ORIGINAL ARTICLE
Heart, Lung and Circulation (2014) 23, 325–331 1443-9506/04/$36.00 http://dx.doi.org/10.1016/j.hlc.2013.10.082
Mortality Prediction in Patients with Dialysis-dependent Acute Kidney Injury after Cardiac Surgery with Cardiopulmonary Bypass Andrey A. Malov, MD, PhD*, Alexander S. Borisov, MD, PhD, Vladimir V. Lomivorotov, MD, PhD, Sergey M. Efremov, MD, PhD, Dmitry N. Ponomarev, MD, Tamara V. Mukhoedova, MD, PhD, Alexander M. Karaskov, MD, PhD Department of Anesthesiology and Intensive Care, Research Institute of Circulation Pathology, Novosibirsk, Russian Federation Received 17 June 2013; received in revised form 12 October 2013; accepted 17 October 2013; online published-ahead-of-print 25 October 2013
Background
To build a predictive model for patients with dialysis-dependent acute kidney injury (AKI-D) after cardiac surgery with the cardiopulmonary bypass (CPB), according to disease severity.
Methods
A single-centre, retrospective cohort study was performed to determine the demographic and clinical parameters (including the specific factor, CPB duration) for risk of poor outcome in patients requiring RRT after cardiac surgery with CPB. A new model was built for mortality prediction in these patients on the basis of the identified risk factors and Sequential Organ Failure Assessment score.
Results
The newly developed model showed good discriminatory ability for predicting death in patients with AKID after cardiac surgery with CPB. The area under the receiver-operating characteristic (ROC) curve for the score was 0.892 (95% confidence interval, 0.852–0.925). We also determined the criterion for the choice of RRT modality by applying this model. On applying the new model in intermittent haemodialysis patients, a score of 3.2 was found safe for selecting the RRT modality.
Conclusions
The new scoring system was valid and accurate in predicting death for AKI-D patients after open-heart surgery. This system and value for choice of RRT were determined for guidance only, to facilitate decisionmaking in difficult situations.
Keywords
Acute kidney injury Mortality Cardiac surgery Dialysis-dependent Renal replacement therapy modality
Introduction Acute kidney injury (AKI) occurring after cardiac surgery can entail serious complications associated with significant morbidity. The mortality rate among AKI patients after cardiac surgery is high, with a wide range 2.4–19.0% [1–4]. In the most severe form of AKI, i.e., dialysis-dependent AKI
(AKI-D), mortality increases significantly, at more than 50% of the total affected individuals [5]. Currently, renal replacement therapy (RRT) has emerged as an essential component of the critical care of patients with severe AKI. No evidence has conclusively shown that continuous RRT (CRRT) is superior to intermittent haemodialysis (IHD) in reducing the mortality rate [6,7]. Nevertheless, data from the
*Corresponding author at: Department of Anesthesiology and Intensive Care, Research Institute of Circulation Pathology, 630055, Rechkunovskaya Str. 15, Novosibirsk, Russian Federation. Tel.: +7 383 332 7622; fax: +7 383 332 7622., Email: [email protected] © 2013 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier Inc. All rights reserved.
326
Program to Improve Care in Acute Renal Disease (PICARD) study showed that patients treated with IHD continued to accumulate fluid during the period between two dialysis sessions and that fluid balance was better controlled by CRRT [8]. This implies that IHD may be less suitable for haemodynamically unstable patients with oliguria. Further, a recent meta-analysis [9] showed that initial treatment with intermittent RRT might be associated with higher rates of dialysis dependence than CRRT. Evidence from some studies suggests that for haemodynamically unstable and oliguric patients, CRRT is preferable to IHD to achieve fluid and metabolic control [10,11]. This is a very important point to be considered during the choice of RRT in patients who have undergone cardiac surgery with cardiopulmonary bypass (CPB) because AKI in these patients is often associated with multiple organ failure. Under these circumstances, logic demands that the choice of the treatment be based on the severity of illness and associated organ dysfunctions. However, the treatment selection is based on subjective criteria. Several prognostic tools have been developed to assess disease severity and the risk of poor outcome, but none of them account for risk factors specific to patients who have undergone cardiac surgery, such as CPB duration and surgery type. Both general illness severity scores and kidney-specific disease severity scores do not identify these risk factors; therefore, such scoring systems are not entirely suitable for patients with AKI-D after cardiac surgery with CPB. Most known among the recent disease-specific models for AKI-D patients are the Cleveland Clinic Foundation score [12] and the score developed in the Veterans Affairs/National Institutes of Health (VA/NIH) trial [13]. The former was designed to predict the postoperative AKI-D, but not the outcome, while the latter was developed for the common population of AKI-D and does not account for the important, specific factor pertaining to cardiac surgery, i.e., the CPB duration. Assessment of the disease severity with consideration of the specific risk factors may be helpful for the choice of the RRT modality in patients with AKI-D after cardiac surgery. The development of a specific scoring model based on disease severity will facilitate the safe choice of an appropriate modality of RRT and, consequently, avoid the needless economic costs associated with the unnecessary prescription of an expensive treatment.
