Acute kidney injury (AKI), which occurs in as many as 30% of patients after cardiac surgery, is associated with impaired outcome.1,2 AKI is the consequence of an interplay of several pathophysiological mechanisms, including preoperative patient-related, intraoperative and postoperative factors.3,4 Factors influencing renal perfusion may play an important role in the AKI process. During the postoperative period, low cardiac output syndrome, vasoactive drugs, diuretic use and fluid balance can also alter renal function.5,6 Recent reports in the field of medical cardiology have associated venous congestion due to right ventricular dysfunction (RVd) with deterioration of renal function.7–10 RVd is a well known complication of cardiac surgery11–15 and several studies have confirmed alteration of right ventricular systolic and diastolic function, resulting in impaired venous return and postoperative circulatory failure.16–19 Despite the high incidence of RVd after cardiac surgery and the documented relationship in medical cardiology between RVd and renal dysfunction, no studies have evaluated the correlation between right ventricular function and renal function after cardiac surgery in adult patients.
The primary objective of this study was to investigate a temporal association between RVd assessed by transthoracic echocardiography (TTE) on admission to the ICU and a subsequent increase of postoperative serum creatinine concentration (sCr). We also investigated the hypothesis that venous congestion associated with RVd might be the main mechanism involved.
Materials and methods
This study was approved by the Amiens University Hospital Institutional Review Board (IRB) for humans, which waived the need for informed consent (CEERNI no. 53 de 2011, Comité de Protection des Personnes Nord-Ouest II CHU - Place V. Pauchet, 80054 AMIENS Cedex 1). The STROBE Statement checklist for cohort studies was used to establish this manuscript.20
A prospective, observational study was conducted in the Amiens University Hospital surgical ICU over a 3-month period in 2011. All patients who had undergone scheduled or emergency on-pump coronary artery bypass graft (CABG) surgery, valvular surgery, combined surgery or thoracic aortic procedures were included in the study. Patients who had undergone off-pump CABG, pulmonary or tricuspid valve surgery, repeat surgery or who underwent immediate postoperative renal replacement therapy were excluded.
Anaesthesia and cardiopulmonary bypass management were standardised for all patients. Induction of anaesthesia was performed with etomidate or propofol, and sufentanil. Tracheal intubation was facilitated with atracurium. Maintenance of anaesthesia was achieved by target-controlled infusions of propofol and sufentanil. Titration of hypnotics and opioids was based on bispectral index (BIS; Covidien, Boulder, Colorado, USA) according to the published algorithms. Cardiopulmonary bypass with a heart-lung machine (Stockert Sorin S5 Heart Lung, Milan, Italy) was performed at a target blood flow of 2.4 l min−1 m−2 and mean arterial blood pressure (MAP) was maintained at more than 60 mmHg first by increasing pump flow or, if blood pressure did not improve, by bolus administration of phenylephrine or norepinephrine. During cardiopulmonary bypass, normothermia (bladder temperature >36°C) was maintained with a venous perfusion temperature of 37°C. Homologous red blood cell transfusion was given to patients with a haemoglobin value less than 8 g dl−1. Myocardial protection was ensured with anterograde (via the aortic root or coronary ostia) and/or retrograde (coronary sinus) cardioplegia, depending on the procedure. Intraoperative haemofiltration was not performed in any of the patients in the study.
At the end of surgery, all patients were sedated with propofol and the lungs were mechanically ventilated until haemodynamic stability and normothermia were obtained and blood loss was considered acceptable (less than 1 ml kg−1 h−1). Patients were treated by physicians trained in postoperative cardiac surgical care, including a cardiologist. Circulatory support was guided by institutional protocols to achieve predefined endpoints: MAP more than 70 mmHg, cardiac index (CI) more than 2.2 l min−1 m−2 and urine output more than 0.5 ml kg−1 h−1.
Demographic, intraoperative and postoperative echocardiographic and haemodynamic data were recorded prospectively during the study period. Echocardiographic data were measured on the day before surgery and immediately after surgery on admission to ICU. Fluid balance, urine output, diuretic and vasoactive treatments were recorded between admission to ICU and the first postoperative day (POD1).
