Background
Although small changes in creatinine level during hospitalization have been associated with risk of short-term mortality, associations with posthospitalization end-stage renal disease (ESRD) and long-term mortality are unknown. We assessed the relationship between change in serum creatinine levels up to 3.0 mg/dL and death and ESRD among elderly survivors of hospitalization for acute myocardial infarction.
Methods
Retrospective cohort study of a nationally representative sample of Medicare beneficiaries admitted with acute myocardial infarction to nonfederal US hospitals between February 1994 and July 1995. Outcomes were mortality and ESRD through June 2004.
Results
The 87 094 eligible patients admitted to 4473 hospitals had a mean age of 77.1 years; for the 43.2% with some creatinine increase, quartiles of increase were 0.1, 0.2, 0.3 to 0.5, and 0.6 to 3.0 mg/dL. Incidence of ESRD and mortality ranged from 2.3 and 139.1 cases per 1000 person-years, respectively, among patients with no increase to 20.0 and 274.9 cases per 1000 person-years in the highest quartile of creatinine increase. Compared with patients without creatinine increase, adjusted hazard ratios by quartile of increase were 1.45, 1.97, 2.36, and 3.26 for ESRD and 1.14, 1.16, 1.26, and 1.39 for mortality, with no 95% confidence intervals overlapping 1.0 for either end point.
Conclusion
In a nationally representative sample of elderly patients discharged after hospitalization for acute myocardial infarction, small changes in serum creatinine level during hospitalization were associated with an independent higher risk of ESRD and death.
Acute kidney injury (AKI), defined by an acute increase in baseline serum creatinine level,1 is a common complication of hospitalization, occurring in 5% to 7% of patients.2,3 It is becoming more frequent within the Medicare population, having risen in incidence by 11% per year since 1992.4 This is of concern because AKI is associated with a mortality rate of 30% to greater than 50% among patients with AKI severe enough to require dialysis during hospitalization.3-5
Recent studies6,7 demonstrate that relatively small increases in serum creatinine level during hospitalization, even among patients not requiring dialysis, are associated with adverse outcomes, and that the incidence of short-term mortality is related to the magnitude of the acute change in serum creatinine level. Among patients undergoing coronary angiography, for example, changes in serum creatinine level as small as +0.25 mg/dL (to convert to micromoles per liter, multiply by 88.4) were associated with higher in-hospital mortality.6 A single-center analysis also demonstrated that a rise in serum creatinine level of 0.3 mg/dL or more during hospitalization was associated with higher in-hospital mortality, longer stay, and higher costs.7 However, despite this higher mortality during hospitalization, most patients with reductions in kidney function survive to discharge. Little is known about the long-term risks of subsequent end-stage renal disease (ESRD) and mortality among elderly patients who experience an increase in serum creatinine level during hospitalization.
The goal of this analysis was to assess ESRD and mortality risk associated with changes in serum creatinine level among survivors of a hospitalization for acute myocardial infarction (AMI) within a nationally representative cohort of Medicare beneficiaries.
The Cooperative Cardiovascular Project (CCP) was a national quality-improvement project to improve the care of patients hospitalized with AMI.8,9 The original CCP data set contains 234 754 randomly selected hospital admissions for AMI for 210 981 Medicare beneficiaries from all 50 states hospitalized between February 1994 and July 1995.10 Patients with an International Classification of Diseases, Ninth Revision, Clinical Modification11 principal discharge diagnosis code of 410 (AMI) were sampled from 6684 nonfederal hospitals, composing virtually all acute care hospitals in the United States. Trained abstractors recorded hospital chart data, with extensive data quality monitoring.10 Patients from the CCP cohort were linked to the US Renal Data System by means of health insurance claim number and beneficiary identification code to identify ESRD. Valid personal identifier numbers (health insurance claim/beneficiary identification code) were available for 94.7% of the CCP patients. Matches were subsequently individually verified on the basis of a 5-point scoring system designed to test validity by using age, sex, and US Renal Data System–verified date of death. Date of death for patients without ESRD was obtained from the Medicare Enrollment Database. Follow-up for ESRD and mortality included the period from the inception of the CCP data collection (February 1994) through June 2004. Our study was approved by the institutional review board of the University of Alabama at Birmingham.