Patients and Methods Study Population We retrospectively analysed the data of 321 patients who developed AKI-D after open-heart surgery with cardiopulmonary bypass (CPB). Data were extracted from the patient records of surgeries performed at the Novosibirsk Research Institute of Circulation Pathology (NRICP) between September 2003 and July 2010. A total of 10,685 patients aged above 18 years underwent CPB during the time of the study. Of these, 314 (2.94%) patients developed AKI-D. All patients receiving RRT after surgery were assessed for inclusion in the study.
A.A. Malov et al.
Table 1 General Characteristics of the Patients. Data Are Presented as Numbers or Median (25th; 75th Percentile). N = 300
%
Gender Male
170
56.7
Female
130
43.3
Age, years
59 (51; 66)
Surgery type Valve surgery
150
50
CABGa Aortic arch surgery Combined surgery
49
16.3
39
13
62
20.7
Cardiopulmonary bypass
180 (120; 236)
duration, min a
Coronary artery bypass grafting.
Exclusion criteria were preoperative need for RRT, including those with chronic renal disease (CRD) requiring dialysis (n = 14) and those with non-renal indications of RRT (n = 7). The derivation dataset comprised data obtained retrospectively from 300 patients aged between 20 and 79 years. The patients had undergone coronary artery bypass graft surgery (CABG), heart valve (HV) surgery, aortic arch (AA) surgery, or combined procedures (Table 1). The RRT modality distribution showed that 124 patients (41.3%) received IHD and 176 (58.7%) received CRRT. In the case of 61 patients (20.3%), the RRT modality was switched during treatment. The mortality rates among the IHD patients and CRRT patients were 15.4% and 59.3%, respectively. Finally, 60 patients who developed AKI-D after undergoing open-heart surgery with CPB in NRICP between September 2011 and April 2012 were assessed prospectively, and the data of these patients was used to form the validation dataset. CRRT was performed using the Multifiltrate (Fresenius Medical Care) and Prisma and Prismaflex (Gambro Medical). The blood flow rate maintained during the procedures was 100–200 mL min1, and the dialysate flow rate was 1.5– 2.5 L h1. The CRRT duration varied widely from 18 h to 28 days (mean SD, 72 32 h). The effluent dose ranged from 30 to 100 mL kg1 min1. IHD therapy was carried out on stationary dialysis machines (models 4008B, 4008S, and 5008; Fresenius Medical Care). The treatment duration varied widely from 2.5 to 8 h (3.4 2.2 h). Therapies were generally performed with a dialysate flow of 6, 12, and 18 L h1, and sometimes, 30 L h1 and a blood flow rate of 100– 250 mL min1. The decision to begin dialysis was based on clinical judgement, and the indications included azotemia, volume overload, and biochemical abnormalities.
Predictive Model Building First, we estimated the predictive ability of five well-known severity scores for comparison with that of the newly
327
Mortality in Post-CPB AKI
developed model. The five severity scores were as follows: Acute Physiology and Chronic Health Evaluation (APACHE) II as a general score [14]; Multiple Organ Dysfunction (MODS) [15], Sequential Organ Failure Assessment (SOFA) [16], and Organ System Failure (OSF) [17], as organfailure scores; and Liano’s score as a kidney-specific disease severity score [18]. Preference was given to organ failure scores because AKI-D after cardiac surgery is usually a component of multiple organ failure. Liano’s score was chosen because among all the kidney-specific severity scores, this score has been externally validated most frequently [19]. All scores were calculated at the time of RRT initiation. We examined 18 variables as possible predictors of a fatal outcome: demographic data (age, gender); preoperative serum creatinine level, as an indicator of baseline renal function; surgical data (CPB duration, types of cardiac surgery, and RRT initiation day); and 12 postoperative variables (serum creatinine level, oliguria, number of sympathomimetic agents administered, mechanical lung ventilation [MLV], perioperative myocardial infarction [MI], oxygenation index [PO2/FiO2], bilirubinaemia, albuminaemia, platelet count, cerebral coma [Glasgow score of
Heart, Lung and Circulation (2014) 23, 325–331 1443-9506/04/$36.00 http://dx.doi.org/10.1016/j.hlc.2013.10.082
Mortality Prediction in Patients with Dialysis-dependent Acute Kidney Injury after Cardiac Surgery with Cardiopulmonary Bypass Andrey A. Malov, MD, PhD*, Alexander S. Borisov, MD, PhD, Vladimir V. Lomivorotov, MD, PhD, Sergey M. Efremov, MD, PhD, Dmitry N. Ponomarev, MD, Tamara V. Mukhoedova, MD, PhD, Alexander M. Karaskov, MD, PhD Department of Anesthesiology and Intensive Care, Research Institute of Circulation Pathology, Novosibirsk, Russian Federation Received 17 June 2013; received in revised form 12 October 2013; accepted 17 October 2013; online published-ahead-of-print 25 October 2013
Background
To build a predictive model for patients with dialysis-dependent acute kidney injury (AKI-D) after cardiac surgery with the cardiopulmonary bypass (CPB), according to disease severity.
