Secondary Logo

Journal Logo

Clinical Aspects



Chang, Chih-Hsiang; Lin, Chan-Yu; Tian, Ya-Chung; Jenq, Chang-Chyi; Chang, Ming-Yang; Chen, Yung-Chang; Fang, Ji-Tseng; Yang, Chih-Wei

Author Information
Shock 33(3):p 247-252, March 2010. | DOI: 10.1097/SHK.0b013e3181b2fe0c
  • Free



Acute kidney injury (AKI) is a common and serious complication in critically ill patients. The mortality rate remains high despite improved renal replacement techniques. A possible cause of the high mortality rate is that intensive care unit (ICU) patients tend to be older and more debilitated than before. Pathophysiological factors associated with AKI are also incriminated in the failure of other organs, indicating that AKI is often part of a multiple organ failure syndrome (1, 2). Before the development of the RIFLE classification (acronym indicating Risk of renal failure, Injury to the kidney, Failure of kidney function, Loss of kidney function, and End-stage renal failure) system, widely varying definitions of AKI limited epidemiologic investigations of incidence and outcomes in critically ill patients (3). These varying definitions also have generated clinical confusion and complicated comparisons of data between studies (4, 5). The RIFLE classification was first proposed by the Acute Dialysis Quality Initiative (ADQI) group to standardize acute renal failure study in 2004 (Table 1) (6). To date, the classification has proven useful not only for unique populations such as cirrhotic patients and those requiring extracorporeal membrane oxygenation for postcardiotomy cardiogenic shock (7, 8), but also for diagnosing and classifying the severity of AKI in heterogeneous hospitalized patients in wards and ICUs (9-12).

Table 1:
RIFLE and AKIN classification schemes for AKI

Recently, the Acute Kidney Injury Network (AKIN) group, which is composed pf nephrologists and intensivists, has proposed modifying the RIFLE criteria. In AKIN stage 1 (analogous to RIFLE-Risk), a smaller change within 48 h in serum creatinine (SCr) greater than 0.3 mg/dL (≥26.2 μmol/L) was suggested as an AKI threshold (Tables 1 and 2). Additionally, patients receiving renal replacement therapy were reclassified as AKIN stage 3 (RIFLE-Failure). Finally, the loss and end-stage kidney disease classification were deleted from the RIFLE system. The reason for these changes was to increase the sensitivity of the RIFLE criteria (13). The few reported studies comparing the two measures have shown little difference between them (14, 15). Therefore, this retrospective study compared the efficacy of each scoring system in predicting outcome in ICU patients.

Table 2:
Comparison of the RIFLE and AKIN definition and classification schemes for acute kidney injury


Patient information and data collection

The local institutional review board approved this study and waived the need for informed consent. This investigation was performed in ICUs at a tertiary care referral center in Taiwan. Post hoc analysis of an accumulated database enrolled 291 heterogeneous critically ill patients with septic shock (16), acute respiratory distress syndrome (ARDS) (17), or hepatic cirrhosis (8) requiring intensive monitoring and/or treatment unavailable elsewhere. The following patients were excluded: pediatric patients (aged ≤18 years), chronic uremic patients undergoing renal replacement therapy, and patients whose hospital stay was less than 24 h. Readmitted patients were also excluded from this study. The RIFLE, AKIN category, Acute Physiology and Chronic Health Evaluation (APACHE) II (18), and Sequential Organ Failure Assessment (SOFA) (19) were also evaluated in the study. Records were collected from patients admitted to medical ICUs between March 2003 and February 2006. Retrospectively collected data were the following: demographic information; underlying diseases; AKIN, RIFLE category, APACHE II, and SOFA scores at admission to an ICU; and length of hospitalization. The primary outcome was hospital mortality. Follow-up at 6 months after hospital discharge was performed via a telephone interview. When necessary, the hospital registry office provided information regarding patient survival or date of death.


Septic shock was defined according to modified American College of Chest Physicians and Society of Critical Care Medicine consensus criteria (112 patients) (20). Fifty-seven patients met ARDS criteria defined by the American-European consensus conference (21). Cirrhosis (122 patients) was diagnosed by liver histology or by combined findings of physical, biochemical, and ultrasonographic examination. The worst physiological and biochemical values on day of initial ICU admission were recorded.