We based much of our exclusion criteria on previous studies that used the CCP data.12-14 We included the first hospitalization for the 210 981 patients in the CCP, and we excluded patients who reached the primary or censoring outcome during hospitalization, including those with a first date of long-term renal replacement therapy during hospitalization (n = 523) and death before hospital discharge (n = 31 408). We further excluded patients for the following reasons: a diagnosis of ESRD before hospital admission (n = 3106), AMI not confirmed by clinical criteria (n = 26 249), transfer to the index hospital (n = 29 712), transfer from the index hospital within 24 hours of admission (n = 34 759), race other than African American or white (n = 8178), and age less than 65 years (n = 15 552). Because our focus was the risk associated with change in kidney function among patients not requiring dialysis during hospitalization, we excluded patients who received acute hemodialysis during hospitalization as determined by the International Classification of Diseases, Ninth Revision, Clinical Modification procedure code 3995, shown to have high predictive validity for AKI requiring dialysis,15 as well as patients in the 99th percentile of increase in serum creatinine level during hospitalization (n = 4176). In addition, we excluded patients with missing or invalid data, as follows: admission or peak serum creatinine level missing or less than 0.1 mg/dL (n = 10 057), missing admission hematocrit (n = 10 115), and inability to identify outcome due to invalid health insurance claim/beneficiary identification code (n = 11 257). Owing to overlap, the total number of patients with any missing data was 24 955. We excluded patients who were not African American or white for 2 main reasons: race/ethnicity is an important confounding variable, and races/ethnicities other than African American or white cannot be classified with reasonable reliability from Medicare administrative data sets.16,17
We used the first serum creatinine level collected within 24 hours of hospital admission as the baseline serum creatinine level at the time of hospitalization. Estimated glomerular filtration rate, calculated by means of the abbreviated Modification of Diet in Renal Disease Study equation,18,19 was used as a measure of kidney function at admission. We divided patients into the following glomerular filtration rate categories by means of a modification of the National Kidney Foundation's Kidney Disease Outcomes Quality Initiative stages of kidney disease20,21: 60 or more, 45 to 59, 30 to 44, 15 to 29, and less than 15 mL/min/1.73 m2. To calculate the maximum change in serum creatinine level, we subtracted the starting serum creatinine level from the peak serum creatinine level measured during hospitalization. We classified patients into 5 mutually exclusive categories based on change in serum creatinine level during hospitalization as follows: a decrease or no change in serum creatinine level during admission, or an increase in serum creatinine level of 0.1, 0.2, 0.3 to 0.5, or 0.6 to 3.0 mg/dL. The cutoff points for increase in creatinine level were selected to approximate the quartiles of creatinine change among patients who experienced an increase in serum creatinine level.
We defined anemia as a hematocrit of less than 39% in men and less than 36% in women.22 We created 4 mutually exclusive levels of hospital technology mix as described previously.12
We compared patient baseline characteristics across the levels of serum creatinine change with the χ2 trend statistic for dichotomous variables and analysis of variance for continuous variables.23 We used Spearman rank correlation coefficients to test for a trend of higher death and ESRD incidence rates among patients with greater levels of serum creatinine change. To assess the association of change in creatinine with ESRD or mortality, we performed time-to-event analysis with Cox proportional hazards models using serum creatinine change as the primary independent variable and ESRD or mortality as the dependent variable. With ESRD as the end point, patients not developing ESRD were censored at the time of death or the end of follow-up. With mortality as the end point, patients were censored at the time of ESRD or the end of follow-up. In both multivariate regression models, initial adjustment was made for identical risk factors to include common risk factors for competing risks,24 including baseline patient characteristics, in-hospital and discharge medical treatment, and hospital characteristics; only covariates that retained statistical significance at the P < .05 level were kept in the final models. For regression models, we used a robust sandwich estimate to account for clustering within hospitals. We tested for violations of the proportional hazards assumption and accounted for violations by means of time-dependent covariates25 and tested for interactions between change in serum creatinine level and demographic characteristics, comorbidities, kidney function, and receipt of percutaneous coronary interventions or coronary artery bypass graft (CABG). In the regression model for the outcome of ESRD, there were no violations of the proportionality of hazards assumption, and, in the regression model for the outcome of mortality, there was a violation of the proportionality assumption for age, which we accounted for by using a time-dependent covariate. To assess the influence of missing data on our results, we imputed values for missing data elements,26 and the results demonstrated no change in statistical significance or substantive change in point estimates of risk for either end point.