Methods
A single-centre, retrospective cohort study was performed to determine the demographic and clinical parameters (including the specific factor, CPB duration) for risk of poor outcome in patients requiring RRT after cardiac surgery with CPB. A new model was built for mortality prediction in these patients on the basis of the identified risk factors and Sequential Organ Failure Assessment score.
Results
The newly developed model showed good discriminatory ability for predicting death in patients with AKID after cardiac surgery with CPB. The area under the receiver-operating characteristic (ROC) curve for the score was 0.892 (95% confidence interval, 0.852–0.925). We also determined the criterion for the choice of RRT modality by applying this model. On applying the new model in intermittent haemodialysis patients, a score of 3.2 was found safe for selecting the RRT modality.
Conclusions
The new scoring system was valid and accurate in predicting death for AKI-D patients after open-heart surgery. This system and value for choice of RRT were determined for guidance only, to facilitate decisionmaking in difficult situations.
Keywords
Acute kidney injury Mortality Cardiac surgery Dialysis-dependent Renal replacement therapy modality
Introduction Acute kidney injury (AKI) occurring after cardiac surgery can entail serious complications associated with significant morbidity. The mortality rate among AKI patients after cardiac surgery is high, with a wide range 2.4–19.0% [1–4]. In the most severe form of AKI, i.e., dialysis-dependent AKI
(AKI-D), mortality increases significantly, at more than 50% of the total affected individuals [5]. Currently, renal replacement therapy (RRT) has emerged as an essential component of the critical care of patients with severe AKI. No evidence has conclusively shown that continuous RRT (CRRT) is superior to intermittent haemodialysis (IHD) in reducing the mortality rate [6,7]. Nevertheless, data from the
*Corresponding author at: Department of Anesthesiology and Intensive Care, Research Institute of Circulation Pathology, 630055, Rechkunovskaya Str. 15, Novosibirsk, Russian Federation. Tel.: +7 383 332 7622; fax: +7 383 332 7622., Email: [email protected] © 2013 Australian and New Zealand Society of Cardiac and Thoracic Surgeons (ANZSCTS) and the Cardiac Society of Australia and New Zealand (CSANZ). Published by Elsevier Inc. All rights reserved.
326
Program to Improve Care in Acute Renal Disease (PICARD) study showed that patients treated with IHD continued to accumulate fluid during the period between two dialysis sessions and that fluid balance was better controlled by CRRT [8]. This implies that IHD may be less suitable for haemodynamically unstable patients with oliguria. Further, a recent meta-analysis [9] showed that initial treatment with intermittent RRT might be associated with higher rates of dialysis dependence than CRRT. Evidence from some studies suggests that for haemodynamically unstable and oliguric patients, CRRT is preferable to IHD to achieve fluid and metabolic control [10,11]. This is a very important point to be considered during the choice of RRT in patients who have undergone cardiac surgery with cardiopulmonary bypass (CPB) because AKI in these patients is often associated with multiple organ failure. Under these circumstances, logic demands that the choice of the treatment be based on the severity of illness and associated organ dysfunctions. However, the treatment selection is based on subjective criteria. Several prognostic tools have been developed to assess disease severity and the risk of poor outcome, but none of them account for risk factors specific to patients who have undergone cardiac surgery, such as CPB duration and surgery type. Both general illness severity scores and kidney-specific disease severity scores do not identify these risk factors; therefore, such scoring systems are not entirely suitable for patients with AKI-D after cardiac surgery with CPB. Most known among the recent disease-specific models for AKI-D patients are the Cleveland Clinic Foundation score [12] and the score developed in the Veterans Affairs/National Institutes of Health (VA/NIH) trial [13]. The former was designed to predict the postoperative AKI-D, but not the outcome, while the latter was developed for the common population of AKI-D and does not account for the important, specific factor pertaining to cardiac surgery, i.e., the CPB duration. Assessment of the disease severity with consideration of the specific risk factors may be helpful for the choice of the RRT modality in patients with AKI-D after cardiac surgery. The development of a specific scoring model based on disease severity will facilitate the safe choice of an appropriate modality of RRT and, consequently, avoid the needless economic costs associated with the unnecessary prescription of an expensive treatment.
