour preoperatively and continuing until the patient was started on CPB, then reduced to 1 mL/kg/h during and for 6 hours after CPB. McGuinness et al8 used a dose of 0.5 mmol/kg body weight (¼bolus) over 1 hour immediately after the induction of anesthesia followed by continuous IV infusion of 0.2 mmol/kg/h over 23 hours (total dose 5.1 mmol/kg). Turner et al9 used an initial bolus consisting of 0.154 M NaCl or 0.150 M NaHCO3 at 5.0 mL/kg given over 15 minutes prior to initiation of CPB then infused with normal saline (NS) 0.154 M or 0.150 M NaHCO3, at a rate of 1.0 mL/kg/h for a total of 10 hours, and decreased to 0.4 mL/kg/h for the remaining 12 hours during and after CPB (Table 1). In 3 of the studies, AKI was defined as a creatinine level that was elevated more than 25% from the preop level in 5 days.6–8 The other 2 studies defined it as an increase of greater than one-and-a-half fold in the creatinine level compared to the preop
Need for renal replacement therapy.
1253
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
Fig 6.
Length of stay in the ICU (hours).
level or increase of serum creatinine by more than 0.3 mg/dL or documented reduction in urine output (o0.5 mL/kg/h for 46 h) (Acute Kidney Injury Network criteria stage 1).9,10 The authors present detailed characteristics and differences among studies in Table 1.
according to each definition (Fig 3). They also showed no difference between groups.
Secondary Endpoints Mortality
Quality of the Included Studies The risk of bias of the included studies was evaluated as low with the exception of the study by Turner et al.9 The authors present the Cochrane traffic lights summary of bias assessment as Figure 2. They performed sensitivity analyses of 4 outcomes to evaluate the effect of the study with high or unclear risk of bias on this meta-analysis (Table 2). Turner et al’s study result did not seem to skew overall results of metaanalysis.
There was no statistical difference in hospital mortality between the sodium bicarbonate group and the control group (3 trials, 573 patients; n ¼ 21 of 288 in the sodium bicarbonate group versus 14 of 285 in the control group; RR, 1.44; 95% CI, 0.76-2.72). There was moderate statistical heterogeneity (χ2 ¼ 3.94; I2 ¼ 49%; p ¼ 0.14) (Fig 4). Also, 30-day mortality and 90-day mortality did not show statistical differences (Fig 4).
Need for Renal Replacement Therapy Primary Outcome Cardiac Surgery-Associated Acute Kidney Injury Author-defined CSA-AKI occurred in a total of 233 patients among the 547 patients in the group treated with sodium bicarbonate compared with a total of 225 events in the 545 patients treated with sodium chloride. There was no difference in the development of CSA-AKI among patients in the sodium bicarbonate group compared with those in the control group (5 trials, 1,092 patients; n ¼ 233 of 547 in the sodium bicarbonate group versus 225 of 545 in the control group; RR, 0.95; 95% CI, 0.74-1.22). There was moderate statistical heterogeneity (χ2 ¼ 9.84 I2 ¼ 59%; p ¼ 0.04) (Fig 3). In particular, the studies of Haase et al6,7 mainly showed heterogeneity. These 5 studies used different definitions of AKI (Table 1), so the authors did subgroup analyses
Fig 7.
There was no statistical difference in the need for renal replacement therapy when comparing the sodium bicarbonate group and the control group (4 trials, 1,000 patients; n ¼ 21 of 503 in the sodium bicarbonate group versus 23 of 497 in the control group; RR, 0.90; 95% CI, 0.50-1.60). There was no statistical heterogeneity (χ2 ¼ 0.99; I2 ¼ 0%; p ¼ 0.80) (Fig 5).
Length of Stay in the ICU (Hours) There was no statistical difference in the ICU length of stay between the sodium bicarbonate group and the control group (4 trials, n ¼ 969 patients, weighted mean difference [WMD], 2.17; 95% CI, 1.15 to 5.49). There was no statistical heterogeneity (χ2 ¼ 0.69; I2 ¼ 0%; p ¼ 0.88) (Fig 6).
Length of ventilation (hours).