The ADQI group first proposed the RIFLE system at the second ADQI Conference in Vicenza, Italy in May 2002. This classification system uses individual criteria for SCr levels and urine output (UO). The criteria resulting in the worst possible classification are used. Patients are classified into three severity categories (risk, injury, and failure) and two clinical outcome categories (loss and end-stage renal disease). Notably, the F component of RIFLE is present even when the increase in SCr is less than 3-fold as long as the subsequent SCr is greater than 4.0 mg/dL (350 μmol/L) in the setting of an acute increase of greater than 0.5 mg/dL (44 μmol/L). In this study, baseline SCr concentration was measured first during hospitalization. The Modification of Diet in Renal Disease formula was applied for 20 patients who were admitted directly to an ICU, and their SCr concentrations at admission were unknown (6). Patients were classified by RIFLE into risk, injury, or failure categories.

The AKIN classification (13) differs from the RIFLE classification as follows: it reduces the need for baseline creatinine but does require at least two creatinine values within 48 h; AKIN stage 1 is similar to RIFLE-R but includes abrupt (within 48 h) reduction in kidney function (increase in SCr ≥0.3 mg/dL); injury and failure are the same as stages 2 and 3, respectively; stage 3 also includes patients who need renal replacement therapy in any stage; two outcome classes, loss and end-stage kidney disease, are omitted. We used a simple model for mortality-non-AKI and AKIN stage 0 (0 points); RIFLE-R and AKIN stage 1 (1 point), RIFLE-I and AKIN stage 2 (2 points), and RIFLE-F and AKIN stage 3 (3 points) for day 1 of ICU admission (7, 8).

Clinical management

Briefly, cirrhosis patients with gastrointestinal bleeding caused by esophageal varices were initially treated with emergency sclerotherapy combined with vasopressin derivatives administration. Patients with peptic ulcer either with active bleeding or clot were treated with injection of sclerosing agents, followed by proton pump inhibitors. Intravenous fluids were administered to all patients depending on their volume status. Blood transfusion was administered according to the criteria of the attending physician or whenever the hemoglobin decreased below 8 g/dL. In all patients, the presence of bacterial infections at admission and their development during hospitalization were investigated with appropriate diagnostic methods and cultures. Patients were then started on appropriate empiric antibiotic therapy intravenously. In patients developing signs of acute renal failure (oliguria and/or increase in serum creatinine), blood volume expansion with packed red blood cells, albumin, and/or artificial plasma expanders was given to improve renal function and increase urine volume. If oliguria persisted after volume depletion had been corrected or excluded, vasoactive agents with or without the addition of a loop diuretic were prescribed. When acute renal failure was severe or progressive and measured to improve renal function had been unsuccessful, renal replacement therapy was performed.

For ARDS patients, the strategy for the ventilation consists of an initial low tidal volume of 6 to 8 mL/kg for either volume- or pressure-controlled ventilation. Mechanical ventilatory adequacy is monitored by arterial blood gas measurements with the ventilator settings changed as needed. Pulse oximetry is used to monitor oxygen saturation, and the FiO2 is adjusted to maintain pulse oximetric saturation greater than 90% and avoid raising the plateau airway pressure greater than 35 cm of water. The ventilatory rate ranged from 20 to 25/min, peak airway pressure ranged from 26 to 38 cm of water, and peak end-expiratory pressure ranged from 10 to 16 cm of water. The fraction of inspirited oxygen ranged from 0.9 to 1 in all patients.

For sepsis patients, the early goal-directed resuscitation of the septic patient during the first 6 h after recognition; appropriate diagnostic studies to ascertain causative organisms before starting antibiotics; early administration of broad-spectrum antibiotic therapy; reassessment of antibiotic therapy with microbiology and clinical data to narrow coverage, when appropriate; a usual 7 to 10 days of antibiotic therapy guided by clinical response; use of crystalloid, colloid resuscitation, and aggressive fluid challenge to restore mean circulating filling pressure; vasopressor preference for norepinephrine and dopamine; avoidance of supranormal oxygen delivery as a goal of therapy; stress-dose steroid therapy for septic shock; targeting a hemoglobin of 7 to 9 g/dL; appropriate use of fresh frozen plasma and platelets; a semirecumbent bed position unless contraindicated; and maintenance of blood glucose <150 mg/dL after initial stabilization.