TTE was performed by a physician experienced in echocardiography and blinded to the study outcomes using a Philips Envisor ultrasound system (Philips Medical System, Suresnes, France). Left ventricular ejection fraction (LVEF) was measured by Simpson's biplane method on a four-chamber view. The aortic annulus diameter (AoD) was measured on a long-axis parasternal view at the time of inclusion of the patient. Aortic area was calculated by: SAo (cm2) = (π × AoD2)/4. Aortic velocity-time integral (VTIAo) was measured with pulsed Doppler on a five-chamber apical view. Stroke volume (SV) was calculated by: SV (ml) = VTIAo × SAo. Cardiac output (CO) was calculated using the formula: CO (ml min−1) = SV × heart rate (HR); CI was calculated as CI (ml min−1 m−2) = CO/body surface area. Right ventricular ejection fraction (RVEF) was measured by Simpson's biplane method on a four-chamber view. Tricuspid annular systolic velocity at the lateral wall [Sr(t)] and M-mode annular systolic excursion plane [tricuspid annular plane systolic excursion (TAPSE)] were measured by placing the tissue-Doppler pulse wave and M-mode sample volume at the level of the basal right ventricular free wall. Right ventricular dilatation was estimated qualitatively from multiple views and graded as ‘no dilatation’ or ‘dilatation’. Inferior vena cava (IVC) diameter was measured on a subcostal view. Mean echocardiographic parameters were the average of five measurements (regardless of the respiratory cycle). Central venous pressure (CVP) and arterial pressure were measured with a transducer zeroed at the level of the mid-axillary line in the supine position.
Due to the complex geometry and lack of accepted standards for echocardiographic evaluation of right ventricular function, RVd was defined as alteration in at least two variables among RVEF, TAPSE, Sr(t) and right ventricular dilatation.8 As no cut-off values have been defined for TAPSE, Sr(t) and RVEF after cardiac surgery, TAPSE, Sr(t) and RVEF values were divided into quartiles and values within the lowest quartile were considered to reflect significant RVd. Quartiles were defined as follows: RVEF less than 35, 35 to 40, 40 to 50 and more than 50%; TAPSE less than 8.8, 8.8 to 12.2, 12.2 to 14.8 and more than 14.8 mm; Sr(t) less than 6.3, 6.3 to 8.1, 8.1 to 9.2 and more than 9.2 cm s−1.
In the context of cardiac surgery, the expected course of sCr in the absence of renal dysfunction is a decrease in sCr on POD1 as a result of cardiopulmonary bypass induced haemodilution.19 Because an increase in sCr and its duration during the postoperative period have been associated with morbidity and mortality, the primary endpoint of this study was defined as a positive variation of sCr on POD1 compared with the preoperative value.21 The postoperative variation of sCr was calculated as the variation between the preoperative value and the value on the first postoperative day (POD1): sCrvar = [(sCr measured on POD1 – preoperative sCr)/preoperative sCr] x 100.
Secondary endpoints were CVP and CO on admission to ICU, duration of mechanical ventilation, vasopressor and inotropic drug use, length of ICU stay and incidence of AKI. AKI was defined as an increase of sCr greater than 26.5 μmol l−1 (0.3 mg dl−1) within 48 h and/or an increase of sCr greater than 1.5 times baseline over 1 week (KDIGO criteria).22 Postoperative fluid balance from admission to ICU until POD1 (T0-POD1) was calculated as: fluid balance (ml) = fluid volume infused – (urine output + chest tube drainage).