1254
KIM ET AL
Length of Ventilation (Hours) There was no statistical difference in the length of mechanical ventilation when comparing the sodium bicarbonate group and the control group (4 trials, 969 patients; WMD, 0.34; 95% CI, 0.80 to 1.48). There was no statistical heterogeneity (χ2 ¼ 0.80; I2 ¼ 0%; p ¼ 0.85) (Fig 7). DISCUSSION
In this meta-analysis of 5 randomized controlled trials including 1,092 patients, sodium bicarbonate infusion did not reduce CSA-AKI significantly. Excluding the first pilot study6 and second multicenter study,7 the other studies showed similar results, and the meta-analysis also showed the same results. There were no significant differences in mortality, need for renal replacement therapy, ICU length of stay, and length of ventilation, compared to control groups. In the authors’ analysis, control event rate was 0.413 (225/ 545), experimental event rate was 0.426 (233/547), and risk ratio was 1.03. Assuming alpha of 0.05 and beta of 0.10, the estimated population size to tell the difference of the effect is 60,581; assuming alpha of 0.05 and beta of 0.20, it will be 45,188.11 There was one meta-analysis article12 published after the authors’ search period, which included the same randomized controlled trials as theirs and one prospective cohort study.13 Although including the cohort study increased the number of patients, this cohort study included a relatively large number of patients compared to other RCTs, so it can increase heterogeneity. Furthermore, Cochrane organization does not recommend combining RCTs and non-randomized trials.14 The authors’ outcomes may have been skewed by early death. For example, a patient who died shortly after surgery might show a shorter duration of ICU stay and a shorter length of ventilation. Recently, in an individual patient meta-analysis,15 authors combined individual patient data of 3 RCTs,6–8 which were included in our meta-analysis. The study provided 7-day mortality of each group (sodium bicarbonate group 1.4% [6/429] v sodium chloride group 0.2% [1/438], p = 0.06), which was not significantly different.15 Nonetheless, relatively higher early mortality in the sodium bicarbonate group could skew its outcomes in favorable directions, which must be considered. Even though it was not included in this meta-analysis, the prospective observational cohort study conducted by Heringlake et al,13 which compared 280 patients who received an infusion of 4 mmol/kg of sodium bicarbonate perioperatively for 24 hours and 304 patients who received an infusion of sodium chloride, also did not show a significant reduction in CSA-AKI. The pilot study of Hasse et al6 reported that perioperative sodium bicarbonate infusion reduced CSA-AKI risk by 20%. It was the only RCT that showed reduction of CSA-AKI after sodium bicarbonate infusion. Maybe it was caused by the short mean CPB time in the experimental group (131.1 minutes in the sodium bicarbonate group, 142.6 minutes in the sodium chloride group), although it was not statistically different (p ¼ 0.19). Also, the study did not describe any adverse effects
related to sodium bicarbonate infusion or amounts of vasopressors and inotropes. The prospective study of Heringlake et al,13 which used the same dose of sodium bicarbonate, reported decreased arterial pressure as well as requirements for fluids and vasopressors in patients who received a bolus of sodium bicarbonate. Sodium bicarbonate has been thought to have renalprotective effects in AKI after CPB, but its rationale was not clearly proven. In animal models of hemoglobinuria or myoglobinuria, high tubular pH was beneficial in preventing renal injury.16 Urine alkalinization also showed a beneficial effect in rhabdomyolysis.17 Based on these findings, many institutions conventionally have administered sodium bicarbonate on a perioperative basis during cardiac surgery.13 Sodium bicarbonate in contrast-induced nephropathy (CIN) seems to have a beneficial effect by reducing the acidification of the urine and the medulla, reducing free radical production. Through that function, it protects the kidneys from renal ischemia and oxidative injury.18,19 In cardiac surgery, however, the causes of kidney injuries are much more complex and multifactorial, so the effect of sodium bicarbonate cannot be the same as in CIN. Various factors can affect free hemoglobin release, which constitutes the possible cause of CPB-related AKI. Differences in patient characteristics, CPB time, the integrity of the RBCs, the structure of the cardiotomy suction, and the negative pressure of the suction can result in various amounts of free hemoglobin release.20 Sodium bicarbonate may be beneficial to the patients who have higher free hemoglobin levels and may have a harmful or, at best, equivocal effect in those who have lower free hemoglobin levels. Therefore, if the authors could evaluate the effect of sodium bicarbonate only in a cohort of patients with high free hemoglobin, they might obtain a different result. But among the studies included in this metaanalysis, only 2 studies described CPB time,7,8 and even these 2 showed similar CPB time. So the authors were not able to perform subgroup analysis. Further studies relating free hemoglobin level and the effect of sodium bicarbonate would be interesting. All studies except Kristeller et al10 used a longer infusion period than CIN studies19 (almost 24 hours vs 6 hours) (Table 1), resulting in an increased total sodium bicarbonate dose. As a matter of fact, a shorter infusion period or smaller dose of sodium bicarbonate did not effectively elevate urine pH in patients who underwent cardiac surgery (Table 1). However, maybe these large doses of sodium bicarbonate compared to CIN studies caused different results. This meta-analysis included relatively few studies, and there were no statistically significant results. According to the authors’ analysis, however, the practice of perioperative sodium bicarbonate administration to reduce the incidence of AKI in patients who undergo cardiac surgery using CPB may not be recommended.
ACKNOWLEDGMENTS
This work was supported by the 2013 Inje University research grant.
1255
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
EMBASE
1. ‘bicarbonate’/exp 40136 2. “Sodium bicarbonate”:ab,ti OR 'Carbonic Acid Monosodium Salt':ab,ti 6151 3. 1 OR 2 41760 4. (Heart:ab,ti OR Cardiopulmonary:ab,ti OR Coronary:ab,ti OR Myocardial:ab,ti OR cardiac:ab,ti) AND (Revascularization:ab,ti OR Bypass:ab,ti OR Repair:ab,ti OR surgeries: ab,ti OR surgery:ab,ti) 213039 5. ‘heart surgery’/exp OR ‘cardiopulmonary bypass’/exp 276210 6. 4 OR 5 366706 7. 6 AND 3 863 8. 7 NOT (‘letter’/it OR ‘review’/it) 731 COCHRANE
1. MeSH descriptor: [Sodium Bicarbonate] explode all trees 495 2. “Sodium bicarbonate” or “Carbonic Acid Monosodium Salt”:ti,ab,kw (Word variations have been searched) 824 3. 1 or 2 824 4. Heart or Cardiopulmonary or Coronary or Myocardial or cardiac:ti,ab,kw (Word variations have been searched) 92664 5. Revascularization or Bypass or Repair or surgeries or
6. 7. 8. 9. 10. 11.