Statistical analysis

Descriptive statistics were expressed as mean ± SE. Primary analysis compared hospital survivors with nonsurvivors. All variables were tested for normal distributions using the Kolmogorov-Smirnov test. The Student t test was applied to compare the means of continuous variables and normally distributed data; otherwise, the Mann-Whitney U test was used. Categorical data were tested using the chi-square test. This study used the chi-square test for trend to assess categorical data associated with RIFLE classification. Finally, risk factors were assessed by univariate analysis.

Calibration was assessed using the Hosmer-Lemeshow goodness-of-fit test to compare the number of observed and predicted deaths in risk groups for the entire range of death probabilities. Discrimination was assessed using the area under a receiver operating characteristic curve (AUROC). Areas under receiver operating characteristic curves were compared by a nonparametric approach. The AUROC analysis was also conducted to calculate cutoff values, sensitivity, specificity, and overall correctness. Finally, cutoff points were calculated by acquiring the best Youden index (sensitivity + specificity - 1). Cumulative survival curves as a function of time were generated using the Kaplan-Meier approach and compared using the log-rank test. All statistical tests were two-tailed; a value of P < 0.05 was considered statistically significant. Data were analyzed using SPSS 12.0 for Windows (SPSS Inc., Chicago, Ill).


Subject characteristics

Between March 2003 and April 2006, 291 ICU patients with septic shock, ARDS, or hepatic cirrhosis were enrolled. Median patient age was 62 years; 204 (70%) were men and 87 (30%) were female. Overall in-hospital mortality was 60.8% (177/291). Table 3 lists patient demographic data, clinical characteristics, and underlying diseases of both survivors and nonsurvivors. In this study, 160 (55%) patients were in sepsis. Thirty-five (12%) patients had underlying cancer, 122 (42%) patients had liver cirrhosis, and 80 (28%) patients had diabetes mellitus.

Table 3:
Patient demographic data and clinical characteristics

Hospital mortality and short-term prognosis

In RIFLE classification, hospital mortality was 36.8% (42/114) for non-AKI patients, 63.2% (24/38) for RIFLE-R (risk), 69.2% (36/52) for RIFLE-I (injury), and 86.2% (75/87) for RIFLE-F (failure; chi-square for trend; P < 0.001; Table 4). A progressive and significant elevation in mortality was correlated with increasing RIFLE classification severity among all patients. Odds ratios for RIFLE criteria were 2.94 (P = 0.006) for RIFLE-R versus non-AKI, 3.86 (P < 0.001) for RIFLE-I versus non-AKI, and 10.71 (P < 0.001) for RIFLE-F versus non-AKI (Table 5). Hospital mortality differed significantly (P < 0.001) according to less than and greater than cutoffs of non-acute renal failure/RIFLE-R, RIFLE-I, and RIFLE-F for septic shock (40% vs. 71.9%), liver cirrhosis (38.6% vs. 92.2%), and ARDS (27.3% vs. 61.2%).

Table 4:
Clinical outcomes stratified by the RIFLE and AKI definition/classification schemes
Table 5:
Predictive ability for hospital mortality by separate logistic regression models for the RIFLE and AKIN definition/classification schemes

In AKIN classification, hospital mortality was 38.7% (36/93) for stage 0 patients, 52.6% (30/57) for stage 1, 67.3% (33/49) for stage 2, and 84.8% (78/92) for stage 3 (chi-square for trend; P < 0.001; Table 3). The AKIN classification also correlated with progressive and significant elevation in mortality. Odds ratios for RIFLE criteria were 1.76 (P = 0.097) for stage 1 versus stage 0, 3.07 (P = 0.002) for stage 2 versus stage 0, and 9.50 (P < 0.001) for stage 3 versus stage 0 (Table 5). Hospital mortality rates significantly differed (P < 0.001) less than and greater than cutoffs of AKIN stage 0/AKIN stages 1 to 3 for septic shock (41.5% vs. 67.6%), liver cirrhosis (40.6% vs. 82.9%), and ARDS (30% vs. 58.8%).

