Skip to main content
NCBI home page
Medicine logoLink to Medicine
. 2019 May 24;98(21):e15789. doi: 10.1097/MD.0000000000015789

Development of model to predict end-stage renal disease after coronary artery bypass grafting

The ACHE score

Yeonhee Lee a, Jiwon Park b, Myoung-Jin Jang b, Hong Ran Moon a, Dong Ki Kim a, Kook-Hwan Oh a, Kwon Wook Joo a, Chun Soo Lim a,c, Yon Su Kim a, Ki Young Na a,d, Seung Seok Han a,∗
Editor: Alfonso H Santos
PMCID: PMC6571385  PMID: 31124973

Supplemental Digital Content is available in the text

Keywords: acute kidney injury, chronic kidney disease, coronary artery bypass grafting, endstage renal disease, hypertension

Abstract

Because end-stage renal disease (ESRD) increases the risks of morbidity and mortality, early detection and prevention of ESRD is a critical issue in clinical practice. However, no ESRD-prediction models have been developed or validated in patients undergoing coronary artery bypass grafting (CABG).

This is a retrospective multicenter cohort study, recruited between January 2004 and December 2015. A cohort of 3089 patients undergoing CABG in two tertiary referral centers was analyzed to derive a risk-prediction model. The model was developed using Cox proportional hazard analyses, and its performance was assessed using C-statistics. The model was externally validated in an independent cohort of 279 patients.

During the median follow-up of 6 years (maximum 13 years), ESRD occurred in 60 patients (2.0%). Through stepwise selection multivariate analyses, the following three variables were finally included in the ESRD-prediction model: postoperative Acute kidney injury, underlying Chronic kidney disease, and the number of antiHypertensive drugs (ACHE score). This model showed good performance in predicting ESRD with the following C-statistics: 0.89 (95% confidence interval [CI] 0.84–0.94) in the development cohort and 0.82 (95% CI 0.60–1.00) in the external validation cohort.

The present ESRD-prediction model may be applicable to patients undergoing CABG, with the advantage of simplicity and preciseness.

1. Introduction

End-stage renal disease (ESRD) increases the risks of morbidity and mortality.[1] ESRD is a contributor to high medical costs, poor quality of life, and several comorbidities including ischemic heart disease, stroke, infection, and frailty.[2–5] More than two million patients are being treated for ESRD worldwide, and the prevalence is expected to rise sharply in the next decade despite the fact that patients with ESRD are receiving increasingly more attention because of worsening outcomes.[1,6] When kidney function falls, kidney transplantation confers survival advantages over dialysis[7]; however, accessibility to transplantation is limited because of donor scarcity. Although dialysis successfully replaces kidney function, patients frequently encounter various complications and high mortality even under dialysis.[1,6] Therefore, early detection and prevention of ESRD are urgent issues.[1,8] However, there are no validated models with which to predict ESRD, although models such as the Cleveland Clinic score and the Society of Thoracic Surgery risk score had been proposed to predict acute kidney injury (AKI) after cardiac surgery.[9,10]

Coronary artery bypass grafting (CABG) is one of the most commonly performed major surgeries for the treatment of severe multivessel coronary artery disease.[11] Nevertheless, the overall outcomes of CABG are not perfect with regard to postsurgical morbidity and mortality.[12] This issue may be particularly problematic for patients with kidney dysfunction including AKI, chronic kidney disease (CKD), and ESRD.[12,13] Although AKI, CKD, and ESRD have relationships within a continuum of disease processes, ESRD is more closely related to mortality than are the others.[14] If patients undergoing CABG develop postoperative ESRD, the mortality risk increases to 25%.[15] Possible explanations for the higher mortality among patients with ESRD include accelerated atherosclerosis, anemia, cardiomyopathy, and repeated events of ischemic heart disease.[16] Therefore, it is essential to predict the risk of ESRD after CABG. To the best of our knowledge, no ESRD-prediction models have been developed or evaluated in patients undergoing CABG. This study was performed to derive the most effective and clinically applicable model and validate it in an independent cohort.