Among the 87 094 patients in the analytic sample discharged from 4473 hospitals, the mean (SD) change in serum creatinine level was 0.16 (0.35) mg/dL, and 49 502 patients (56.8%) experienced a decrease or no change in serum creatinine level during hospitalization. For the 37 592 patients (43.2%) who experienced an increase in serum creatinine level during hospitalization, the distribution of creatinine increase is shown in Figure 1.
Mean (SD) age of the sample was 77.1 (7.5) years. The mean (SD) age among patients who experienced a decrease or no change in serum creatinine level was 76.8 (7.5) years (Table 1) and rose to 78.4 (7.4) years among patients with the greatest change in creatinine level. The proportion of African Americans and patients with diabetes mellitus, hypertension, previous myocardial infarction, history of congestive heart failure, and history of stroke was greatest among patients with larger changes in creatinine level. Patients with the greatest change in serum creatinine level also had worse kidney function at baseline, as reflected by a higher admission serum creatinine level and lower estimated glomerular filtration rate; greater changes in serum creatinine level were also associated with a lower admission hematocrit. Serum creatinine change was associated with more frequent CABG procedures during hospitalization and transfers to the intensive care unit, as well as less frequent percutaneous coronary interventions and thrombolytic therapy. Patients with increases in serum creatinine level were less likely to receive a β-blocker or aspirin at discharge, but more likely to receive an angiotensin-converting enzyme inhibitor and to be treated at hospitals with a higher technology index.
Median follow-up was 4.1 years (interquartile range, 0.6-5.2 years), amounting to 199 263.1 person-years. After 10 years of follow-up, 1.8% and 73.3% of patients developed ESRD and died, respectively, and the incidence rates of ESRD and death were 3.8 and 154.7 cases per 1000 person-years, respectively (Table 2). The incidences of ESRD and mortality were greatest among patients with larger changes in creatinine level; for example, patients with a creatinine increase of 0.6 to 3.0 mg/dL experienced an ESRD and mortality incidence of 20.0 and 274.9 cases per 1000 person-years, respectively.
After adjustment for demographic characteristics, comorbid medical conditions (history of stroke, hypertension, diabetes mellitus, previous myocardial infarction or CABG, and previous or current smoking), and admission characteristics (presence of reduced kidney function and presence of anemia), all levels of serum creatinine increase were independently and significantly associated with a greater ESRD risk, with the greatest risk being among patients with more extreme creatinine changes (Figure 2A and Figure 3A). The adjusted risk of ESRD associated with a creatinine increase of 0.6 to 3.0 mg/dL (vs a decrease or no change) was 3.26 (95% confidence interval [CI], 2.73-3.71). Although there were statistically significant interactions between creatinine change and race (P = .04), age (P < .001), history of hypertension (P = .01), and history of diabetes mellitus (P < .001) with respect to the outcome of ESRD, these interactions did not influence the main findings of greater ESRD risk associated with serum creatinine change.
After adjustment for demographic characteristics, comorbid medical conditions (history of stroke, hypertension, diabetes mellitus, previous myocardial infarction or CABG, and previous or current smoking), admission characteristics (presence of reduced kidney function, presence of anemia, and Acute Physiology and Chronic Health Evaluation II score), in-hospital treatment (receipt of thrombolytic therapy, percutaneous coronary interventions, CABG, and intensive care unit admission), and discharge medications (angiotensin-converting enzyme inhibitor, β-blocker, and aspirin prescription), there was a statistically significant association between all levels of serum creatinine increase and mortality after discharge (Figures 2B and 3B). The adjusted risk of death associated with a serum creatinine change of 0.6 to 3.0 mg/dL (vs a decrease or no change) was 1.39 (95% CI, 1.35-1.43). Although there were statistically significant interactions between creatinine change and age (P < .001), history of hypertension (P = .01), history of diabetes mellitus (P = .04), and presence of reduced kidney function (P = .04) with respect to the outcome of mortality, these interactions did not influence the main findings of the association of creatinine change with mortality.