Table 6 shows the goodness of fit, as measured by the Hosmer-Lemeshow chi-square for predicted mortality risk, and the predictive accuracy of the RIFLE category, AKIN classification, APACHE II, and SOFA scores. Calibration of APACHE II and SOFA scores were superior to RIFLE category and AKIN classification. Table 6 also lists the discrimination for those scores. The ROC curve confirmed that the discriminatory power of the RIFLE classification (AUROC = 0.738 ± 0.030 [95% confidence interval {CI}, 0.680-0.796]; P < 0.001) was superior to that of AKIN score (AUROC = 0.720 ± 0.030 [95% CI, 0.680-0.796]; P < 0.001). In addition, SOFA score (AUROC = 0.796 ± 0.026 [95% CI, 0.746-0.846]; P < 0.001) had the best discrimination for ROC curve among them (APACHE II AUROC = 0.747 ± 0.029 [95% CI, 0.690-0.803]; P < 0.001).

Table 6:
Calibration and discrimination for the scoring methods in predicting hospital mortality

To compare the selected cutoff points for predicting hospital mortality, the sensitivity, specificity, and overall correctness of prediction were all determined. All four scoring systems were tested by Youden index (Table 7). Hospital mortality rates differed significantly (P < 0.001) less than and greater than cutoffs of non-AKI category of RIFLE classification, AKIN stage 1, 21 APACHE II points, and 9 SOFA points.

Table 7:
Subsequent hospital mortality predicted after ICU admission

Cumulative survival rates differed significantly (P < 0.05) for non-AKI versus RIFLE-R, RIFLE-I, and RIFLE-F (Fig. 1A). Figure 1B also shows that cumulative survival rates differed significantly (P < 0.05) for stage 0 versus stages 1 to 3 in the AKIN group.

Fig. 1:
Cumulative survival rate for 291 critically ill patients based on their RIFLE classification (2A) and AKIN stage (2B).


This retrospective study included 291 heterogeneous patients with critical illnesses. Overall in-hospital mortality rate was 60.8%, which is consistent with that obtained by previous studies, indicating poor prognosis of ICU patients with septic shock, ARDS, or hepatic cirrhosis. All patients were classified by RIFLE to identify and classify the severity of AKI and to compare its ability to predict hospital outcome in our study population, which, like those in earlier published studies, was composed of homogeneous or heterogeneous patients (2, 7, 8, 11, 16, 17, 22-31). The AKIN group aimed to improve the sensitivity and reproducibility of the AKI criteria and defined the AKIN classification. Bagshaw et al. (15) found that AKIN criteria may not improve sensitivity and predictive ability. Lopes et al. (14) later compared both scoring systems and found that AKIN classification had superior sensitivity to AKI but was inferior for outcome prediction in critically ill patients.

In the present study, compared with RIFLE, AKIN criteria identified 7.9% more patients (RIFLE 60.1% vs. AKIN 68.0%) as having AKI and classified more patients as stage 1 (RIFLE 13.1%-19.6%). Table 5 shows a trend toward significantly increased mortality rates associated with increasing RIFLE score for all patients (risk, 2.94; injury, 3.86; and failure, 10.71), which was also observed in AKIN classification (stage 1, 1.79; stage 2, 3,07; stage 3, 9.50).

Analytical results in this study demonstrated that both RIFLE and AKIN criteria precisely predicted hospital mortality and short-term prognosis (Fig. 1) in this heterogeneous subset of critically ill patients. The RIFLE score had better discriminatory power and overall correctness than the AKIN score. No material differences were noted between RIFLE and AKIN (Tables 6 and 7). Among RIFLE, AKIN, APACHE II, and SOFA scores, the SOFA score had the highest overall predictive accuracy and discrimination. Lack of extrarenal predictors in RIFLE and AKIN criteria may explain their discriminative inferiority to SOFA.

