Low oxygen delivery during cardiopulmonary bypass is related to a range of adverse outcomes. Previous research specified certain critical oxygen delivery levels associated with acute kidney injury. However, a single universal critical oxygen delivery value is not sensible, as oxygen consumption has to be considered when determining critical delivery values. This study examined the associations between oxygen delivery and oxygen consumption and between oxygen delivery and kidney function in patients undergoing cardiopulmonary bypass.
Oxygen delivery, oxygen consumption and kidney function decrease were retrospectively studied in 65 adult patients.
Mean oxygen consumption was 56 ± 8 ml/min/m2, mean oxygen delivery was 281 ± 39 ml/min/m2. Twenty-seven patients (42%) had an oxygen delivery lower than the previously mentioned critical value of 272 ml/min/m2. None of the patients developed acute kidney injury according to RIFLE criteria. However, in 10 patients (15%) a decrease in the estimated glomerular filtration rate of more than 10% was noted, which was not associated with oxygen delivery lower than 272 ml/min/m2. Eighteen patients had a strong correlation (r >0.500) between DO2 and VO2, but this was not related to low oxygen delivery. Central venous oxygen saturation (77 ± 3%), oxygen extraction ratio (21 ± 3%) and blood lactate levels at the end of surgery (1.2 ± 0.3 mmol/l) showed not to be indicative of insufficient oxygen delivery either.
This study could not confirm an evident correlation between O2 delivery and O2 consumption or kidney function decrease, even at values below previously specified critical levels. The variability in O2 consumption however, is an indication that every patient has individual O2 needs, advocating for an individualized O2 delivery goal.
Citation: Hendrix RHJ, Ganushchak YM, Weerwind PW (2019) Oxygen delivery, oxygen consumption and decreased kidney function after cardiopulmonary bypass. PLoS ONE 14(11): e0225541. https://doi.org/10.1371/journal.pone.0225541
Editor: Jaap A. Joles, University Medical Center Utrecht, NETHERLANDS
Received: August 22, 2019; Accepted: November 5, 2019; Published: November 22, 2019
Copyright: © 2019 Hendrix et al. This is an open access article distributed under the terms of the Creative Commons Attribution License, which permits unrestricted use, distribution, and reproduction in any medium, provided the original author and source are credited.
Data Availability: The data file is available from the DANS EASY database via the following link: https://doi.org/10.17026/dans-xw4-a6kc.
Funding: The authors received no specific funding for this work.
Competing interests: The authors have declared that no competing interests exist.
Oxygen delivery (DO2) during cardiopulmonary bypass (CPB) has received considerable attention in recent years, as it is one of the few modifiable factors related to acute kidney injury (AKI) and other morbidities. DO2 is dependent on hemoglobin concentration, oxygen saturation and blood flow. When either of these factors decreases the reduction is met by an increase in organ O2-extraction rate (OER) to meet metabolic needs, until the point of maximum O2-extraction. When DO2 falls beyond this point, referred to as the critical DO2 level (DO2crit), a pathologic supply dependency arises as organ oxygen consumption (VO2) decreases proportionally to decreasing DO2. At this point, cells enter the anaerobic metabolism phase and lactate levels rise [1–3].
Priming volume and crystalloid cardioplegic solution are the main contributors to a decreased hemoglobin concentration during CPB. The drop in oxygen carrying capacity due to hemodilution is often not considered when CPB flow rates are determined. The subsequent drop in DO2 does not have to be a problem however, as VO2 is also decreased by anaesthesia, possible hypothermia, and even by excluding and arresting the heart. This was rationalised by Ganushchak et al , who found no correlation between VO2 and DO2 during CPB. In contrast, other studies showed a correlation between perioperative DO2 and postoperative AKI, a severe complication with high morbidity and mortality, specifying DO2crit levels in the range of 200–300 ml/min/m2 [5–7].
Since DO2 below a certain critical level can undoubtedly lead to a range of morbidities including AKI, goal directed perfusion strategies have aimed at keeping DO2 above critical levels. Even though this seems to decrease the rate of AKI [8, 9], performing CPB purely based on the goal of keeping DO2 above a universal value might not be the optimal perfusion strategy for every patient, as one may argue that using a single DO2crit value for all patients is not sensible. DO2crit cannot be considered without a patient’s VO2, which in turn depends on patient characteristics and is additionally influenced by factors like anaesthesia technique and temperature management, leading to different VO2 levels and subsequently different DO2 requirements for every patient. We therefore explored the relationship of DO2, VO2 and AKI in a group of patients undergoing CPB for cardiac surgery at our institution, and determined if one of the previously found DO2crit levels can be applicable to all these patients as well.