In univariate and adjusted analyses, receipt of β-blockers and aspirin at the time of discharge was associated with a decreased adjusted risk of mortality (adjusted hazard ratios, 0.74 [95% CI, 0.72-0.75] and 0.83 [95% CI, 0.81-0.84], respectively). There was no statistically significant interaction between medication receipt and change in serum creatinine level with respect to mortality, suggesting that the association between these medications and mortality was comparable regardless of creatinine change during hospitalization.
In a nationally representative sample of Medicare beneficiaries discharged from the hospital after AMI, we have demonstrated an association between small increases in serum creatinine level during hospitalization and subsequent long-term risk of both ESRD and mortality independent of known risk factors. Not only were these risks manifested over 10 years of follow-up, but they were evident for even small increases in serum creatinine level and were present in a dose-response manner.
Although this is, to our knowledge, the first systematic description of creatinine increase and longer-term ESRD and mortality risk in a national study, there has been growing interest in the association between reductions in kidney function during hospitalization and risk of relatively short-term mortality.6,7,27-32 Growing evidence suggests that small changes in creatinine level are associated with higher postdischarge mortality.27,33 Loef et al,33 among patients undergoing cardiac surgery at a single center, demonstrated that patients experiencing a 25% increase in serum creatinine level were 1.63 times more likely to die over an 8-year period. Brown et al,27 in a study involving 8 medical centers, demonstrated a greater risk of death 90 days after discharge, including in-hospital mortality, associated with a 50% creatinine increase during hospitalization for CABG surgery. Our results within a national sample of patients from more than 4000 hospitals support these findings and further demonstrate that incremental changes in serum creatinine level are associated with a greater risk of death over a 10-year period.
Our results also demonstrate that small degrees of creatinine change among elderly patients not requiring dialysis during hospitalization are associated with a higher ESRD risk after discharge. Few data exist, but previous studies suggest that between 1.6% and 16.2% of patients surviving hospital-associated AKI severe enough to require dialysis may develop ESRD.34-36 A recent single-center study in Scotland demonstrated that, even among patients not requiring renal replacement therapy, decreases in kidney function as classified by the RIFLE criteria (risk, injury, failure, loss, and ESRD) of the Acute Dialysis Quality Initiative37 were associated with risk of renal function loss at 90 days.38 In recognition of the adverse events associated with small changes in creatinine level, the Acute Kidney Injury Network has suggested that the least severe AKI stage be identified as an abrupt (< 48-hour) increase in serum creatinine level of 0.3 mg/dL or more, or 1.5- to 2-fold or more from baseline.1 The long-term risks that we observed suggest that even the least severe stage of kidney injury may indicate a worse prognosis.
The long-term clinical impact of AKI among the elderly is of particular importance given its high prevalence within this age group38-40 and its increasing incidence.4 In our study, increases in serum creatinine level were associated with higher risks of both ESRD and death after AMI, but it is important to note that the absolute risk of mortality was far greater than that of ESRD (154.7 vs 3.8 cases per 1000 person-years, respectively), consistent with other studies.41-43 Clinicians caring for elderly patients after AMI should be aware of changes in creatinine level during hospitalization and target these patients for close monitoring and aggressive treatment. In our sample, the prescription of β-blockers and aspirin, consistent with guidelines contemporaneous with the baseline cohort,44 exhibited a protective association with future mortality regardless of creatinine change during hospitalization, but patients with greater creatinine changes were undertreated.