Despite the promising analytical results obtained in this study, several important limitations should be recognized. First, this retrospective study was performed at a single tertiary-care medical center, which limits generalization of its findings. Second, the patient group was mainly collected from patients with septic shock, ARDS, or hepatic cirrhosis and were being treated in medical ICUs; these conditions are associated with poor prognosis; thus, the results may not be directly extrapolated to other patient populations. Third, sequential measurement of these scoring systems (e.g. daily, weekly) may reflect the dynamic aspects of clinical diseases, thus providing superior information on mortality risk. Finally, use of predicted mortality as a clinical performance benchmark is limited by factors not directly related to care quality.


In summary, although capable of improving the sensitivity of the AKI diagnosis, the AKIN criteria do not improve the ability to predict a short-term outcome such as in-hospital mortality in critically ill patients. Further study of the RIFLE criteria in prospective randomized controlled clinical trials is needed to establish specific interventions for controlling the progression of AKI.


1. Uchino S, Kellum JA, Bellomo R, Doig GS, Morimatsu H, Morgera S, Schetz M, Tan I, Bouman C, Macedo E, et al.: Acute renal failure in critically ill patients-a multinational, multicenter study. JAMA 294:813-818, 2005.
2. Ostermann M, Chang RWS: Acute kidney injury in the intensive care unit according to RIFLE. Crit Care Med 35:1837-1843, 2007.
3. Kellum JA, Levin N, Bouman C, Bouman C: Developing a consensus classification system for acute renal failure. Curr Opin Crit Care 8:509-514, 2002.
4. Lameire N, Van Biesen W, Vanholder R: Acute renal failure. Lancet 365:417-430, 2005.
5. Schrier RW, Wang W, Poole B, Mitra A: Acute renal failure: definitions, diagnosis, pathogenesis, and therapy. J Clin Invest 114:5-14, 2004.
6. Bellomo R, Ronco C, Kellum JA, Mehta RL, Palevsky P, Acute Dialysis Quality Initiative Workgroup: 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 8:R204-R212, 2004.
7. Lin CY, Tsai FC, Tian YC, Jenq CC, Chen YC, Fang JT, Yang CW: Evaluation of outcome scoring systems for patients on extracorporeal membrane oxygenation. Ann Thorac Surg 84:1256-1263, 2007.
8. Jenq CC, Tsai MH, Tian YC, Lin CY, Yang C, Liu NJ, Lien JM, Chen YC, Fang JT, Chen PC, et al.: RIFLE classification can predict short-term prognosis in critically ill cirrhotic patients. Intensive Care Med 33:1921-1930, 2007.
9. Uchino S, Bellomo R, Goldsmith D, Bates S, Ronco C: An assessment of the RIFLE criteria for acute renal failure in hospitalized patients. Crit Care Med 34:1913-1917, 2006.
10. Ricci Z, Cruz D, Ronco C: The RIFLE criteria and mortality in acute kidney injury: A systematic review. Kidney Int 73:538-546, 2008.
11. Hoste EAJ, Clermont G, Kersten A, Venkataraman R, Angus DC, De Bacquer D, Kellum JA: RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Crit Care 10:R73, 2006.
12. Bellomo R, Kellum JA, Ronco C: Defining and classifying acute renal failure: from advocacy to consensus and validation of the RIFLE criteria. Intensive Care Med 33:409-413, 2007.
13. Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, Levin A: Acute Kidney Injury Network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 11:R31, 2007.
14. Lopes JA, Fernandes P, Jorge S, Goncalves S, Alvarez A, Costa e Silva Z, Franca C, Prata MM: Acute kidney injury in intensive care unit patients: a comparison between the RIFLE and the Acute Kidney Injury Network classifications. Critical Care 12:R110, 2008.
15. Bagshaw SM, George C, Bellomo R: A comparison of the RIFLE and AKIN criteria for acute kidney injury in critically ill patients. Nephrol Dial Transplant 23:1569-1574, 2008.