A dataset of 67 adult patients who underwent elective CPB between May and November 2014 at the Maastricht University Medical Centre was retrospectively analysed in this study. Patients undergoing re-operations and procedures requiring deep hypothermic circulatory arrest were excluded from this study. This concerned 2 patients, resulting in data from 65 procedures. Data acquisition at the time and current data analyses were performed anonymously and included only routine measurements performed during CPB without the need for any intervention. In accordance with the Dutch law for approving medical research, Institutional Review Board approval was therefore waived.
Anesthesia and perfusion techniques
General anaesthesia was induced using weight-related infusion of sufentanil (1.0 μg/kg), midazolam (0.1 mg/kg) and rocuronium bromide (1.0 mg/kg). Subsequently, maintenance doses of 5.0 mg/kg/h of propofol and 1.0 μg/kg/h of sufentanil were used. A Stöckert S5 heart-lung machine with a CP5 pump unit (LivaNova Deutschland GmbH, Munich, Germany) was used in all cases, and comprised a phosphorylcholine-coated adult oxygenator tubing pack (LivaNova, London, England) and a Revolution centrifugal pump (LivaNova). Before initiation of CPB the patient received a bolus (300 IU/kg body weight) of heparin (Leo Pharma B.V., Breda, Netherlands). CPB was started when activated clotting time (ACT, measured using HemoTec ACT II, Medtronic, Minneapolis, Minnesota, USA) was extended to at least 400 seconds. After central cannulation, pulsatile CPB (pulse frequency 70 bpm, pulse width 50%, base flow 30%) was started with a pump flow indexed at 2.4 l/min/m2. Arterial and venous blood gases were measured using the CDI500 blood parameter monitoring system (Terumo Corporation, Tokyo, Japan), and arterial pO2 was kept between 10–15 kPa as measured by CDI500. Mean arterial blood pressure was kept between 70 and 80 mmHg using phenylephrine titration when applicable. Cardiac arrest was induced using either cold crystalloid cardioplegia (St. Thomas II solution, Apotheek Haagse Ziekenhuizen, Den Haag, the Netherlands) or warm blood cardioplegia (pharmacy Catharina hospital, Eindhoven, the Netherlands). After cessation of CPB, protamine in a proportion of 1.0–1.2 mg/100 IU of the initial heparin dose was used as an antidote. If after this dose the ACT was still prolonged compared to baseline, the patient received an additional dose of protamine accordingly.
Data collection and analysis
Continuous inline monitoring of arterial and venous blood gas parameters during cardiac arrest was done using the CDI500, which was calibrated according to the instructions for use. In addition, arterial and venous blood samples were sent to the laboratory when on full CPB support and before initiating weaning from CPB. Laboratory results were used to recalculate the CDI500 data for hemoglobin towards more accurate results by means of regression. These, and all other CPB data were digitally gathered in a patient data management system (PDMS, Chipsoft B.V., Amsterdam, the Netherlands). All data were automatically synchronized by the PDMS and combined into one output file at a rate of 6 data points per minute.
Where hb is hemoglobin concentration in g/dl; SaO2 and SvO2 are arterial and venous O2 saturations (decimal) and PaO2 and PvO2 are the partial pressures of O2 in arterial and venous blood respectively (in mmHg).
Where both VO2 and DO2 are in ml/dl/m2.
As most perfusion goals are aimed at keeping DO2 above 270–300 ml/min/m2 [8, 9] we chose to test if the DO2crit level found by Ranucci et al  (272 ml/min/m2) applies to our patient population. Therefore, the population was split in a group with a mean DO2 higher than 272 ml/min/m2 during the aortic cross clamp time (group DO2 >272) and a group with a mean DO2 lower than 272 ml/min/m2 (group DO2 <272). The two groups were then compared on the prevalence of postoperative kidney function decrease, which was assessed by calculating the estimated glomerular filtration rate (eGFR) using the simplified 'modification of diet in renal disease' formula ,
Where Scr is the serum creatinine level in mg/dl and A is the age in years. According to RIFLE criteria, patients in the lowest AKI class have a more than 25% decrease in eGFR persisting >24h [11, 12]. However, our study population contained no patients meeting these criteria. To still be able to evaluate the relationship between DO2 and kidney function decrease in this population, we defined a 10% decrease in eGFR (ΔeGFR) compared to the preoperative value as declined postoperative kidney function.