surgery:ti,ab,kw (Word variations have been searched) 78410 4 and 5 19978 MeSH descriptor: [Cardiac Surgical Procedures] explode all trees 11662 MeSH descriptor: [Cardiopulmonary Bypass] explode all trees 2406 7 or 8 12681 6 or 9 23267 3 and 10 23 KOREAMED
1. “Sodium bicarbonate”[ALL] OR “Carbonic Acid Monosodium Salt”[ALL] 112 2. (Heart[ALL] OR Cardiopulmonary[ALL] OR Coronary [ALL] OR Myocardial[ALL] OR cardiac[ALL]) AND (Revascularization[ALL] OR Bypass[ALL] OR Repair [ALL] OR surgery[ALL]) 3. 1 AND 2 139 MEDLINE 154 EMBASE 731 COCHRANE 23 KOREAMED 139 Total 1047 Duplicated 112 Total – Duplicated ¼ 935
REFERENCES 1. Mao H, Katz N, Ariyanon W, et al: Cardiac surgery-associated acute kidney injury. Cardiorenal Med 3:178-199, 2013 2. Uchino S, Kellum JA, Bellomo R, et al: Acute renal failure in critically ill patients: A multinational, multicenter study. JAMA 294: 813-818, 2005 3. Sear JW: Kidney dysfunction in the postoperative period. Br J Anaesth 95:20-32, 2005 4. Schiffl H: Sodium bicarbonate infusion for prevention of acute kidney injury: No evidence for superior benefit, but risk for harm? Int Urol Nephrol 47:321-326, 2014 5. Lindinger MI, Franklin TW, Lands LC, et al: NaHCO(3) and KHCO(3) ingestion rapidly increases renal electrolyte excretion in humans. J Appl Physiol (1985) 88:540-550, 2000 6. Haase M, Haase-Fielitz A, Bellomo R, et al: Sodium bicarbonate to prevent increases in serum creatinine after cardiac surgery: A pilot double-blind, randomized controlled trial. Crit Care Med 37:39-47, 2009 7. Haase M, Haase-Fielitz A, Plass M, et al: Prophylactic perioperative sodium bicarbonate to prevent acute kidney injury following open heart surgery: A multicenter double-blinded randomized controlled trial. PLoS Med. 10:e1001426, 2013 8. McGuinness SP, Parke RL, Bellomo R, et al: Sodium bicarbonate infusion to reduce cardiac surgery-associated acute kidney injury: A phase II multicenter double-blind randomized controlled trial. Crit Care Med 41:1599-1607, 2013 9. Turner KR, Fisher EC, Hade EM, et al: The role of perioperative sodium bicarbonate infusion affecting renal function after cardiothoracic surgery. Front Pharmacol 5:127, 2014 10. Kristeller JL, Zavorsky GS, Prior JE, et al: Lack of effectiveness of sodium bicarbonate in preventing kidney injury in patients
undergoing cardiac surgery: A randomized controlled trial. Pharmacotherapy 33:710-717, 2013 11. Pogue JM, Yusuf S: Cumulating evidence from randomized trials: Utilizing sequential monitoring boundaries for cumulative metaanalysis. Control Clin Trials 18:580-593, 1997 12. Tian M-L, Hu Y, Yuan J, et al: Efficacy and safety of perioperative sodium bicarbonate therapy for cardiac surgeryassociated acute kidney injury: a meta-analysis. J Cardiovasc Pharmacol 65:130-136, 2014 13. Heringlake M, Heinze H, Schubert M, et al: A perioperative infusion of sodium bicarbonate does not improve renal function in cardiac surgery patients: A prospective observational cohort study. Crit Care Lond Engl 16:R156, 2012 14. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org. 15. Bailey M, McGuinness S, Haase M, et al: Sodium bicarbonate and renal function after cardiac surgery: A prospectively planned individual patient meta-analysis. Anesthesiology 122: 294-306, 2015 16. Atkins JL: Effect of sodium bicarbonate preloading on ischemic renal failure. Nephron 44:70-74, 1986 17. Bosch X, Poch E, Grau JM: Rhabdomyolysis and acute kidney injury. N Engl J Med 361:62-72, 2009 18. Navaneethan SD, Singh S, Appasamy S, et al: Sodium bicarbonate therapy for prevention of contrast-induced nephropathy: A systematic review and meta-analysis. Am J Kidney Dis Off J Natl Kidney Found 53:617-627, 2009
1256
19. Jang J-S, Jin H-Y, Seo J-S, et al: Sodium bicarbonate therapy for the prevention of contrast-induced acute kidney injury – a systematic review and meta-analysis. Circ J Off J Jpn Circ Soc 76: 2255-2265, 2012
KIM ET AL
20. Hansbro SD, Sharpe DA, Catchpole R, et al: Haemolysis during cardiopulmonary bypass: An in vivo comparison of standard roller pumps, nonocclusive roller pumps and centrifugal pumps. Perfusion 14:3-10, 1999
group (SC); risk ratio (RR), 0.95; 95% confidence interval (CI), 0.74-1.22. Also, there were no statistical differences in inhospital mortality (3 trials, 573 patients; n ¼ 21 of 288 in SB versus 14 of 285 in SC; RR, 1.44; 95% CI, 0.76-2.72), need for renal replacement therapy (4 trials, 1,000 patients; n ¼ 21 of 503 in SB versus 23 of 497 in SC; RR, 0.90; 95% CI, 0.50-1.60), length of stay in the intensive care unit (ICU) (hours) (4 trials, n ¼ 969 patients, weighted men difference (WMD), 2.17; 95% CI, -1.15-5.49), and length of ventilation (hours) (4 trials, 969 patients; WMD, 0.34; 95% CI, 0.80-1.48). Conclusions: Perioperative administration of sodium bicarbonate did not reduce the rate of CSA-AKI in randomized controlled trials. Therefore, use of perioperative administration of sodium bicarbonate for the prevention of CSAAKI is questionable. & 2015 Elsevier Inc. All rights reserved.