Patients and Methods Study Population We retrospectively analysed the data of 321 patients who developed AKI-D after open-heart surgery with cardiopulmonary bypass (CPB). Data were extracted from the patient records of surgeries performed at the Novosibirsk Research Institute of Circulation Pathology (NRICP) between September 2003 and July 2010. A total of 10,685 patients aged above 18 years underwent CPB during the time of the study. Of these, 314 (2.94%) patients developed AKI-D. All patients receiving RRT after surgery were assessed for inclusion in the study.
A.A. Malov et al.
Table 1 General Characteristics of the Patients. Data Are Presented as Numbers or Median (25th; 75th Percentile). N = 300
%
Gender Male
170
56.7
Female
130
43.3
Age, years
59 (51; 66)
Surgery type Valve surgery
150
50
CABGa Aortic arch surgery Combined surgery
49
16.3
39
13
62
20.7
Cardiopulmonary bypass
180 (120; 236)
duration, min a
Coronary artery bypass grafting.
Exclusion criteria were preoperative need for RRT, including those with chronic renal disease (CRD) requiring dialysis (n = 14) and those with non-renal indications of RRT (n = 7). The derivation dataset comprised data obtained retrospectively from 300 patients aged between 20 and 79 years. The patients had undergone coronary artery bypass graft surgery (CABG), heart valve (HV) surgery, aortic arch (AA) surgery, or combined procedures (Table 1). The RRT modality distribution showed that 124 patients (41.3%) received IHD and 176 (58.7%) received CRRT. In the case of 61 patients (20.3%), the RRT modality was switched during treatment. The mortality rates among the IHD patients and CRRT patients were 15.4% and 59.3%, respectively. Finally, 60 patients who developed AKI-D after undergoing open-heart surgery with CPB in NRICP between September 2011 and April 2012 were assessed prospectively, and the data of these patients was used to form the validation dataset. CRRT was performed using the Multifiltrate (Fresenius Medical Care) and Prisma and Prismaflex (Gambro Medical). The blood flow rate maintained during the procedures was 100–200 mL min1, and the dialysate flow rate was 1.5– 2.5 L h1. The CRRT duration varied widely from 18 h to 28 days (mean SD, 72 32 h). The effluent dose ranged from 30 to 100 mL kg1 min1. IHD therapy was carried out on stationary dialysis machines (models 4008B, 4008S, and 5008; Fresenius Medical Care). The treatment duration varied widely from 2.5 to 8 h (3.4 2.2 h). Therapies were generally performed with a dialysate flow of 6, 12, and 18 L h1, and sometimes, 30 L h1 and a blood flow rate of 100– 250 mL min1. The decision to begin dialysis was based on clinical judgement, and the indications included azotemia, volume overload, and biochemical abnormalities.
Predictive Model Building First, we estimated the predictive ability of five well-known severity scores for comparison with that of the newly
327
Mortality in Post-CPB AKI
developed model. The five severity scores were as follows: Acute Physiology and Chronic Health Evaluation (APACHE) II as a general score [14]; Multiple Organ Dysfunction (MODS) [15], Sequential Organ Failure Assessment (SOFA) [16], and Organ System Failure (OSF) [17], as organfailure scores; and Liano’s score as a kidney-specific disease severity score [18]. Preference was given to organ failure scores because AKI-D after cardiac surgery is usually a component of multiple organ failure. Liano’s score was chosen because among all the kidney-specific severity scores, this score has been externally validated most frequently [19]. All scores were calculated at the time of RRT initiation. We examined 18 variables as possible predictors of a fatal outcome: demographic data (age, gender); preoperative serum creatinine level, as an indicator of baseline renal function; surgical data (CPB duration, types of cardiac surgery, and RRT initiation day); and 12 postoperative variables (serum creatinine level, oliguria, number of sympathomimetic agents administered, mechanical lung ventilation [MLV], perioperative myocardial infarction [MI], oxygenation index [PO2/FiO2], bilirubinaemia, albuminaemia, platelet count, cerebral coma [Glasgow score of