Patients were divided into a group with >10% kidney function decline and a group with less decline or no decline at all. Differences between these groups were then analysed.
Statistical analysis was performed using IBM SPSS, version 23. Data are presented as mean ± standard deviation (sd) where applicable. To compare the DO2 and ΔeGFR groups on pre-, intra- and postoperative characteristics, Mann-Whitney U tests and Chi-squared tests were performed. Analysis of the relation between DO2 and ΔeGFR was done by crosstab analysis with the exact test module [13, 14] and a receiver operating characteristic curve. Oldham correlation analysis was performed to look for relations between DO2 and VO2 per patient, and crosstab analysis with an exact test was used to analyse the effect of low DO2 on these correlations. Blood lactate levels at the start of CPB and at the end of CPB were compared with a paired sample t-test.
The majority of the patients (n = 48) included in this study underwent isolated coronary artery bypass grafting (CABG) or isolated aortic valve replacement (AVR, n = 13). One case was a combined intervention (CABG + AVR) and one case was a mitral valve plasty (MVP). All these procedures were performed under normothermic CPB. Additionally, two procedures were performed using mild hypothermia (one Bentall procedure at 32 degrees Celsius, and one combined MVP + CABG at 34 degrees Celsius).
Mean VO2 was 56 ± 8 ml/min/m2, varying between 38 and 76 ml/min/m2, whereas mean OER was 20 ± 3%. Mean oxygen delivery was 281 ± 39 ml/min/m2, ranging from 181 ml/min/m2 to 415 ml/min/m2. Twenty-seven patients (42%) had an average DO2 <272 ml/min/m2, the other 38 patients (58%) had an average DO2 >272 ml/min/m2. The preoperative EuroSCORE II and the percentage of patients with diabetes were significantly higher in the DO2 <272 group (Table 1), whereas both preoperative and intraoperative hemoglobin levels were significantly higher in the DO2 >272 group. Even though values were still well within the normal physiologic ranges, patients in the DO2 <272 group had a significantly lower PvO2 and SvO2. Intraoperative VO2 and OER did not differ between the low and high DO2 groups, nor were there any significant differences in postoperative eGFR, the percentage of patients with a ΔeGFR >10% or postoperative lactate levels.
The correlation between DO2 and VO2 was strong (r >0.500) in 18 patients, of whom 5 patients were in the DO2 <272 group. Crosstab analysis showed that DO2 <272 was not related to having a correlation between DO2 and VO2 (p = 0.260).
When split on a ΔeGFR of 10%, fifty-five patients (85%) had a ΔeGFR <10%, whereas 10 patients (15%) had a ΔeGFR >10%. Crosstab analysis showed no significant differences between the two DO2 groups concerning the percentage of patients with a ΔeGFR >10% (p = 0.865, Table 2). Moreover, a receiver operating characteristic analysis showed an area under the curve of 0.491 (Fig 1), indicating that DO2 did not predict ΔeGFR >10%.
Comparing the two ΔeGFR groups in terms of other patient and CPB characteristics (Table 3) revealed that patients in the ΔeGFR >10% group had a significantly lower pre-operative hemoglobin level. Intraoperatively however, there was no significant difference in the hemoglobin level. DO2, VO2 and OER did not significantly differ between the ΔeGFR groups.
DO2 –oxygen delivery; ΔeGFR >10%—more than 10% decrease in estimated glomerular filtration rate.
This study was intended to explore the relationship between DO2 and VO2 and AKI in patients undergoing CPB for cardiac surgery at our institution and to verify if the previously found DO2crit of 272 ml/min/m2 was applicable. Twenty-seven patients (42%) had an average DO2 lower than the DO2crit of 272 ml/min/m2, caused by a significantly lower perioperative hemoglobin level. Even though DO2 in this group was lower than the proposed critical DO2 value of 272 ml/min/m2, there were no patients with AKI according to the RIFLE criteria (>25% decrease in eGFR). To still be able to analyse if the low DO2 led to decreased kidney function, we chose to define decreased kidney function as a more than 10% decrease in eGFR compared to the preoperative value. Whether this decrease is clinically relevant, needs to be further determined. Previous research, however, has shown that even small decreases in eGFR and/or small increases in serum creatinine, and even detection of renal injury markers without actual renal function decrease are related to increased morbidity and mortality . Other research, however, has indicated that serum markers might not be reliable during hemodilution . There were 10 patients (17%) with a ΔeGFR >10%, but there was no association between this decreased kidney function and low DO2.