2. Materials and methods

2.1. Patients and study design

The study design was approved by the institutional review boards of all involved centers (nos. H-1702–050–831, B-1702/384–103, and 20170531/10–2017–1/071) and complied with the Declaration of Helsinki. The study was performed in two consecutive parts. In part 1, we conducted a retrospective multicenter cohort study involving 3089 patients who comprised the development and internal validation cohort. Patients who had undergone CABG in two tertiary referral centers (Seoul National University Hospital and Seoul National University Bundang Hospital) were recruited from January 2004 to December 2015. We excluded patients if they had undergone renal replacement therapy before surgery or had ESRD (n = 63), had undergone concomitant valve surgery or redo-CABG (n = 10), or were ≤18 years of age (n = 1). Finally, the cohort comprised of 3015 patients.

Baseline clinical preoperative, intraoperative, and postoperative data were recorded, including age; sex; body mass index; systolic and diastolic blood pressures; current smoking status; hypertension; diabetes mellitus; history of myocardial infarction, stroke, or peripheral vascular disease; medications such as angiotensin-converting enzyme inhibitors, angiotensin receptor blockers, beta-blockers, calcium channel blockers, diuretics, and others; contrast use within 1 week before surgery; perioperative use of an intra-aortic balloon pump or cardiopulmonary bypass; and total surgery time. The number of antihypertensive drugs used was categorized into 0, 1–2, and ≥3. The left ventricular ejection fraction was determined by Simpson's modified biplane method from the apical two- and four-chamber views on echocardiography. Laboratory data, such as the serum creatinine, albumin, cholesterol, hemoglobin, and white blood cell count, were obtained. The estimated glomerular filtration rate (eGFR) was calculated using the Chronic Kidney Disease Epidemiology Collaboration Eq.[17] AKI was defined as an increase in the serum creatinine level to ≥0.3 mg/dL or ≥1.5 times baseline within 48 h after surgery and was classified according to the guideline proposed by Kidney Disease Improving Global Outcomes[18] as follows: stage 1, an increase in the serum creatinine level to ≥0.3 mg/dL or 1.5–1.9 times baseline; stage 2, an increase in the serum creatinine level of 2.0–2.9 times baseline; and stage 3, an increase in the serum creatinine level to 4.0 mg/dL or 3.0 times baseline. The postoperative serum creatinine level recorded was the highest level measured within a 48-hour timeframe. CKD was defined as an eGFR of <60 mL/min/1.73 m2 for 3 months and classified as G3a (45–60 mL/min/1.73 m2), G3b (30–45 mL/min/1.73 m2), or G4–5 (<30 mL/min/1.73 m2) according to the guideline.[19]

We developed a prediction model using the above cohort and validated its performance in both the internal cohort and separate dataset (part 2), wherein a total of 279 patients undergoing CABG were recruited from an independent tertiary referral center (Boramae Medical Center). Patient selection flow charts for the development (internal) and external cohorts are shown in Fig. S1, Supplemental Content).

2.2. Outcome variables

The primary outcome was ESRD. The onset of ESRD was defined as the initiation of renal replacement therapy or kidney transplantation due to failed kidney function after CABG. Patients were followed until April 2017 or death. The information on ESRD was obtained from the Korean Renal Registry, which is a database of all patients who have undergone renal replacement therapy in Korea.[20]

The secondary outcome was all-cause mortality. The data on death were obtained from the National Database of Statistics Korea.