The reasons for the association between small changes in serum creatinine level and adverse outcomes are ill defined. An inflammatory response and important changes in many physiologic functions and organ systems accompany an acute reduction in kidney function.45-49 Subsequent risk of adverse events, therefore, may be associated with these primary effects of decreases in kidney function. On the other hand, these small increases in creatinine level may also reflect unmeasured comorbidities or subclinical pathologic processes within these patients. In addition, as our study demonstrates, these individuals are at risk of ESRD, and it is possible that changes in kidney function during hospitalization increase the risk of kidney disease progression among these patients and, in turn, the risk of cardiovascular morbidity and mortality associated with kidney disease.50,51
The smallest changes in serum creatinine level described herein are within the range that may be due to chance variation at the individual level or the “critical difference” (CD) of a serum creatinine measurement.52,53In our study, the CD for mean admission serum creatinine level likely subsumes the range of increase in serum creatinine level for the 2 smallest categories of creatinine change (0.1 and 0.2 mg/dL) and includes a portion of the range for the third-smallest category (0.3-0.5 mg/dL). On the basis of the 1994 College of American Pathology Survey’s54 summary serum creatinine level coefficients of variation for US hospitals, and a 4% intraindividual coefficient for variation of creatinine level,55 the CD for mean admission serum creatinine level in this sample is estimated at ±0.33 mg/dL for patients in the first 2 groups (0.1 and 0.2 mg/dL) and ±0.36 mg/dL in the third (0.3-0.5 mg/dL). Notably, according to the more recent 2006 College of American Pathology survey,56 these increases in serum creatinine level would still be included within the CD with the use of current methods. Our study therefore describes greater risks of events that, in the aggregate, are associated with changes in serum creatinine level below the detectable threshold of meaningful variation at the individual level by means of current assays, demonstrating the need for more sensitive markers of kidney function and biomarkers of AKI in clinical practice.1
There are notable limitations to our study. First, our study sample was a population of elderly patients with AMI at high risk of adverse events after hospitalization, and our results may not be applicable to younger populations or patients without cardiovascular disease. Further research is necessary to determine whether these results apply to populations with less comorbidity. Second, this was a retrospective observational analysis from which causal inference cannot be derived and which is subject to bias from unmeasured factors. The size and national scope of the data provide a high statistical power, and the ability to provide extensive adjustment given the large number of available variables helps to reduce but not eliminate the potential for residual confounding. Additional research should focus on replicating these results in smaller, prospective cohorts and determining patient subgroups that may be at most risk. Third, we could not identify the cause of creatinine change or patients' outpatient serum creatinine levels. Rather, our analysis addressed the relationship between serum creatinine increase experienced during hospitalization from all causes and subsequent adverse events. Most causes of AKI may be multifactorial57 and, because our sample consists of elderly patients hospitalized with AMI, likely related to congestive heart failure,58 percutaneous transluminal coronary angioplasty,6 CABG surgery,27 medication toxicity,59 and volume depletion.2 Fourth, we did not have access to information related to subsequent medical care. However, the risks of mortality and ESRD associated with creatinine change were constant over time, suggesting that events subsequent to hospitalization had little impact on a patient's future event risk. Finally, although the laboratory instruments used to measure serum creatinine in each hospital were not calibrated to a standard, the exposure of interest in this analysis is change in serum creatinine level rather than the creatinine level itself. To the extent that calibration would influence admission serum creatinine level (a potential confounder), it is unlikely that it would be differentially related to the exposure of interest (ie, change in serum creatinine level), biasing the results toward the null.
In a nationally representative sample of Medicare beneficiaries discharged from the hospital after admission for AMI, we have demonstrated an independent association between small changes in serum creatinine level during hospitalization and subsequent risk of mortality and ESRD. These risks were evident in a dose-response manner and were present during 10 years of follow-up. Further research should focus on etiologic mechanisms connecting AKI and risk of adverse events, increasing the sensitivity of current methods of detecting AKI, and improving the care of patients discharged after AMI to reduce long-term mortality and risk of ESRD.
Correspondence: Britt B. Newsome, MD, MPH, MSPH, Division of Nephrology, University of Alabama at Birmingham, 1530 Third Ave S, Room 524, Birmingham, AL 35294.
Accepted for Publication: October 23, 2007.
Author Contributions:Study concept and design: Newsome, McClellan, and Allison. Acquisition of data: Newsome, McClellan, Eggers, and Allison. Analysis and interpretation of data: Newsome, Warnock, McClellan, Herzog, Kiefe, and Allison. Drafting of the manuscript: Newsome, Warnock, Herzog, and Allison. Critical revision of the manuscript for important intellectual content: Newsome, Warnock, McClellan, Herzog, Kiefe, Eggers, and Allison. Statistical analysis: Newsome, Kiefe, and Allison. Administrative, technical, and material support: Newsome, Warnock, Eggers, and Allison. Study supervision: Newsome, McClellan, and Allison.