C
as increased vascular resistance, altered coagulation activity, platelet dysfunction, and renal tubular injury.1 The exact mechanism of how sodium bicarbonate might be helpful to reduce AKI after CPB has not been revealed clearly, but there was a report that sodium bicarbonate reduces acidification of the urine and the renal medulla and, thus, decreases free radical production to protect the kidneys from injury.5 Therefore, in this meta-analysis, the authors’ primary objective was to evaluate the effect of sodium bicarbonate in the prevention of CSA-AKI. Authors decided to include only randomized controlled trials (RCTs).
ARDIAC SURGERY-ASSOCIATED acute kidney injury (CSA-AKI) is a common and serious postoperative complication following exposure to cardiopulmonary bypass (CPB). It has been reported to occur in 8.9% to 39% of cardiac surgery patients who undergo surgery with CPB.1 This is related to increased mortality, morbidity, and hospital cost.1 In addition, CSA-AKI is the second most common cause of acute kidney injury (AKI) in the intensive care unit (ICU).2 The pathogenetic features of CSA-AKI are complex and multifactorial. There are several intraoperative elements of CSA-AKI following CPB, including hypoperfusion, ischemiareperfusion injury, decreased oxygen supply by hemodilution, activation of systemic inflammatory pathways, exogenous catecholamines, release of reactive oxygen species, and release of free hemoglobin and iron. These can reduce arterial oxygen content or impair renal oxygen delivery and contribute to AKI.3 Small microemboli formed during CPB can damage renal capillaries directly.4 Also, hemolysis and release of free hemoglobin may result in a variety of serious sequelae such
KEY WORDS: acute kidney injury, cardiac surgery, sodium bicarbonate
MATERIALS AND METHODS
This study was based on the Cochrane Review Methods. The authors report their findings according to the PRISMA (Preferred Reporting Items for Systematic Reviews and MetaAnalyses) statement. Data and Literature Sources
From the Departments of *Anesthesiology and Pain Medicine, Ilsan Paik Hospital, Inje University, Goyang, Korea, †Preventive Medicine, College of Medicine, Korea University, Seoul, Korea, ‡Literary Arts, Brown University, Providence, RI; and §Anesthesiology and Pain Medicine, Korea University Medical Center, Seoul, Korea. Jun Hyun Kim and Hyun Jung Kim contributed equally as first authors. Address reprint requests to Ji Yeon Kim, Department of Anesthesiology and Pain Medicine, Ilsan Paik Hospital, Inje University, 170 JuHwaRo, IlsanSeoGu, Goyang, GyungGiDo, South Korea. E-mail: [email protected] © 2015 Elsevier Inc. All rights reserved. 1053-0770/2601-0001$36.00/0 http://dx.doi.org/10.1053/j.jvca.2015.03.007 1248
The authors searched MEDLINE, EMBASE, the Cochrane Central Register of Controlled Trials (CENTRAL), and KoreaMed without language and date restrictions. They used both MeSH and free-text terms to identify relevant studies. Electronic searches were undertaken on July 31, 2014. The following keywords were searched: “Sodium Bicarbonate,” “Cardiac Surgical Procedures,” or “Cardiopulmonary Bypass and Sodium Chloride”. Search strategies were adapted for other databases based on the MEDLINE strategy (the authors present search terms in the Appendix). After the initial electronic search, they manually searched reference lists of identified studies to look for relevant studies their electronic searches might have missed. They also searched ClinicalTrials. gov to identify ongoing studies.
Journal of Cardiothoracic and Vascular Anesthesia, Vol 29, No 5 (October), 2015: pp 1248–1256
1249
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
Study Selection Two reviewers (J.Y.K. and J.H.K.) independently screened titles and abstracts for eligible studies. If eligibility could not be determined, the authors screened the full text. Inclusion criteria for studies were as follows: Studies were randomized controlled trials for prevention of CSA-AKI; conducted during cardiac surgery the intervention was perioperative infusion with sodium bicarbonate versus sodium chloride. Final inclusion of studies was based on the agreement of both reviewers. Data Collection
Idenficaon
From eligible randomized controlled trials, 2 reviewers independently collected information on study characteristics using a predesigned data extraction form. Any discrepancy unresolved by discussion was put under the review of a third author. The following variables were extracted from studies: AKI, mortality, need for renal replacement therapy, length of ventilation, length of stay in the ICU, and length of stay in the hospital. Dichotomous data (AKI, mortality, need for renal replacement therapy) were analyzed using the risk ratio (RR) measure and its 95% confidence interval (CI), and continuous variables (length of ventilation, length of stay in the ICU,
Records idenfied through database searching ( n = 1047 )
length of stay in the hospital) were analyzed using the weighted mean difference and its 95% CI. Assessment of Methodologic Quality The authors used the Cochrane Collaboration’s tool (sequence generation, allocation concealment, blinding, incomplete outcome data, selective outcome reporting, other sources of bias) to assess the risk of bias of randomized controlled trials. Two blinded reviewers independently assessed the risk of bias for each study. Disagreements were resolved by discussion between the reviewers. The authors did not assess publication bias because of the low statistical power when the number of included studies was fewer than 10. Statistical Analysis The primary outcome of the authors’ review was the difference in the presence of CSA-AKI between patients treated with sodium bicarbonate and those treated with sodium chloride. Secondary outcomes were mortality, the length of stay in the ICU, length of stay in the hospital, length of time on artificial ventilation, and whether or not postoperative renal replacement therapy was required. The presence of AKI, mortality, and postoperative renal replacement therapy were
Addional records idenfied through other sources (n=0)
Eligibility
Screening
Records aer duplicates removed ( n = 935 )
Records screened ( n = 13 )
Records excluded ( n = 922 ) Not relevant
Full-text arcles assessed for eligibility (n=6)
Full-text arcles excluded, with reasons (n=7) Not RCTs, no numeric data, could not be retrieved
Included
Studies included in qualitave synthesis (n=5)
Studies included in quantave synthesis (meta-analysis) (n=5) Fig 1.