The distribution of variables was assessed using the D’Agostino–Pearson test. Results are expressed as median with 25th to 75th percentiles, or numbers with proportions (%), as appropriate. The study population was divided into groups according to the presence or absence of RVd. Wilcoxon, Mann–Whitney, Kruskal–Wallis or Fisher's exact tests were used, as appropriate. Spearman's method was used to test linear correlations. Association between RVd and sCrvar was assessed using a multivariate logistic model with sCrvar as the dependent variable. Logistic regression was performed with a dichotomous variable: presence or absence of a postoperative increase of sCr. Variables with a P value less than 0.10 were included in the multivariate model with a backward selection procedure. Four variables were selected: RVd, diuretic treatment, norepinephrine treatment and postoperative transfusion. Because the incidence of chronic renal failure and baseline sCr values did not differ between the two groups, these variables were not included in the logistic regression model. Validity of the logistic regression was evaluated using the percentage of patients correctly classified, the C-index and the Hosmer–Lemeshow test. The percentage of cases correctly classified was 74%, the C-index was 0.82 [95% confidence interval (95% CI) 0.72 to 0.9] and the Lemeshow χ2 test was 1.1 (P = 0.56). Differences with a P value less than 0.05 were considered statistically significant. Statistical analysis was performed with SPSS 21 (IBM, Chicago, Illinois, USA).
Ninety-nine of 109 consecutive patients undergoing cardiac surgery during the study period were included. Ten patients were not included due to lack of investigators trained in echocardiography, notably in the evening and at weekends. Among the patients initially included, 25 patients were excluded due to incomplete echocardiographic data related to poor echogenicity on admission to ICU. Data from 74 patients were therefore analysed (Fig. 1). The patients’ preoperative and intraoperative characteristics are summarised in Table 1.
Right ventricular echocardiographic variables
All echocardiographic variables decreased after surgery (P < 0.05); LVEF decreased from 60% (50 to 65) to 50% (45 to 60) (P = 0.07), RVEF from 50% (49 to 60) to 40% (35 to 50) (P < 0.001), TAPSE from 22.3 mm (19.4 to 25.3) to 12.2 mm (8.8 to 14.8) (P < 0.001) and Sr(t) from 15.0 cm s−1 (12.0 to 18.0) to 8.1 cm s−1 (6.3 to 9.2) (P < 0.001). IVC diameter increased from 11.5 mm (9.6 to 15.0) to 15.2 mm (11.8 to 19.1) (P = 0.004). Fourteen (19%) patients had right ventricular dilatation. Cardiopulmonary bypass duration and echocardiographic parameters of right ventricular/left ventricular function were not correlated (P > 0.05).
On admission to ICU, all right ventricular variables were moderately correlated: Sr(t) and TAPSE (r = 0.38, P = 0.001), Sr(t) and RVEF (r = 0.43, P = 0.001), TAPSE and RVEF (r = 0.47, P = 0.001), TAPSE and IVC (r = –0.28, P = 0.02), IVC and RVEF (r = –0.42, P = 0.001), IVC and Sr(t) (r = –0.25, P = 0.04). RVd, defined by composite criteria, was present in 23 out of 74 (31%) patients.
Relationship between right ventricular dysfunction and acute kidney injury
At baseline and on admission to ICU, sCr was not significantly different between patients with RVd and those without RVd, whereas on POD1, sCr was higher in patients with RVd (Table 2). Patients with RVd had a higher incidence of AKI and poorer outcomes than patients without RVd. None of the variables of right ventricular and left ventricular function measured at baseline were correlated with sCrvar (P > 0.05). On admission to ICU, only echocardiographic parameters of right ventricular function and CVP were correlated with sCrvar (Table 3). CVP values increased significantly with each echocardiographic parameter of RVd (RVEF, TAPSE, Sr(t) and right ventricular dilatation) (P < 0.05) (Fig. 2). On POD1, sCrvar increased significantly with each echocardiographic positive value of RVd (P < 0.05), as shown in Fig. 2. Forty (54%) of the patients had a positive sCrvar on POD1. Four variables were included in multivariate logistic regression analysis: RVd, diuretic use, postoperative transfusion and norepinephrine treatment. An association was shown only between sCrvar and RVd [odds ratio (OR) 12.7, 95% CI 2.6 to 63.4, P = 0.02] and diuretic use (OR 5.2, 95% CI 1.5 to 18.3, P = 0.01) (Table 4).