Financial Disclosure: None reported.
Previous Presentation: The results reported in this article were presented in poster form at the American Society of Nephrology meeting; November 4, 2007; San Francisco, California.
Additional Contributions: Allan Collins, MD, FASN, and Shu-Cheng Chen, MS, from the US Renal Data System Coordinating Center, assisted and collaborated in collecting these data. Anupam Agarwal, MD, and Robert Hardy, PhD, provided suggestions and discussions during the preparation of the manuscript. Sijan Zhang, MS, assisted with the conduct of the analyses.
1.Mehta
RLKellum
JShah
S
et al. Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury.
Crit Care 2007;11
(2)
R31
PubMedGoogle ScholarCrossref 2.Hou
SHBushinsky
DWish
JCohen
JHarrington
J Hospital-acquired renal insufficiency: a prospective study.
Am J Med 1983;74
(2)
243- 248
PubMedGoogle ScholarCrossref 4.Xue
JLDaniels
FStar
RA
et al. Incidence and mortality of acute renal failure in Medicare beneficiaries, 1992 to 2001.
J Am Soc Nephrol 2006;17
(4)
1135- 1142
PubMedGoogle ScholarCrossref 5.Mehta
RLPascual
MTSoroko
S
et al. Spectrum of acute renal failure in the intensive care unit: the PICARD experience.
Kidney Int 2004;66
(4)
1613- 1621
PubMedGoogle ScholarCrossref 6.Weisbord
SDChen
HStone
RA
et al. Associations of increases in serum creatinine with mortality and length of hospital stay after coronary angiography.
J Am Soc Nephrol 2006;17
(10)
2871- 2877
PubMedGoogle ScholarCrossref 7.Chertow
GMBurdick
EHonour
MBonventre
JVBates
DW Acute kidney injury, mortality, length of stay, and costs in hospitalized patients.
J Am Soc Nephrol 2005;16
(11)
3365- 3370
PubMedGoogle ScholarCrossref 8.Jencks
SFWilensky
GR The health care quality improvement initiative: a new approach to quality assurance in Medicare.
JAMA 1992;268
(7)
900- 903
PubMedGoogle ScholarCrossref 9.Ellerbeck
EFJencks
SFRadford
MJ
et al. Quality of care for Medicare patients with acute myocardial infarction: a four-state pilot study from the Cooperative Cardiovascular Project.
JAMA 1995;273
(19)
1509- 1514
PubMedGoogle ScholarCrossref 10.Marciniak
TAEllerbeck
EFRadford
MJ
et al. Improving the quality of care for Medicare patients with acute myocardial infarction: results from the Cooperative Cardiovascular Project.
JAMA 1998;279
(17)
1351- 1357
PubMedGoogle ScholarCrossref 11. International Classification of Diseases, Ninth Revision, Clinical Modification. 6th ed. Washington, DC Public Health Service, US Dept of Health and Human Services1997;
12.Allison
JJKiefe
CIWeissman
NW
et al. Relationship of hospital teaching status with quality of care and mortality for Medicare patients with acute MI.
JAMA 2000;284
(10)
1256- 1262
PubMedGoogle ScholarCrossref 13.Berger
AKRadford
MJKrumholz
HM Factors associated with delay in reperfusion therapy in elderly patients with acute myocardial infarction: analysis of the Cooperative Cardiovascular Project.
Am Heart J 2000;139
(6)
985- 992
PubMedGoogle ScholarCrossref 14.Berger
AKDuval
SKrumholz
HM Aspirin, beta-blocker, and angiotensin-converting enzyme inhibitor therapy in patients with end-stage renal disease and an acute myocardial infarction.
J Am Coll Cardiol 2003;42
(2)
201- 208
PubMedGoogle ScholarCrossref 15.Waikar
SSWald
RChertow
GM
et al. Validity of
International Classification of Diseases, Ninth Revision, Clinical Modification codes for acute renal failure.
J Am Soc Nephrol 2006;17
(6)
1688- 1694
PubMedGoogle ScholarCrossref 16.Pan
CXGlynn
RJMogun
HChoodnovsky
IAvorn
J Definition of race and ethnicity in older people in Medicare and Medicaid.