When analysing the VO2/DO2 relationship per individual patient, 18 patients had a positive correlation between DO2 and VO2. This however, showed not to be related to DO2 <272 ml/min/m2. Moreover, other parameters supported that the positive correlation between DO2 and VO2 in these patients did not arise because of pathologic supply dependency and that DO2 was adequate for aerobic metabolism. Central venous oxygen saturation (SvO2) is a routinely used marker of systemic tissue perfusion, reflecting the balance between DO2 and VO2. In our study population SvO2 was 77 ± 3%, which is higher than associated with tissue hypoxia and increased postoperative morbidity and mortality . The OER of 21 ± 3% was in the normal range, indicating sufficient DO2 as well. Furthermore, the lactate level before weaning (1.2 ± 0.3 mmol/l) was in the physiologic range, after a clinically not relevant increase from 1.0 ± 0.3 mmol/l at the start of CPB. A possible explanation for this trivial increase in lactate levels prior to CPB cessation might be washout of lactate accumulated in the heart during the aortic occlusion time. All this indicates that, despite mostly normothermic CPB and “low” DO2 in 42% of the patient population, DO2 seems to have been sufficient to prevent kidney function decrease or the commence of anaerobic metabolism.
Analysing the differences between the two ΔeGFR groups, we found that all the tested parameters were similar, except for the preoperative hemoglobin level. This value however (8.6 ± 0.8 mmol/l in the ΔGFR <10% group and 8.1 ± 0.7 mmol/l in the ΔGFR >10% group), was still well above anaemia levels associated with increased risk of AKI . Moreover, the perioperative hemoglobin levels were similar between groups. The lack of differences between the two groups could be a result of the defined ΔeGFR of 10% which was necessary as the study population contained no patients with AKI according to RIFLE criteria. Some factors that may have contributed to the absence of AKI are the relatively high perioperative mean arterial blood pressure (70 ± 7 mmHg and 71 ± 5 mmHg in the low and high DO2 groups, respectively), which was similar to the optimal blood pressure during CPB found by Hori et al.  and the routine use of pulsatile blood flow that might be protective of renal function . In addition, the patients scored relatively low at some AKI risk factors e.g.: no emergent surgery or patients with cardiogenic shock, 17% female gender; 8% COPD; 5% LVEF <35%; 17% peripheral vascular disease; 18% preoperative renal insufficiency (GFR <60 ml/min/m2) . Whether the average CPB and aortic occlusion time of 99 ± 36 minutes and 66 ± 23 minutes respectively increase or decrease the risk of AKI is not clear, as some studies found that a CPB duration longer than 60 or 90 minutes increases the risk of AKI, whereas others compared AKI patients to non-AKI patients and found that the 100 minute CPB duration of the non-AKI group was lower than the CPB duration in the AKI group [22–25]. An explanation for the higher kidney function decrease in the ΔeGFR >10% group might be the higher incidence of diabetes, a known risk factor for AKI [21, 26].
We were not able to calculate DO2crit levels for patients in this study, as we did not find an association between DO2 and kidney function decrease, nor did we find an indication of the onset of the pathologic supply dependency between DO2 and VO2. The large range of VO2 values (ranging from 38 to 76 ml/min/m2) makes clear though, that it is a highly variable factor depending on many elements and that therefore, a perfusion goal targeting a generic pre-established value for DO2 just above a general DO2crit level cannot guarantee appropriate DO2 for all patients at all times. Indexing DO2 to patient body surface area (as is done for DO2crit levels) allows for differences in physique, but does not take into account e.g. differences in gender and age. Furthermore, the oxygen necessity during CPB depends not only on patient characteristics, but also on factors like anaesthesia technique and temperature management. Thus, when targeting or studying DO2, VO2 should always be taken into account.