Study flowchart.
64
Abbreviations: SB, sodium bicarbonate; SC, sodium chloride; AKI, acute kidney injury; Cr, creatinine; NGAL, neutrophil gelatinase-associated lipocalin; Pl, plasma; Ur, urinary; ICU, intensive care unit; RRT, renal replacement therapy.
22 3
24 5.1
Turner KR et al9 2014
59 Z0.3 mg/dL↑ or 1.5 fold↑ Cr in 48 h or oliguria of o0.5 mL/kg/h for 46 h
7 1.35
48 Z0.3 mg/dL↑ or 1.5 fold↑ Cr in 48 h or oliguria of o0.5 mL/kg/h for 46 h McGuinness SP et al8 2013 215 212 Z25%↑ or 0.5 mg/dL↑ Cr in 5 days 44
Kristeller JL et al10 2013
24 5.1 174 176 Z25%↑ or 0.5 mg/dL↑ Cr in 5 days Haase M et al7 2013
Urine pH
Unknown 24 4 50 Haase M et al6 2009
50 Z25%↑ Cr in 5 days
(mmol/Kg) SC
size
SB Study (reference no.)
Sample
Definition of AKI
Total SB Dose
Infusion Time (hours)
(SB vs SC)
Table 1. Characteristics of Individual Studies Included in Meta-analysis
Outcomes
KIM ET AL AKI, Pl Cr & Urea, Ur NGAL, NGAL/Cr, length of ventilation, length of ICU stay, hospital length stay, mortality, need for RRT 6.5 v 6.0 (6h) AKI, Pl Cr & Urea, Ur NGAL, length of ventilation, length of 7.5 v 5.5 (24h) ICU stay, hospital length stay, mortality 5.75 v 5.63 (3-4h) AKI, length of ventilation, length of ICU stay, hospital length stay, mortality Unknown AKI, length of ventilation, length of ICU stay, hospital length stay, mortality, need for RRT Unknown AKI, hospital length stay, mortality, need for RRT
1250
measured in pooled risk ratios via the Mantel-Haenszel method, using the number of events and total from both groups (sodium bicarbonate and sodium chloride groups). Length of stay in the ICU, length of stay in the hospital, and length of time on artificial ventilation were measured in weighted mean difference via the inverse variance method, using the mean and standard deviation from each study. For studies that presented median and interquartile ranges (IQRs), the authors followed the guidelines from the Cochrane Handbook to translate the values to mean and standard deviations. Heterogeneity was measured using the I2 statistic, the ratio of variation arising not from chance but from true difference among studies. I2 values above 50% were considered moderately heterogenous and were analyzed via the random-effect model. In this meta-analysis, the authors used The Cochrane Collaboration’s Review Manager (RevMan) computer program, Version 5.2 (Copenhagen, The Nordic Cochrane Center, 2012). RESULTS
Identification of Studies Searches of the databases resulted in 1,047 articles (154 from MEDLINE, 731 from EMBASE, 23 from Cochrane, 139 from KoreaMed) (Fig 1). After removal of 112 duplicated studies, the authors screened 935 studies and excluded 922 publications as they were clearly not relevant according to the titles and abstracts. The remaining 13 studies were analyzed.
Fig 2.
The Cochrane traffic lights summary of bias assessment.
1251
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
Table 2. Sensitivity Analysis Outcome
Studies
Hospital mortality Author-defined AKI Required renal replace therapy Hospital length stay
All (3) not including All (5) not including All (4) not including All (5) not including
Turner KR 2014 Turner KR 2014 Turner KR 2014 Turner KR 2014
Patients
573 450 1092 969 1000 877 1092 969
Effect Estimate (95% CI)
1.44 3.03 0.95 0.97 0.90 0.94 0.60 0.62
(0.76, 2.72) (0.99, 9.25) (0.74, 1.22) (0.72, 1.29) (0.50, 1.60) (0.49, 1.80) (-0.22, 1.42) (-0.21, 1.45)
I2
49% 0% 58% 66% 0% 0% 0% 0%
Abbreviation: AKI, acute kidney injury.
After review of the full-text articles, 7 publications were excluded from the meta-analysis because they were not randomized trials, reported no numeric data, or could not be retrieved. Therefore, the total number of studies included in the review was 5. Study Characteristics and Patient Populations Patient Populations Four studies6–9 included patients who fulfilled the following characteristics, which are predisposing factors for AKI: age above
Fig 3.
70 years, preexisting renal impairment, New York Heart Association class III/IV or left ventricular ejection fraction o35%, valvular surgery or concomitant valvular and coronary artery bypass graft surgery, redo cardiac surgery and insulin-dependent type 2 diabetes mellitus. Kristeller et al10 included patients with baseline estimated GFR of less than 60 mL/minutes/1.73 m2 using the Modification of Diet in Renal Disease equation. Some authors7,9 also presented results in per-protocol analysis, but the authors analyzed and presented all data on an intention-to-treat basis.
Cardiac surgery-associated acute kidney injury. AKI, acute kidney injury; Scr, serum creatinine.
1252
KIM ET AL
Fig 4.