2.3. Statistical analysis

The data are described as mean with standard deviation for continuous variables and as proportion for categorical variables. We compared the baseline characteristics using chi-square test, one-way ANOVA, and Fisher's exact test, as appropriate. Kaplan–Meier curves were drawn to compare the risks between groups. A Cox proportional hazard ratio (HR) model was used to estimate the HR of ESRD risk. Restricted cubic splines were used to check the assumption of proportional linearity between variables and their log hazards. Variables with a P value of <.2 in the univariate model were applied to multivariate stepwise analyses. To facilitate the calculation of risks, we converted estimates of log HRs (coefficients) to integer scores, which were calculated as (rounding of the coefficients × 100)/(sum of the largest coefficient in each predictor) to obtain a total score of 100. The performance of the model was assessed with respect to calibration and discrimination using development (internal) and external validation cohorts. Discrimination ability was evaluated with Harrell's C-statistics from the jackknife method in the somersd package of STATA (StataCorp, College Station, TX, USA).[21,22] Calibration was assessed by plotting the predicted 5-year risk of ESRD against the observed risk with 95% confidence intervals (CIs) for the quartiles of the predicted value.

As a sensitivity analysis, we have applied a formal competing-risks (Fine-Gray) proportional HR regression model, which estimated the cumulative incidence of an ESRD or death before ESRD by considering the sub-distribution hazard. All tests were two-sided and performed at the .05 significance level. The analyses were conducted with the statistical software packages (SPSS [version 22; IBM Corp., Armonk, NY, USA], SAS [version 9.3; SAS Institute, Cary, NC, USA], and STATA [version 12]).

3. Results

3.1. Baseline characteristics

All baseline characteristics are presented in Table 1. Of 3015 patients, 2222 (73.7%) were male and 1322 (43.8%) had diabetes mellitus. The mean baseline eGFR was 70.1 ± 20.17 mL/min/1.73 m2. Postoperative AKI occurred in 798 patients (26.5%), including stage 1 in 23.8% and stages 2 and 3 in 2.7%. Preoperative CKD was identified in 890 patients (29.5%). Among them, 303 patients developed postoperative AKI. The median follow-up duration was 6.1 years (interquartile range, 2.9–9.2 years; maximum, 13.3 years).

Table 1.

Baseline characteristics of patients according to the presence of acute kidney injury and chronic kidney disease.

3.1.

Open in a new tab

3.2. Risk factors for ESRD

ESRD occurred in 60 patients (2.0%) and 5 patients (1.8%) in the development cohort and the external validation cohort, respectively. Estimation of the ESRD risk of each variable showed that the following were significantly associated with the risk: diabetes mellitus, a history of cardiovascular disease (sum of myocardial infarction, stroke, and peripheral vascular disease), postoperative AKI, underlying CKD, the number of antihypertensive drugs, the use of cardiopulmonary bypass, total surgery time, hemoglobin level, albumin level, and left ventricular ejection fraction (Table 2). After stepwise selection, the following three variables were significant: postoperative AKI, underlying CKD, and the number of antihypertensive drugs.

Table 2.

Cox proportional hazards regression results for the risk of end-stage renal disease.

3.2.

Open in a new tab

Next, we re-analyzed the ESRD risk according to the three above-mentioned factors. The proportions of ESRD were as follows: non-AKI/non-CKD group, 0.4%; AKI/non-CKD group, 1.2%; non-AKI/CKD group, 3.4%; and AKI/CKD group, 9.2% (Table 3). The incidence rates of ESRD were 0.6 per 1000 person-years in the non-AKI/non-CKD group, 1.7 in the AKI/non-CKD group, 5.8 in the non-AKI/CKD group, and 19.4 in the AKI/CKD group. Figure 1 shows the cumulative risk curves of ESRD. As shown in Table 3, both AKI (hazard ratio [HR] 3.2, 95% CI 1.01–9.80) and CKD (HR 13.8, 95% CI 5.36–35.44) were associated with the risk of ESRD. Particularly, the extent to which AKI increased the risk of ESRD was greater in the CKD group (HR 3.4, 95% CI 1.91–6.04) than in the non-CKD group (HR 2.9, 95% CI 0.95–9.16).

Table 3.

Risk of end-stage renal disease according to the presence of acute kidney injury and chronic kidney disease.