This study was solely designed to look for the relationship between DO2, VO2 and kidney function in patients undergoing CPB at our institution, as we questioned the usefulness of universal DO2crit levels from previous research. The limited database compared to the large study populations in previous studies [5–7] resulted in a small sample size. Nevertheless, with research suggesting an incidence of AKI up to 40% after cardiac surgery (depending on the type of surgical procedure and the used definition of AKI) [27–29], we expected to have some patients in our population developing AKI, but the absence thereof resulted in the inability to determine patient specific DO2crit values.
In conclusion, this study could not confirm an evident correlation between low O2 delivery and kidney function decrease in patients undergoing CPB for cardiac surgery. In addition, there was no indication for the presence of a pathologic supply dependency between DO2 and VO2. The variability in O2 consumption, however, is an indication that every patient has individual O2 needs, advocating for an individualized O2 delivery goal instead of a generic DO2crit level.
- 1. Dantzker DR, Foresman B, Gutierrez G. Oxygen supply and utilization relationships. A reevaluation. The American review of respiratory disease. 1991;143(3):675–9. pmid:2001082.
- 2. Pinsky MR. Beyond global oxygen supply-demand relations: in search of measures of dysoxia. Intensive care medicine. 1994;20(1):1–3. pmid:8163751.
- 3. Shibutani K, Komatsu T, Kubal K, Sanchala V, Kumar V, Bizzarri DV. Critical level of oxygen delivery in anesthetized man. Critical care medicine. 1983;11(8):640–3. pmid:6409505.
- 4. Ganushchak YM, Maessen JG, de Jong DS. The oxygen debt during routine cardiac surgery: illusion or reality? Perfusion. 2002;17(3):167–73. pmid:12017383.
- 5. de Somer F, Mulholland JW, Bryan MR, Aloisio T, Van Nooten GJ, Ranucci M. O2 delivery and CO2 production during cardiopulmonary bypass as determinants of acute kidney injury: time for a goal-directed perfusion management? Critical care. 2011;15(4):R192. pmid:21831302; PubMed Central PMCID: PMC3387634.
- 6. Ranucci M, Romitti F, Isgro G, Cotza M, Brozzi S, Boncilli A, et al. Oxygen delivery during cardiopulmonary bypass and acute renal failure after coronary operations. The Annals of thoracic surgery. 2005;80(6):2213–20. pmid:16305874.
- 7. Magruder JT, Dungan SP, Grimm JC, Harness HL, Wierschke C, Castillejo S, et al. Nadir Oxygen Delivery on Bypass and Hypotension Increase Acute Kidney Injury Risk After Cardiac Operations. The Annals of thoracic surgery. 2015;100(5):1697–703. pmid:26271583.
- 8. Magruder JT, Crawford TC, Harness HL, Grimm JC, Suarez-Pierre A, Wierschke C, et al. A pilot goal-directed perfusion initiative is associated with less acute kidney injury after cardiac surgery. The Journal of thoracic and cardiovascular surgery. 2017;153(1):118–25 e1. Epub 2016/11/12. pmid:27832832; PubMed Central PMCID: PMC5517016.
- 9. Ranucci M, Johnson I, Willcox T, Baker RA, Boer C, Baumann A, et al. Goal-directed perfusion to reduce acute kidney injury: A randomized trial. The Journal of thoracic and cardiovascular surgery. 2018;156(5):1918–27 e2. Epub 2018/05/21. pmid:29778331.
- 10. Foundation NK. K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification. Am J Kidney Dis. 2002;39(2 Suppl 1):S1–266. pmid:11904577.
- 11. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, Acute Dialysis Quality Initiative w. Acute renal failure—definition, outcome measures, animal models, fluid therapy and information technology needs: the Second International Consensus Conference of the Acute Dialysis Quality Initiative (ADQI) Group. Critical care. 2004;8(4):R204–12. Epub 2004/08/18. pmid:15312219; PubMed Central PMCID: PMC522841.
- 12. Lopes JA, Jorge S. The RIFLE and AKIN classifications for acute kidney injury: a critical and comprehensive review. Clin Kidney J. 2013;6(1):8–14. Epub 2013/02/01. pmid:27818745; PubMed Central PMCID: PMC5094385.
- 13. Agresti A. Categorical data analysis. Statistics in Medicine. 1990;11(13):1791–2.
- 14. Cochran WG. Some methods for strengthening the common χ 2 tests. Biometrics. 1954;10(4):417–51.
- 15. Ronco C, Kellum JA, Haase M. Subclinical AKI is still AKI. Critical care. 2012;16(3):313. Epub 2012/06/23. pmid:22721504; PubMed Central PMCID: PMC3580601.