Relationship between right ventricular dysfunction and secondary endpoints
On admission to ICU, patients with RVd had higher values of CVP, whereas MAP, HR and CI were not significantly different between patients with and without RVd (P > 0.05) (Table 2). CVP values were correlated with TAPSE (r = -0.49, P = 0.001), RVEF (r = -0.45, P = 0.001) and IVC (r = 0.59, P = 0.001). CVP values were not correlated with Sr(t) (r = –0.19, P = 0.1), cardiopulmonary bypass duration (r = 0.19, P = 0.11) or intraoperative fluid volume (r = 0.17, P = 0.14). CVP increased significantly with each echocardiographic parameter of RVd, as shown in Fig. 2. Patients with RVd had longer intubation durations, longer ICU length of stay and higher rates of norepinephrine treatment (Table 2).
This study showed that RVd, defined by a composite criterion comprising more than two altered echocardiographic parameters diagnosed on admission to ICU, was associated with an increased risk of AKI. Patients with RVd also had higher CVP values, although CI was similar between the two groups. In this cohort, CI may not be the main factor influencing renal function, as venous congestion associated with RVd may also alter renal function.
Although RVd is a well known complication of cardiac surgery, no previous studies have evaluated its impact on renal function in this setting. The degree of right ventricular dysfunction has been correlated with haemodynamic outcomes (difficult cardiopulmonary bypass weaning, circulatory failure), which are known AKI risk factors.23,24 Several authors have demonstrated an alteration of right ventricular function lasting up to several months after surgery.11–14 However, using three-dimensional echocardiography, Tamborini et al.15 advised caution when interpreting TAPSE and Doppler tissue imaging. These authors demonstrated an alteration of the long-axis function of the right ventricle (RV) related to geometrical changes due to cardiac surgery. Due to its complex geometry, the lack of accepted standard definitions for right ventricular dysfunction and the consequences of anaesthesia/mechanical ventilation/surgery on echocardiographic variables, no consensual echocardiographic definition of RVd after cardiac surgery is available. We therefore used a ‘conservative approach’, favouring specificity over sensitivity to define RVd that included severely altered echocardiographic parameters and association of at least two alterations. The justification for this approach was based on the modest statistical correlations between various criteria of right ventricular function (less than 20% of the variability of one criterion explained by the values of another criterion) and, given the reproducibility of echocardiography, the use of more than one altered variable could avoid false-positive diagnosis by excluding discordant measures. Finally, using these high specificity/low sensitivity criteria, 31% of our patients were classified as having RVd, and this group had a significantly poorer postoperative course after cardiac surgery (higher incidence of complications, higher vasoactive drug use, longer ICU stay).
One of the possible haemodynamic mechanisms for impaired renal function could be venous congestion.7 Since the 1990s, abnormalities of systemic venous physiology after cardiac surgery have been demonstrated by jugular or hepatic blood flow studies.17–19 Wranne et al.17 demonstrated a correlation between alterations of venous blood flow and right ventricular function. On the basis of a series of 26 patients, they suggested that a marked decrease in tricuspid annular motion was diagnostic of early alteration of right heart filling, as reflected by altered hepatic venous blood flow. Our results confirm an association between RVd and venous congestion. In the present study, patients with RVd had higher CVP and IVC diameter values. More specifically, all right ventricular echocardiographic variables were correlated with CVP values, and CVP values increased with the number of echocardiographic parameters of RVd (Fig. 2). High CVP values may reflect several mechanisms. In the present study, high CVP could be secondary to increased right heart filling pressure due to impaired right ventricular diastolic function. Elevated CVP values may also reflect fluid overload or decreased venous compliance secondary to activation of the sympathetic nervous system and/or norepinephrine infusion. However, on admission to ICU, intraoperative fluid balance was not associated with CVP values, and norepinephrine requirements did not differ between patients with or without RVd.
Our results show a statistical association between diuretic use and an increased risk of renal dysfunction on POD1. Diuretics are a well known risk factor for AKI in the cardiac surgery setting.25 High CVP values were probably caused by decreased venous compliance related to RVd and not to increased preload/hypervolaemia that could have justified prescription of diuretics. In this context, diuretics may not improve compliance and may decrease blood volume, thereby potentially worsening renal function. Interpretation of these results is only speculative at the present time, and further prospective interventional trials are required.