3.2.

Open in a new tab

Figure 1.

Figure 1

Open in a new tab

Kaplan–Meier curves for the cumulative probability of end-stage renal disease according to the presence of acute kidney injury and chronic kidney disease. ∗P < .001 compared with the non-AKI/non-CKD group by the log-rank test.

Figure 2 shows the risk curves of ESRD according to the stages of AKI or CKD. The ESRD risk increased with advancing stages, especially in stage 3 AKI (HR 6.59, 95% CI 2.80–15.52) and grades 4 to 5 CKD (HR 54.2, 95% CI 26.21–112.22).

Figure 2.

Figure 2

Open in a new tab

Kaplan–Meier curves for the cumulative probability of end-stage renal disease according to the classification of (A) acute kidney injury and (B) chronic kidney disease. ∗P < .05; †P < .01; ‡P < .001 compared with the non-AKI or non-CKD group by the log-rank test.

The risk of ESRD was higher in patients using ≥3 of antihypertensive drugs (HR 2.7, 95% CI 1.34–5.62) than in those using no antihypertensive drugs. The risks were similar between patients using no antihypertensive drugs and 1–2 of antihypertensive drugs (HR 1.3, 95% CI 0.66–2.54) (Fig. 3).

Figure 3.

Figure 3

Open in a new tab

Kaplan–Meier curves for the cumulative probability of end-stage renal disease according to the number of antihypertensive drugs. ∗P < 0.001 compared with the group without the use of antihypertensive drugs by the log-rank test.

3.3. Development of prediction model

Collectively, three variables (AKI, CKD, and the number of antihypertensive drugs) were included in the 5-year ESRD-prediction model. The Cox model provided the estimated probability of failure for the 5-year prediction of ESRD. This probability was equal to P = 1 − .9978Exp(X) where X = 0.7149 × (1; stage 1 AKI) + 1.2892 × (1; stage 2 AKI) + 1.8858 × (1; stage 3 AKI) + 1.0115 × (1; grade 3a CKD) + 2.3460 × (1; grade 3b CKD) + 3.9932 × (1; grade 4–5 CKD) + 0.2602 × (1; 1–2 antihypertensive drugs) + 1.0095 × (1; ≥ 3 antihypertensive drugs). Based on this model, a nomogram was developed using three variables to allow the clinician to easily estimate the 5-year risk of ESRD (see Fig. S2, Supplemental Content, nomogram). The original model was assessed in the development (internal) and external validation cohorts by evaluating the discrimination ability using C-statistics (see Table S1, Supplemental Content, which shows regression coefficient and integer-based simplified score).

We developed an integer-based simplified scoring index (ACHE score; postoperative Acute kidney injury, underlying Chronic kidney disease, and the number of antiHypertensive drugs in the ESRD-prediction model) to easily apply this ESRD-prediction model in clinical practice (Fig. 4). Points assigned to the values of each variable can be summed to obtain a patient's total score, which can be used to assess the predicted risk of ESRD. The C-statistics of simplified scoring model were 0.89 (95% CI 0.84–0.94) and 0.82 (95% CI 0.60–1.00) for the development and external validation cohorts, respectively. Regarding the calibration plots (Fig. 5), the predicted probability appeared to be almost consistent with the actual probability of ESRD, indicating good calibration. When patients were categorized by the quartile of scores, the risk curves of ESRD were well separated (see Fig. S3, Supplemental Content, which represents the risk of ESRD in the internal validation cohort, based on the quartile of scores).

Figure 4.

Figure 4

Open in a new tab

Simplified scoring index of the developed end-stage renal disease-prediction model (ACHE score).

Figure 5.

Figure 5

Open in a new tab

Calibration plots in the (A) internal and (B) external validation cohorts.