- 16. Svenmarker S, Haggmark S, Holmgren A, Naslund U. Serum markers are not reliable measures of renal function in conjunction with cardiopulmonary bypass. Interact Cardiovasc Thorac Surg. 2011;12(5):713–7. Epub 2011/02/08. pmid:21297138.
- 17. Reinhart K, Bloos F. The value of venous oximetry. Current opinion in critical care. 2005;11(3):259–63. pmid:15928476.
- 18. Han SS, Baek SH, Ahn SY, Chin HJ, Na KY, Chae DW, et al. Anemia Is a Risk Factor for Acute Kidney Injury and Long-Term Mortality in Critically Ill Patients. Tohoku J Exp Med. 2015;237(4):287–95. Epub 2015/11/27. pmid:26607258.
- 19. Hori D, Hogue C, Adachi H, Max L, Price J, Sciortino C, et al. Perioperative optimal blood pressure as determined by ultrasound tagged near infrared spectroscopy and its association with postoperative acute kidney injury in cardiac surgery patients. Interactive cardiovascular and thoracic surgery. 2016;22(4):445–51. pmid:26763042; PubMed Central PMCID: PMC4801133.
- 20. Milano AD, Dodonov M, Van Oeveren W, Onorati F, Gu YJ, Tessari M, et al. Pulsatile cardiopulmonary bypass and renal function in elderly patients undergoing aortic valve surgerydagger. European journal of cardio-thoracic surgery: official journal of the European Association for Cardio-thoracic Surgery. 2015;47(2):291–8; discussion 8. pmid:24740935.
- 21. Rosner MH, Okusa MD. Acute kidney injury associated with cardiac surgery. Clinical journal of the American Society of Nephrology: CJASN. 2006;1(1):19–32. pmid:17699187.
- 22. Fischer UM, Weissenberger WK, Warters RD, Geissler HJ, Allen SJ, Mehlhorn U. Impact of cardiopulmonary bypass management on postcardiac surgery renal function. Perfusion. 2002;17(6):401–6. Epub 2002/12/10. pmid:12470028.
- 23. Boldt J, Brenner T, Lehmann A, Suttner SW, Kumle B, Isgro F. Is kidney function altered by the duration of cardiopulmonary bypass? The Annals of thoracic surgery. 2003;75(3):906–12. Epub 2003/03/21. pmid:12645715.
- 24. Yi Q, Li K, Jian Z, Xiao YB, Chen L, Zhang Y, et al. Risk Factors for Acute Kidney Injury after Cardiovascular Surgery: Evidence from 2,157 Cases and 49,777 Controls—A Meta-Analysis. Cardiorenal Med. 2016;6(3):237–50. Epub 2016/06/09. pmid:27275160; PubMed Central PMCID: PMC4886037.
- 25. Sirvinskas E, Andrejaitiene J, Raliene L, Nasvytis L, Karbonskiene A, Pilvinis V, et al. Cardiopulmonary bypass management and acute renal failure: risk factors and prognosis. Perfusion. 2008;23(6):323–7. Epub 2009/05/21. pmid:19454560.
- 26. O'Neal JB, Shaw AD, Billings FTt. Acute kidney injury following cardiac surgery: current understanding and future directions. Critical care. 2016;20(1):187. Epub 2016/07/05. pmid:27373799; PubMed Central PMCID: PMC4931708.
- 27. Chang TI, Leong TK, Boothroyd DB, Hlatky MA, Go AS. Acute kidney injury after CABG versus PCI: an observational study using 2 cohorts. Journal of the American College of Cardiology. 2014;64(10):985–94. pmid:25190232.
- 28. Helgason D, Helgadottir S, Viktorsson SA, Orrason AW, Ingvarsdottir IL, Geirsson A, et al. Acute kidney injury and outcome following aortic valve replacement for aortic stenosis. Interactive cardiovascular and thoracic surgery. 2016;23(2):266–72. pmid:27127185.
- 29. Mangano CM, Diamondstone LS, Ramsay JG, Aggarwal A, Herskowitz A, Mangano DT. Renal dysfunction after myocardial revascularization: risk factors, adverse outcomes, and hospital resource utilization. The Multicenter Study of Perioperative Ischemia Research Group. Annals of internal medicine. 1998;128(3):194–203. pmid:9454527.