Study Characteristics Treatment regimen varied among studies. Haase et al6 used an infusion of sodium bicarbonate at a dose of 0.5 mmol/kg body weight (¼bolus) diluted in 250 mL of 5% dextrose and H2O over 1 hour immediately after induction of anesthesia and before the first surgical incision, followed by continuous intravenous infusion of 0.15 mmol/kg/h (¼maintenance) diluted in 1,000 mL 5% dextrose and H2O over 23 hours (total dose of 4 mmol/kg over 24 hours) in 2009. In 2013, Haase et al7 used 0.5 mmol/kg of body weight in 250 mL given over 1 hour, commencing with induction of anesthesia. During the following 23 hours, a continuous intravenous infusion of sodium bicarbonate or sodium chloride at a dose of 0.2 mmol/ kg/h in 1,000 mL was administered. The bolus and continuous infusion achieved a total volume of 1.25 L and a total dose of 5.1 mmol/kg over 24 hours. Kristeller et al10 used 0.9% sodium chloride or sodium bicarbonate, 150 mEq mixed in 850 mL
Fig 5.
Mortality.
of a 5% dextrose solution. The infusion was started at 3 mL/kg/h starting 1 hour preoperatively and continuing until the patient was started on CPB, then reduced to 1 mL/kg/h during and for 6 hours after CPB. McGuinness et al8 used a dose of 0.5 mmol/kg body weight (¼bolus) over 1 hour immediately after the induction of anesthesia followed by continuous IV infusion of 0.2 mmol/kg/h over 23 hours (total dose 5.1 mmol/kg). Turner et al9 used an initial bolus consisting of 0.154 M NaCl or 0.150 M NaHCO3 at 5.0 mL/kg given over 15 minutes prior to initiation of CPB then infused with normal saline (NS) 0.154 M or 0.150 M NaHCO3, at a rate of 1.0 mL/kg/h for a total of 10 hours, and decreased to 0.4 mL/kg/h for the remaining 12 hours during and after CPB (Table 1). In 3 of the studies, AKI was defined as a creatinine level that was elevated more than 25% from the preop level in 5 days.6–8 The other 2 studies defined it as an increase of greater than one-and-a-half fold in the creatinine level compared to the preop
Need for renal replacement therapy.
1253
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
Fig 6.
Length of stay in the ICU (hours).
level or increase of serum creatinine by more than 0.3 mg/dL or documented reduction in urine output (o0.5 mL/kg/h for 46 h) (Acute Kidney Injury Network criteria stage 1).9,10 The authors present detailed characteristics and differences among studies in Table 1.
according to each definition (Fig 3). They also showed no difference between groups.
Secondary Endpoints Mortality
Quality of the Included Studies The risk of bias of the included studies was evaluated as low with the exception of the study by Turner et al.9 The authors present the Cochrane traffic lights summary of bias assessment as Figure 2. They performed sensitivity analyses of 4 outcomes to evaluate the effect of the study with high or unclear risk of bias on this meta-analysis (Table 2). Turner et al’s study result did not seem to skew overall results of metaanalysis.
There was no statistical difference in hospital mortality between the sodium bicarbonate group and the control group (3 trials, 573 patients; n ¼ 21 of 288 in the sodium bicarbonate group versus 14 of 285 in the control group; RR, 1.44; 95% CI, 0.76-2.72). There was moderate statistical heterogeneity (χ2 ¼ 3.94; I2 ¼ 49%; p ¼ 0.14) (Fig 4). Also, 30-day mortality and 90-day mortality did not show statistical differences (Fig 4).
Need for Renal Replacement Therapy Primary Outcome Cardiac Surgery-Associated Acute Kidney Injury Author-defined CSA-AKI occurred in a total of 233 patients among the 547 patients in the group treated with sodium bicarbonate compared with a total of 225 events in the 545 patients treated with sodium chloride. There was no difference in the development of CSA-AKI among patients in the sodium bicarbonate group compared with those in the control group (5 trials, 1,092 patients; n ¼ 233 of 547 in the sodium bicarbonate group versus 225 of 545 in the control group; RR, 0.95; 95% CI, 0.74-1.22). There was moderate statistical heterogeneity (χ2 ¼ 9.84 I2 ¼ 59%; p ¼ 0.04) (Fig 3). In particular, the studies of Haase et al6,7 mainly showed heterogeneity. These 5 studies used different definitions of AKI (Table 1), so the authors did subgroup analyses
Fig 7.
There was no statistical difference in the need for renal replacement therapy when comparing the sodium bicarbonate group and the control group (4 trials, 1,000 patients; n ¼ 21 of 503 in the sodium bicarbonate group versus 23 of 497 in the control group; RR, 0.90; 95% CI, 0.50-1.60). There was no statistical heterogeneity (χ2 ¼ 0.99; I2 ¼ 0%; p ¼ 0.80) (Fig 5).
Length of Stay in the ICU (Hours) There was no statistical difference in the ICU length of stay between the sodium bicarbonate group and the control group (4 trials, n ¼ 969 patients, weighted mean difference [WMD], 2.17; 95% CI, 1.15 to 5.49). There was no statistical heterogeneity (χ2 ¼ 0.69; I2 ¼ 0%; p ¼ 0.88) (Fig 6).
Length of ventilation (hours).