3.4. Composite risk of ESRD and all-cause mortality

During the follow-up period, 785 patients (26.0%) died. When we defined the composite outcome as the total events of ESRD and death, this outcome occurred in 818 patients (27.1%), including 246 (15.1%) in the non-AKI/non-CKD group, 155 (31.3%) in the AKI/non-CKD group, 233 (38.0%) in the non-AKI/CKD group, and 194 (64.0%) in the AKI/CKD group (see Table S2, Supplemental Content, which shows the risk of composite outcome by AKI and CKD). The risk of the composite outcome increased depending on the presence of AKI and CKD (see Fig. S4, Supplemental Content, which represents the risk of composite outcome according to AKI and CKD), similar to the results of Fig. 1. The HR was 1.53 (95% CI 1.24–1.89) in the AKI/non-CKD group, 1.73 (95% CI 1.42–2.10) in the non-AKI/CKD group, and 3.01 (95% CI 2.44–3.71) in the AKI/CKD group, compared with the non-AKI/non-CKD group. The composite risk of ESRD and death was not associated with the number of antihypertensive drugs. Although the death events were considered as a competing risk, all of AKI, CKD, and the number of antihypertensive drugs predicted the risk of ESRD with a similar pattern to previous analyses (see Table S3, Supplemental Content, which shows the risk of ESRD from sub-distribution hazard models with death as a competing risk).

4. Discussion

Prediction models for ESRD have been evaluated in patients with various pathologic conditions, such as CKD,[23–26] light chain amyloidosis,[27] and diabetes.[28–31] Nevertheless, an ESRD-prediction model after CABG has never been developed. The Cleveland Clinic score and the Society of Thoracic Surgery risk score have a limitation in predicting a risk of ESRD over several years.[9,10] Despite the difficulty in deriving such a model because of the relatively high mortality during the perioperative period, we successfully created an ESRD-prediction model with a long-term follow-up. This finding could be of significance in the perioperative management of patients undergoing other type of cardiac surgery. In outpatient setting, clinicians can focus on appropriate care of kidney disease progression in patients undergoing CABG who has AKI, CKD, or high number of anti-hypertensive drugs.

Regarding the best predictor, kidney dysfunction itself (i.e., AKI and CKD) predominantly determined the risk of ESRD, although other factors such as comorbidities and intraoperative problems could be also considered in each case. Furthermore, the number of antihypertensive drugs, representing the severity of hypertension, was chosen in the final model. The estimation of ESRD risk could be simplified by these kidney function and antihypertensive drug data with good performance for the internal and external validations. Furthermore, the predictability of these factors was confirmed despite taking the competing risk for death into account.

AKI and CKD are risk factors for each other. AKI has long been regarded as completely reversible, but recent studies have shown that AKI may cause permanent kidney injury (inducing renal fibrosis) and increase the risk of long-term subsequent progression of kidney dysfunction because of incomplete recovery.[32–34] Patients with CKD may be at risk of AKI due to autoregulation failure, abnormal vasodilatation, and adverse effects of medications.[33] Throughout this series of processes, the loss of renal reserve gradually progresses, and kidney function eventually fails until ESRD develops. This pathophysiologic issue may underlie the primary selection of AKI and CKD during development of the present prediction model.

Hypertension is a well-known risk factor for the progression from CKD to ESRD.[35] Various parameters may reflect the severity of hypertension. Among them, the number of antihypertensive drugs was selected as the best predictive factor in the present model. The advantage of this variable over other hypertension-representative factors, including measured blood pressure, is that it is not temporary. For example, the blood pressure might be unstable during the perioperative period or due to the white coat effect.[36–38] In this regard, 24-h blood pressure monitoring may be the most effective way to reflect the patient's hypertensive condition. However, because this perfect measurement technique is not always feasible, particularly in emergency or surgical settings, the number of antihypertensive drugs may be clinically useful to elucidate the patient's hypertensive condition.