Limitations of this study
This study presents a number of potential weaknesses. In the absence of preliminary data in cardiac surgery, we designed a prospective observational (exploratory) study with a convenience sample of 74 consecutive patients. We chose to assess renal function by sCr variation, because slight elevations are frequently observed and have been demonstrated to be associated with mortality after cardiac surgery.21,26,27 The diagnostic criteria of renal dysfunction have recently been revised.22 The relevance of these criteria for the postcardiac surgery period has been confirmed by several studies.3,4 In the context of cardiac surgery, the expected variation of sCr in the absence of renal dysfunction is a decrease due to cardiopulmonary bypass-induced haemodilution.19 Therefore, any increase in sCr after cardiac surgery is an even more sensitive marker of renal dysfunction than KDIGO stage 1. As expected, the two criteria of renal dysfunction were related, and a higher proportion of patients had a positive change in sCr than stage 1 KDIGO renal dysfunction.26 Similarly, early recovery of renal function after cardiac surgery has been shown to improve long-term survival, and may represent a therapeutic goal.27 The number of patients excluded due to incomplete echocardiographic data may appear to be high, but is consistent with the results of a previous study.28 We did not specifically evaluate right ventricular diastolic function due to methodological difficulties and the lack of universally accepted criteria for right ventricular diastolic dysfunction. Furthermore, we did not use acceleration of isovolumetric contraction, a parameter that has gained increasing popularity for the assessment of right ventricular function, due to its relative load-independency.29 Right ventricular dysfunction was defined by a composite criterion combining more than two echocardiographic parameters of significant right ventricular dysfunction. However, due to the complexity of RV, and the lack of accepted standards for echocardiographic evaluation of right ventricular function, we used a previously described approach.7 Our results are therefore consistent with those of previous studies, but further prospective studies are needed to confirm these results by using complementary techniques such as invasive assessment of pressure–volume loops with right ventricular arterial coupling.
The present study demonstrated that early RVd defined by a composite echocardiographic criterion was associated with an increase in sCr. Patients with RVd had a higher incidence of AKI. One possible mechanism would be via the venous compartment that plays a role in the haemodynamic regulation of renal function. Further studies are required to confirm this association and to test the hypothesis that early intervention could improve outcome.
Acknowledgements relating to this article
Assistance with the study: none.
Financial support and sponsorship: the authors performed this study in the course of their normal duties as full-time employees of public healthcare institutions. The study received no financial support.
Conflicts of interest: none.
Presentation: preliminary data for this study were presented at 25th ESICM Annual Congress, 2012.
1. Chertov GM, Levy EM, Hammermeister KE, et al. Independent association between acute renal failure and mortality following cardiac surgery. Am J Med
2. Thakar CV, Worley S, Arrigain S, et al. Influence of renal dysfunction on mortality after cardiac surgery: modifying effect of preoperative renal function. Kidney Ant
3. Mariscalco G, Lorusso R, Dominici C, et al. Acute kidney injury: a relevant complication after cardiac surgery. Ann Thorac Surg
4. Provenchere S, Plantefeve G, Hufnagel G, et al. Renal dysfunction after cardiac surgery with normothermic cardiopulmonary bypass: incidence, risk factors, and effect on clinical outcome. Anesth Analg
5. Brienza M, Giglio MT, Marucci M, Fiore T. Does perioperative hemodynamic optimization protect renal function in surgical patients? A meta-analytic study. Crit Care Med
6. Shahin J, deVarennes B, Wing Tse C, et al. The relationship between inotrope exposure, six-hour physiological variables, and hospital mortality and renal dysfunction in patients undergoing cardiac surgery. Crit Care