1254
KIM ET AL
Length of Ventilation (Hours) There was no statistical difference in the length of mechanical ventilation when comparing the sodium bicarbonate group and the control group (4 trials, 969 patients; WMD, 0.34; 95% CI, 0.80 to 1.48). There was no statistical heterogeneity (χ2 ¼ 0.80; I2 ¼ 0%; p ¼ 0.85) (Fig 7). DISCUSSION
In this meta-analysis of 5 randomized controlled trials including 1,092 patients, sodium bicarbonate infusion did not reduce CSA-AKI significantly. Excluding the first pilot study6 and second multicenter study,7 the other studies showed similar results, and the meta-analysis also showed the same results. There were no significant differences in mortality, need for renal replacement therapy, ICU length of stay, and length of ventilation, compared to control groups. In the authors’ analysis, control event rate was 0.413 (225/ 545), experimental event rate was 0.426 (233/547), and risk ratio was 1.03. Assuming alpha of 0.05 and beta of 0.10, the estimated population size to tell the difference of the effect is 60,581; assuming alpha of 0.05 and beta of 0.20, it will be 45,188.11 There was one meta-analysis article12 published after the authors’ search period, which included the same randomized controlled trials as theirs and one prospective cohort study.13 Although including the cohort study increased the number of patients, this cohort study included a relatively large number of patients compared to other RCTs, so it can increase heterogeneity. Furthermore, Cochrane organization does not recommend combining RCTs and non-randomized trials.14 The authors’ outcomes may have been skewed by early death. For example, a patient who died shortly after surgery might show a shorter duration of ICU stay and a shorter length of ventilation. Recently, in an individual patient meta-analysis,15 authors combined individual patient data of 3 RCTs,6–8 which were included in our meta-analysis. The study provided 7-day mortality of each group (sodium bicarbonate group 1.4% [6/429] v sodium chloride group 0.2% [1/438], p = 0.06), which was not significantly different.15 Nonetheless, relatively higher early mortality in the sodium bicarbonate group could skew its outcomes in favorable directions, which must be considered. Even though it was not included in this meta-analysis, the prospective observational cohort study conducted by Heringlake et al,13 which compared 280 patients who received an infusion of 4 mmol/kg of sodium bicarbonate perioperatively for 24 hours and 304 patients who received an infusion of sodium chloride, also did not show a significant reduction in CSA-AKI. The pilot study of Hasse et al6 reported that perioperative sodium bicarbonate infusion reduced CSA-AKI risk by 20%. It was the only RCT that showed reduction of CSA-AKI after sodium bicarbonate infusion. Maybe it was caused by the short mean CPB time in the experimental group (131.1 minutes in the sodium bicarbonate group, 142.6 minutes in the sodium chloride group), although it was not statistically different (p ¼ 0.19). Also, the study did not describe any adverse effects
related to sodium bicarbonate infusion or amounts of vasopressors and inotropes. The prospective study of Heringlake et al,13 which used the same dose of sodium bicarbonate, reported decreased arterial pressure as well as requirements for fluids and vasopressors in patients who received a bolus of sodium bicarbonate. Sodium bicarbonate has been thought to have renalprotective effects in AKI after CPB, but its rationale was not clearly proven. In animal models of hemoglobinuria or myoglobinuria, high tubular pH was beneficial in preventing renal injury.16 Urine alkalinization also showed a beneficial effect in rhabdomyolysis.17 Based on these findings, many institutions conventionally have administered sodium bicarbonate on a perioperative basis during cardiac surgery.13 Sodium bicarbonate in contrast-induced nephropathy (CIN) seems to have a beneficial effect by reducing the acidification of the urine and the medulla, reducing free radical production. Through that function, it protects the kidneys from renal ischemia and oxidative injury.18,19 In cardiac surgery, however, the causes of kidney injuries are much more complex and multifactorial, so the effect of sodium bicarbonate cannot be the same as in CIN. Various factors can affect free hemoglobin release, which constitutes the possible cause of CPB-related AKI. Differences in patient characteristics, CPB time, the integrity of the RBCs, the structure of the cardiotomy suction, and the negative pressure of the suction can result in various amounts of free hemoglobin release.20 Sodium bicarbonate may be beneficial to the patients who have higher free hemoglobin levels and may have a harmful or, at best, equivocal effect in those who have lower free hemoglobin levels. Therefore, if the authors could evaluate the effect of sodium bicarbonate only in a cohort of patients with high free hemoglobin, they might obtain a different result. But among the studies included in this metaanalysis, only 2 studies described CPB time,7,8 and even these 2 showed similar CPB time. So the authors were not able to perform subgroup analysis. Further studies relating free hemoglobin level and the effect of sodium bicarbonate would be interesting. All studies except Kristeller et al10 used a longer infusion period than CIN studies19 (almost 24 hours vs 6 hours) (Table 1), resulting in an increased total sodium bicarbonate dose. As a matter of fact, a shorter infusion period or smaller dose of sodium bicarbonate did not effectively elevate urine pH in patients who underwent cardiac surgery (Table 1). However, maybe these large doses of sodium bicarbonate compared to CIN studies caused different results. This meta-analysis included relatively few studies, and there were no statistically significant results. According to the authors’ analysis, however, the practice of perioperative sodium bicarbonate administration to reduce the incidence of AKI in patients who undergo cardiac surgery using CPB may not be recommended.
ACKNOWLEDGMENTS
This work was supported by the 2013 Inje University research grant.
1255
META-ANALYSIS OF SB EFFECTS ON CSA-AKI
EMBASE
1. ‘bicarbonate’/exp 40136 2. “Sodium bicarbonate”:ab,ti OR 'Carbonic Acid Monosodium Salt':ab,ti 6151 3. 1 OR 2 41760 4. (Heart:ab,ti OR Cardiopulmonary:ab,ti OR Coronary:ab,ti OR Myocardial:ab,ti OR cardiac:ab,ti) AND (Revascularization:ab,ti OR Bypass:ab,ti OR Repair:ab,ti OR surgeries: ab,ti OR surgery:ab,ti) 213039 5. ‘heart surgery’/exp OR ‘cardiopulmonary bypass’/exp 276210 6. 4 OR 5 366706 7. 6 AND 3 863 8. 7 NOT (‘letter’/it OR ‘review’/it) 731 COCHRANE
1. MeSH descriptor: [Sodium Bicarbonate] explode all trees 495 2. “Sodium bicarbonate” or “Carbonic Acid Monosodium Salt”:ti,ab,kw (Word variations have been searched) 824 3. 1 or 2 824 4. Heart or Cardiopulmonary or Coronary or Myocardial or cardiac:ti,ab,kw (Word variations have been searched) 92664 5. Revascularization or Bypass or Repair or surgeries or
6. 7. 8. 9. 10. 11.