The strengths of our study are that the model was developed from relatively large multicenter cohorts of adults undergoing CABG and that the primary outcome (ESRD) was well documented. The total period was 13 years, and this enabled us to develop the ESRD-prediction model. The predictor variables in the model were readily available patient demographics and laboratory test parameters. The model was finally validated in an independent cohort. Nevertheless, we acknowledge the limitations inherent to a retrospective study design. Some patients died before ESRD events, although these were appropriately censored. This was reflected by the fact that the number of deaths was 10 times higher than the number of ESRD events. As a sensitivity analysis, we further applied a formal competing-risks (Fine-Gray) model. The difference in the risk of ESRD among various baseline diseases is another issue. The present cohort included only patients undergoing CABG; thus, application of the model to other disease subsets such as patients undergoing non-cardiac surgery or nonsurgical treatment may be limited. Another issue is that external validation was conducted in a single-center cohort of 279 patients, among whom a relatively small number developed ESRD. This resulted in wide CIs in the C-statistics, although the calibration plots showed good performance. Finally, the information on the cause of AKI was not available, which could strengthen the understanding of the relationships.

5. Conclusions

Because the outcomes of CABG are inseparable from ESRD, the present study was performed to develop an ESRD-prediction model following CABG, with the clear advantage of simplicity and preciseness of prediction. Such risk estimation can lead to prognostic assessment and follow-up of each patient, helping to guide appropriate treatment by the nephrologists as well as other clinicians to identify the risky patients. Forthcoming studies will validate the present model in other types of cardiac surgery in addition to CABG.

Acknowledgment

This work was supported by a grant from the Basic Science Research Program through the National Research Foundation of Korea (NRF) funded by the Ministry of Education (NRF-2017R1D1A1B03031642).

Author contributions

YL designed the study, collected the data, analyzed and interpreted the results, and drafted the manuscript. JP and MJJ analyzed and interpreted the results. HRM, DKK, and KHO collected the data. CSL and YSK conceived the study and assisted in the analyses. KYN conceived the study, analyzed the results, and interpreted the data. SSH conceived the study, analyzed the results, interpreted the data, and reviewed the manuscript. All of the authors read and approved the final manuscript.

Conceptualization: Dong Ki Kim, Kwon Wook Joo, Seung Seok Han.

Data curation: Hong Ran Moon, Chun Soo Lim, Ki Young Na, Seung Seok Han.

Formal analysis: Yeonhee Lee, Jiwon Park, Myoung-Jin Jang, Seung Seok Han.

Resources: Yon Su Kim.

Supervision: Kook-Hwan Oh.

Writing – original draft: Yeonhee Lee, Seung Seok Han.

Writing – review & editing: Yeonhee Lee, Seung Seok Han.

Seung Seok Han orcid: 0000-0003-0137-5261.

Supplementary Material

Supplemental Digital Content

Footnotes

Abbreviations: AKI = acute kidney injury, CABG = coronary artery bypass grafting, CI = confidence interval, CKD = chronic kidney disease, eGFR = estimated glomerular filtration rate, ESRD = end-stage renal disease, HR = hazard ratio.

The authors report no conflicts of interest.

Supplemental Digital Content is available for this article.