7. Mullens W, Abrahams Z, Francis GS, et al. Importance of venous congestion for worsening of renal function in advanced decompensated heart failure. J Am Coll Cardiol
8. Testani JM, Khera AV, St John Sutton MG, et al. Effect of right ventricular function and venous congestion on cardio-renal interactions during the treatment of decompensated heart failure. Am J Cardiol
9. Uthoff H, Breidthardt T, Klima T. Central venous pressure and impaired renal function in patients with acute heart failure. Eur J Heart Fail
10. Damman K, Voors AA, Hillege HL, et al. Congestion in chronic systolic heart failure is related to renal dysfunction and increased mortality. Eur J Heart Fail
11. Pegg TJ, Selvanayagam JB, Karamitsos TD, et al. Effects of off-pump versus on-pump coronary artery bypass grafting on early and late right ventricular function. Circulation
12. Alam M, Hedman A, Nordlander R, Samad B. Right ventricular function before and after an uncomplicated coronary artery bypass graft as assessed by pulsed wave Doppler tissue imaging of the tricuspid annulus. Am Heart J
13. Diller GP, Wasan BS, Kyriacou A, et al. Effect of coronary artery bypass surgery on myocardial function as assessed by tissue Doppler echocardiography. Eur J Cardiothorac Surg
14. Siddiqui MM, Jalal A, Sherwani M, Ahmad MZ. Right ventricular dysfunction after coronary artery bypass grafting is a reality of unknown cause and significance. Heart Surg Forum
15. Tamborini G, Muratori M, Brusoni D, et al. Is right ventricular systolic function reduced after cardiac surgery? A two- and three dimensional echocardiographic study. Eur J Echocardiogr
16. Mishra M, Swaminathan M, Malhotra R, et al. Evaluation of right ventricular function during CABG: transesophageal echocardiographic assessment of hepatic venous flow versus conventional right ventricular performance indices. Echocardiography
17. Wranne B, Pinto FJ, Hammarström E, et al. Abnormal right heart filling after cardiac surgery: time course and mechanisms. Br Heart J
18. Normura T, Lebowitz L, Koide Y, et al. Evaluation of hepatic venous flow using transesophageal echocardiography in coronary artery bypass surgery: an index of right ventricular function. J Cardiothorac Vasc Anesth
19. Purkiss SF, Fort S, Graham TR, et al. Hepatic portal venous flow in patients undergoing tricuspid valve surgery. Br Heart J
20. von Elm E, Altman DG, Egger M, et al. STROBE initiative. Strengthening the Reporting of Observational Studies in Epidemiology (STROBE) statement: guidelines for reporting observational studies. BMJ
21. Ishani A, Nelson D, Clothier B, et al. The magnitude of acute serum creatinine increase after cardiac surgery and the risk of chronic kidney disease, progression of kidney disease, and death. Arch Intern Med
22. Kidney Disease: Improving Global Outcomes (KDIGO) acute kidney injury work group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl
23. Costachescu T, Denault A, Guimond J-G, et al. The hemodynamically unstable patient in the intensive care unit: hemodynamic vs transesophageal echocardiographic monitoring. Crit Care Med
24. Denault AY, Couture P, Buithieu J, et al. Left and right ventricular diastolic dysfunction as predictors of difficult separation from cardiopulmonary bypass. Can J Anaesth
25. Lassnigg A, Donner E, Grubhofer G, et al. Lack of renoprotective effects of dopamine and furosemide during cardiac surgery. J Am Soc Nephrol
26. Kolli H, Rajagopalam S, Patel N, et al. Mild acute kidney injury is associated with increased mortality after cardiac surgery in patients with eGFR < 60 mL/min/1.73 m2. Ren Fail
27. Swaminathan M, Hudson CC, Phillips-Bute BG, et al. Impact of early recovery on survival after cardiac surgery-associated acute kidney injury. Ann Thorac Surg
28. Flynn BC, Spellman J, Bodian C, Moitra VK. Inadequate visualization and reporting of ventricular function from transthoracic echocardiography after cardiac surgery. J Cardiothorac Vasc Anesth
© 2015 European Society of Anaesthesiology
29. Vogel M, Schmidt M, Kristiansen S, et al. Validation of myocardial acceleration during isovolumic contraction as a novel noninvasive index of right ventricular contractility. Circulation