surgery:ti,ab,kw (Word variations have been searched) 78410 4 and 5 19978 MeSH descriptor: [Cardiac Surgical Procedures] explode all trees 11662 MeSH descriptor: [Cardiopulmonary Bypass] explode all trees 2406 7 or 8 12681 6 or 9 23267 3 and 10 23 KOREAMED
1. “Sodium bicarbonate”[ALL] OR “Carbonic Acid Monosodium Salt”[ALL] 112 2. (Heart[ALL] OR Cardiopulmonary[ALL] OR Coronary [ALL] OR Myocardial[ALL] OR cardiac[ALL]) AND (Revascularization[ALL] OR Bypass[ALL] OR Repair [ALL] OR surgery[ALL]) 3. 1 AND 2 139 MEDLINE 154 EMBASE 731 COCHRANE 23 KOREAMED 139 Total 1047 Duplicated 112 Total – Duplicated ¼ 935
REFERENCES 1. Mao H, Katz N, Ariyanon W, et al: Cardiac surgery-associated acute kidney injury. Cardiorenal Med 3:178-199, 2013 2. Uchino S, Kellum JA, Bellomo R, et al: Acute renal failure in critically ill patients: A multinational, multicenter study. JAMA 294: 813-818, 2005 3. Sear JW: Kidney dysfunction in the postoperative period. Br J Anaesth 95:20-32, 2005 4. Schiffl H: Sodium bicarbonate infusion for prevention of acute kidney injury: No evidence for superior benefit, but risk for harm? Int Urol Nephrol 47:321-326, 2014 5. Lindinger MI, Franklin TW, Lands LC, et al: NaHCO(3) and KHCO(3) ingestion rapidly increases renal electrolyte excretion in humans. J Appl Physiol (1985) 88:540-550, 2000 6. Haase M, Haase-Fielitz A, Bellomo R, et al: Sodium bicarbonate to prevent increases in serum creatinine after cardiac surgery: A pilot double-blind, randomized controlled trial. Crit Care Med 37:39-47, 2009 7. Haase M, Haase-Fielitz A, Plass M, et al: Prophylactic perioperative sodium bicarbonate to prevent acute kidney injury following open heart surgery: A multicenter double-blinded randomized controlled trial. PLoS Med. 10:e1001426, 2013 8. McGuinness SP, Parke RL, Bellomo R, et al: Sodium bicarbonate infusion to reduce cardiac surgery-associated acute kidney injury: A phase II multicenter double-blind randomized controlled trial. Crit Care Med 41:1599-1607, 2013 9. Turner KR, Fisher EC, Hade EM, et al: The role of perioperative sodium bicarbonate infusion affecting renal function after cardiothoracic surgery. Front Pharmacol 5:127, 2014 10. Kristeller JL, Zavorsky GS, Prior JE, et al: Lack of effectiveness of sodium bicarbonate in preventing kidney injury in patients
undergoing cardiac surgery: A randomized controlled trial. Pharmacotherapy 33:710-717, 2013 11. Pogue JM, Yusuf S: Cumulating evidence from randomized trials: Utilizing sequential monitoring boundaries for cumulative metaanalysis. Control Clin Trials 18:580-593, 1997 12. Tian M-L, Hu Y, Yuan J, et al: Efficacy and safety of perioperative sodium bicarbonate therapy for cardiac surgeryassociated acute kidney injury: a meta-analysis. J Cardiovasc Pharmacol 65:130-136, 2014 13. Heringlake M, Heinze H, Schubert M, et al: A perioperative infusion of sodium bicarbonate does not improve renal function in cardiac surgery patients: A prospective observational cohort study. Crit Care Lond Engl 16:R156, 2012 14. Higgins JPT, Green S (editors). Cochrane Handbook for Systematic Reviews of Interventions Version 5.1.0 [updated March 2011]. The Cochrane Collaboration, 2011. Available from www.cochrane-handbook.org. 15. Bailey M, McGuinness S, Haase M, et al: Sodium bicarbonate and renal function after cardiac surgery: A prospectively planned individual patient meta-analysis. Anesthesiology 122: 294-306, 2015 16. Atkins JL: Effect of sodium bicarbonate preloading on ischemic renal failure. Nephron 44:70-74, 1986 17. Bosch X, Poch E, Grau JM: Rhabdomyolysis and acute kidney injury. N Engl J Med 361:62-72, 2009 18. Navaneethan SD, Singh S, Appasamy S, et al: Sodium bicarbonate therapy for prevention of contrast-induced nephropathy: A systematic review and meta-analysis. Am J Kidney Dis Off J Natl Kidney Found 53:617-627, 2009
1256
19. Jang J-S, Jin H-Y, Seo J-S, et al: Sodium bicarbonate therapy for the prevention of contrast-induced acute kidney injury – a systematic review and meta-analysis. Circ J Off J Jpn Circ Soc 76: 2255-2265, 2012
KIM ET AL
20. Hansbro SD, Sharpe DA, Catchpole R, et al: Haemolysis during cardiopulmonary bypass: An in vivo comparison of standard roller pumps, nonocclusive roller pumps and centrifugal pumps. Perfusion 14:3-10, 1999