References

  • [1].Liyanage T, Ninomiya T, Jha V, et al. Worldwide access to treatment for end-stage kidney disease: a systematic review. Lancet 2015;385:1975–82. [DOI] [PubMed] [Google Scholar]
  • [2].Bello AK, Levin A, Tonelli M, et al. Assessment of global kidney health care status. JAMA 2017;317:1864–81. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [3].Go AS, Chertow GM, Fan D, et al. Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization. N Engl J Med 2004;351:1296–305. [DOI] [PubMed] [Google Scholar]
  • [4].Wang HE, Gamboa C, Warnock DG, et al. Chronic kidney disease and risk of death from infection. Am J Nephrol 2011;34:330–6. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [5].Wilhelm-Leen ER, Hall YN, K Tamura M. Frailty and chronic kidney disease: the Third National Health and Nutrition Evaluation Survey. Am J Med 2009;122:664–71.e2. e662. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [6].Robinson BM, Akizawa T, Jager KJ, et al. Factors affecting outcomes in patients reaching end-stage kidney disease worldwide: differences in access to renal replacement therapy, modality use, and haemodialysis practices. Lancet 2016;388:294–306. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [7].Schnuelle P, Lorenz D, Trede M, et al. Impact of renal cadaveric transplantation on survival in end-stage renal failure: evidence for reduced mortality risk compared with hemodialysis during long-term follow-up. J Am Soc Nephrol 1998;9:2135–41. [DOI] [PubMed] [Google Scholar]
  • [8].Jha V, Garcia-Garcia G, Iseki K, et al. Chronic kidney disease: global dimension and perspectives. Lancet 2013;382:260–72. [DOI] [PubMed] [Google Scholar]
  • [9].Thakar CV, Arrigain S, Worley S, et al. A clinical score to predict acute renal failure after cardiac surgery. J Am Soc Nephrol 2005;16:162–8. [DOI] [PubMed] [Google Scholar]
  • [10].Mehta RH, Grab JD, O’Brien SM, et al. Bedside tool for predicting the risk of postoperative dialysis in patients undergoing cardiac surgery. Circulation 2006;114:2208–16. quiz 2208. [DOI] [PubMed] [Google Scholar]
  • [11].Alexander JH, Smith PK. Coronary-artery bypass grafting. N Engl J Med 2016;374:1954–64. [DOI] [PubMed] [Google Scholar]
  • [12].Parikh DS, Swaminathan M, Archer LE, et al. Perioperative outcomes among patients with end-stage renal disease following coronary artery bypass surgery in the USA. Nephrol Dial Transplant 2010;25:2275–83. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [13].Han SS, Shin N, Baek SH, et al. Effects of acute kidney injury and chronic kidney disease on long-term mortality after coronary artery bypass grafting. Am Heart J 2015;169:419–25. [DOI] [PubMed] [Google Scholar]
  • [14].Chawla LS, Bellomo R, Bihorac A, et al. Acute kidney disease and renal recovery: consensus report of the Acute Disease Quality Initiative (ADQI) 16 Workgroup. Nat Rev Nephrol 2017;13:241–57. [DOI] [PubMed] [Google Scholar]
  • [15].Ivert T, Holzmann MJ, Sartipy U. Survival in patients with acute kidney injury requiring dialysis after coronary artery bypass grafting. Eur J Cardiothorac Surg 2014;45:312–7. [DOI] [PubMed] [Google Scholar]
  • [16].Holzmann MJ, Hammar N, Ahnve S, et al. Renal insufficiency and long-term mortality and incidence of myocardial infarction in patients undergoing coronary artery bypass grafting. Eur Heart J 2007;28:865–71. [DOI] [PubMed] [Google Scholar]
  • [17].Levey AS, Stevens LA, Schmid CH, et al. A new equation to estimate glomerular filtration rate. Ann Intern Med 2009;150:604–12. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [18].KDIGO AKI Working Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int Suppl 2012;2:1–38. [Google Scholar]
  • [19].KDIGO Working Group. KDIGO 2012 clinical practice guideline for evaluation and management of chronic kidney disease. Kidney Int Suppl 2013;3:1–63. [DOI] [PubMed] [Google Scholar]
  • [20].ESRD Registry Committee KSN. Current renal replacement therapy in Korea; 2016. Available at http://www.ksn.or.kr. Accessed at April 19, 2017. [Google Scholar]