J Am Geriatr Soc 1999;47
(6)
730- 733
PubMedGoogle Scholar 17.Lauderdale
DSGoldberg
J The expanded racial and ethnic codes in the Medicare data files: their completeness of coverage and accuracy.
Am J Public Health 1996;86
(5)
712- 716
PubMedGoogle ScholarCrossref 18.Levey
ASBosch
JPLewis
JBGreene
TRogers
NRoth
D A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation.
Ann Intern Med 1999;130
(6)
461- 470
PubMedGoogle ScholarCrossref 19.Levey
ASGreene
TKusek
JBeck
GJGroup
MS A simplified equation to predict glomerular filtration rate from serum creatinine [abstract A0828].
J Am Soc Nephrol 1999;11155A
Google Scholar 20.National Kidney Foundation, K/DOQI clinical practice guidelines for chronic kidney disease: evaluation, classification, and stratification.
Am J Kidney Dis 2002;39
(2)
((suppl 1))
S1- S266
PubMedGoogle ScholarCrossref 21.Go
ASChertow
GMFan
D McCulloch
CEHsu
C-y Chronic kidney disease and the risks of death, cardiovascular events, and hospitalization.
N Engl J Med 2004;351
(13)
1296- 1305
PubMedGoogle ScholarCrossref 22.World Health Organization, Iron Deficiency Anemia. Geneva, Switzerland World Health Organization1959;WHO Technical Report Series
23.Rosner
B Fundamentals of Biostatistics. 4th ed. Belmont, CA Wadsworth Publishing Co1995;
24.Kleinbaum
DGKlein
M Survival Analysis: A Self-learning Text. 2nd ed. New York, NY Springer2005;
25.Allison
PD Survival Analysis Using SAS: A Practical Guide. Cary, NC SAS Institute Inc2004;
26.Rubin
DB Multiple Imputation of Nonresponse in Surveys. New York, NY John Wiley & Sons Inc1987;
27.Brown
JRCochran
RPDacey
LJ
et al. Perioperative increases in serum creatinine are predictive of increased 90-day mortality after coronary artery bypass graft surgery.
Circulation 2006;114
(1)
((suppl))
I409- I413
PubMedGoogle ScholarCrossref 28.Smith
GLVaccarino
VKosiborod
M
et al. Worsening renal function: what is a clinically meaningful change in creatinine during hospitalization with heart failure?
J Card Fail 2003;9
(1)
13- 25
PubMedGoogle ScholarCrossref 29.Lassnigg
ASchmidlin
DMouhieddine
M
et al. Minimal changes of serum creatinine predict prognosis in patients after cardiothoracic surgery: a prospective cohort study.
J Am Soc Nephrol 2004;15
(6)
1597- 1605
PubMedGoogle ScholarCrossref 30.Gottlieb
SSAbraham
WButler
J
et al. The prognostic importance of different definitions of worsening renal function in congestive heart failure.
J Card Fail 2002;8
(3)
136- 141
PubMedGoogle ScholarCrossref 31.Ahlström
ATallgren
MPeltonen
SRäsänen
PPettilä
V Survival and quality of life of patients requiring acute renal replacement therapy.
Intensive Care Med 2005;31
(9)
1222- 1228
PubMedGoogle ScholarCrossref 32.Korkeila
MRuokonen
ETakala
J Costs of care, long-term prognosis and quality of life in patients requiring renal replacement therapy during intensive care.
Intensive Care Med 2000;26
(12)
1824- 1831
PubMedGoogle ScholarCrossref 33.Loef
BGEpema
AHSmilde
TD
et al. Immediate postoperative renal function deterioration in cardiac surgical patients predicts in-hospital mortality and long-term survival.
J Am Soc Nephrol 2005;16
(1)
195- 200
PubMedGoogle ScholarCrossref 34.Metcalfe
WSimpson
MKhan
IH
et al. Acute renal failure requiring renal replacement therapy: incidence and outcome.
QJM 2002;95
(9)
579- 583
PubMedGoogle ScholarCrossref 37.Bellomo
RRonco
CKellum
JARavindra
LMPalevsky
P 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.