  • [21].Harrell FE, Jr, Lee KL, Mark DB. Multivariable prognostic models: issues in developing models, evaluating assumptions and adequacy, and measuring and reducing errors. Stat Med 1996;15:361–87. [DOI] [PubMed] [Google Scholar]
  • [22].Wolbers M, Blanche P, Koller MT, et al. Concordance for prognostic models with competing risks. Biostatistics 2014;15:526–39. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [23].Johnson ES, Thorp ML, Platt RW, et al. Predicting the risk of dialysis and transplant among patients with CKD: a retrospective cohort study. Am J Kidney Dis 2008;52:653–60. [DOI] [PubMed] [Google Scholar]
  • [24].Landray MJ, Emberson JR, Blackwell L, et al. Prediction of ESRD and death among people with CKD: the Chronic Renal Impairment in Birmingham (CRIB) prospective cohort study. Am J Kidney Dis 2010;56:1082–94. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [25].Tangri N, Stevens LA, Griffith J, et al. A predictive model for progression of chronic kidney disease to kidney failure. JAMA 2011;305:1553–9. [DOI] [PubMed] [Google Scholar]
  • [26].Drawz PE, Goswami P, Azem R, et al. A simple tool to predict end-stage renal disease within 1 year in elderly adults with advanced chronic kidney disease. J Am Geriatr Soc 2013;61:762–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [27].Rubinstein S, Cornell RF, Du L, et al. Novel pathologic scoring tools predict end-stage kidney disease in light chain (AL) amyloidosis. Amyloid 2017;24:205–11. [DOI] [PubMed] [Google Scholar]
  • [28].Desai AS, Toto R, Jarolim P, et al. Association between cardiac biomarkers and the development of ESRD in patients with type 2 diabetes mellitus, anemia, and CKD. Am J Kidney Dis 2011;58:717–28. [DOI] [PubMed] [Google Scholar]
  • [29].Jardine MJ, Hata J, Woodward M, et al. Prediction of kidney-related outcomes in patients with type 2 diabetes. Am J Kidney Dis 2012;60:770–8. [DOI] [PubMed] [Google Scholar]
  • [30].Lin CC, Li CI, Liu CS, et al. Development and validation of a risk prediction model for end-stage renal disease in patients with type 2 diabetes. Sci Rep 2017;7:10177. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [31].Wan EYF, Fong DYT, Fung CSC, et al. Prediction of new onset of end stage renal disease in Chinese patients with type 2 diabetes mellitus—a population-based retrospective cohort study. BMC Nephrol 2017;18:257. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [32].Coca SG, Singanamala S, Parikh CR. Chronic kidney disease after acute kidney injury: a systematic review and meta-analysis. Kidney Int 2012;81:442–8. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [33].Chawla LS, Eggers PW, Star RA, et al. Acute kidney injury and chronic kidney disease as interconnected syndromes. N Engl J Med 2014;371:58–66. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [34].Sawhney S, Marks A, Fluck N, et al. Post-discharge kidney function is associated with subsequent ten-year renal progression risk among survivors of acute kidney injury. Kidney Int 2017;92:440–52. [DOI] [PMC free article] [PubMed] [Google Scholar]
  • [35].Gansevoort RT, Correa-Rotter R, Hemmelgarn BR, et al. Chronic kidney disease and cardiovascular risk: epidemiology, mechanisms, and prevention. Lancet 2013;382:339–52. [DOI] [PubMed] [Google Scholar]
  • [36].O’Brien E, Staessen JA. What is “hypertension”? Lancet 1999;353:1541–3. [DOI] [PubMed] [Google Scholar]
  • [37].Verdecchia P, Staessen JA, White WB, et al. Properly defining white coat hypertension. Eur Heart J 2002;23:106–9. [DOI] [PubMed] [Google Scholar]
  • [38].Sweitzer BJ, Howell SJ. The Goldilocks principle as it applies to perioperative blood pressure: what is too high, too low, or just right? Br J Anaesth 2017;119:7–10. [DOI] [PubMed] [Google Scholar]

Associated Data

This section collects any data citations, data availability statements, or supplementary materials included in this article.

Supplementary Materials

Supplemental Digital Content
medi-98-e15789-s001.doc (672KB, doc)

Articles from Medicine are provided here courtesy of Wolters Kluwer Health

RESOURCES