Crit Care 2004;8
(4)
R204- R212
PubMedGoogle ScholarCrossref 38.Ali
TKhan
ISimpson
W
et al. Incidence and outcomes in acute kidney injury: a comprehensive population-based study.
J Am Soc Nephrol 2007;18
(4)
1292- 1298
PubMedGoogle ScholarCrossref 41.Keith
DSNichols
GAGullion
CMBrown
JBSmith
DH Longitudinal follow-up and outcomes among a population with chronic kidney disease in a large managed care organization.
Arch Intern Med 2004;164
(6)
659- 663
PubMedGoogle ScholarCrossref 42.Thomas
MCCooper
MEShahinfar
SBrenner
BM Dialysis delayed is death prevented: a clinical perspective on the RENAAL study.
Kidney Int 2003;63
(4)
1577- 1579
PubMedGoogle ScholarCrossref 43.Foley
RNMurray
AMLi
S
et al. Chronic kidney disease and the risk for cardiovascular disease, renal replacement, and death in the United States Medicare population, 1998 to 1999.
J Am Soc Nephrol 2005;16
(2)
489- 495
PubMedGoogle ScholarCrossref 44.Ryan
TJAnderson
JLAntman
EM
et al. ACC/AHA guidelines for the management of patients with acute myocardial infarction: executive summary: a report of the American College of Cardiology/American Heart Association Task Force on Practice Guidelines (Committee on Management of Acute Myocardial Infarction).
Circulation 1996;94
(9)
2341- 2350
PubMedGoogle ScholarCrossref 45.Druml
W Long term prognosis of patients with acute renal failure: is intensive care worth it?
Intensive Care Med 2005;31
(9)
1145- 1147
PubMedGoogle ScholarCrossref 47.Kramer
AAPostler
GSalhab
KFMendez
CCarey
LCRabb
H Renal ischemia/reperfusion leads to macrophage-mediated increase in pulmonary vascular permeability.
Kidney Int 1999;55
(6)
2362- 2367
PubMedGoogle ScholarCrossref 48.Donnahoo
KKShames
BDHarken
AHMeldrum
DR Review article: the role of tumor necrosis factor in renal ischemia-reperfusion injury.
J Urol 1999;162
(1)
196- 203
PubMedGoogle ScholarCrossref 49.Hoste
EAClermont
GKersten
A
et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis.
Crit Care 2006;10
(3)
R73
PubMedGoogle ScholarCrossref 51.Ritz
E McClellan
WM Increased cardiovascular risk in patients with minor renal dysfunction: an emerging issue with far-reaching consequences.
J Am Soc Nephrol 2004;15
(3)
513- 516
PubMedGoogle ScholarCrossref 52.Costongs
GMJanson
PCBas
BMHermans
Jvan Wersch
JWBrombacher
PJ Short-term and long-term intra-individual variations and critical differences of clinical chemical laboratory parameters.
J Clin Chem Clin Biochem 1985;23
(1)
7- 16
PubMedGoogle Scholar 53.Froissart
MRossert
JAJacquot
CPaillard
MHouillier
P Predictive performance of the Modification of Diet in Renal Disease and Cockroft-Gault equations for estimating renal function.
J Am Soc Nephrol 2005;16
(3)
763- 773
PubMedGoogle ScholarCrossref 54. 1994 CAP Survey Set C3, C4, C5, C6, C7-C. Northfield, IL College of American Pathologists1994;
55.Ricós
CAlvarez
VCava
F
et al. Current databases on biological variation: pros, cons and progress.
Scand J Clin Lab Invest 1999;59
(7)
491- 500
PubMedGoogle ScholarCrossref 56. 2006 CAP Survey Set C-B. Northfield, IL College of American Pathologists2006;
58.Behrend
TMiller
SB Acute renal failure in the cardiac care unit: etiologies, outcomes, and prognostic factors.
Kidney Int 1999;56
(1)
238- 243
PubMedGoogle ScholarCrossref 59.Liaño
FPascual
J Epidemiology of acute renal failure: a prospective, multicenter community-based study.
Kidney Int 1996;50
(3)
811- 818
PubMedGoogle ScholarCrossref