16. Chen YC, Jenq CC, Tian YC, Chang MY, Lin CY, Chang CC, Lin HC, Fang JT, Yang CW, Lin SM: RIFLE classification for predicting in-hospital mortality in critically ill sepsis patients. Shock 31:139-145, 2009.
17. Lin CY, Kao KC, Tian YC, Jenq CC, Chang MY, Chen YC, Fang JT, Huang CC, Tsai YH, Yang CW: The RIFLE score increases the accuracy of outcome prediction in patients with acute respiratory distress syndrome undergoing open lung biopsy. Respiration 77:398-406, 2009.
18. Knaus WA, Draper EA, Wagner DP, Zimmerman JE: APACHE II: a severity of disease classification system. Crit Care Med 13:818-829, 1985.
19. Vincent JL, Moreno R, Takala J, Willatts S, De Mendonca A, Bruining H, Reinhart CK, Suter PM, Thijs LG: The SOFA (Sepsis-related Organ Failure Assessment) score to describe organ dysfunction/failure. On behalf of the Working Group on Sepsis-Related Problems of the European Society of Intensive Care Medicine. Intensive Care Med 22:707-710, 1996.
20. Bone RC, Balk RA, Cerra FB, Dellinger RP, Fein AM, Knaus WA, Schein RMH, Sibbald WJ, Abrams JH, Bernard GR, et al.: American-College of Chest Physicians Society of Critical Care Medicine Consensus Conference-definitions for sepsis and organ failure and guidelines for the use of innovative therapies in sepsis. Crit Care Med 20:864-874, 1992.
21. Bernard GR, Artigas A, Brigham KL, Carlet J, Falke K, Hudson L, Lamy M, Legall JR, Morris A, Spragg R, et al.: Report of the American-European Consensus Conference on ARDS-definitions, mechanisms, relevant outcomes and clinical-trial coordination. Intensive Care Med 20:225-232, 1994.
22. Lopes JA, Jorge S, Silva S, de Almeida E, Abreu F, Martins C, do Carmo JA, Lacerda JF, Prata MM: An assessment of the RIFLE criteria for acute renal failure following myeloablative autologous and allogeneic haematopoietic cell transplantation. Bone Marrow Transplant 38:395, 2006.
23. Lopes JA, Jorge S, Neves FC, Caneira M, da Costa AG, Ferreira AC, Prata MM: An assessment of the rifle criteria for acute renal failure in severely burned patients. Nephrol Dial Transplant 22:285, 2007.
24. Ahlstrom A, Kuitunen A, Peltonen S, Hynninen M, Tallgren M, Aaltonen J, Pettila V: Comparison of 2 acute renal failure severity scores to general scoring systems in the critically ill. Am J Kidney Dis 48:262-268, 2006.
25. Kuitunen A, Vento A, Suojaranta-Ylinen R, Pettila V: Acute renal failure after cardiac surgery: evaluation of the RIFLE classification. Ann Thorac Surg 81:542-546, 2006.
26. Siner JM, Bhandari V, Engle KM, Elias JA, Siegel MD: Elevated serum angiopoietin 2 levels are associated with increased mortality in sepsis. Shock 31:348-353, 2009.
27. Guitard J, Cointault O, Kamar N, Muscari F, Lavayssiere L, Suc B, Ribes D, Esposito L, Barange K, Durand D, et al.: Acute renal failure following liver transplantation with induction therapy. Clin Nephrol 65:103-112, 2006.
28. Nguyen HB, Banta JE, Cho TW, Van Ginkel C, Burroughs K, Wittlake WA, Corbett SW: Mortality predictions using current physiologic scoring systems in patients meeting criteria for early goal-directed therapy and the severe sepsis resuscitation bundle. Shock 30:23-28, 2008.
29. O'Riordan A, Wong V, McQuillan R, McCormick PA, Hegarty JE, Watson AJ: Acute renal disease, as defined by the RIFLE criteria, post-liver transplantation. Am J Transplant 7:168-176, 2007.
30. Rhee JY, Kwon KT, Ki HK, Shin SY, Jung DS, Chung DR, Ha BC, Peck KR, Song JH: Scoring systems for prediction of mortality in patients with intensive care unit-acquired sepsis: a comparison of the Pitt bacteremia score and the acute physiology and chronic health evaluation ii scoring systems. Shock 31:146-150, 2009.
31. Lee CC, Chen SY, Tsai CL, Wu SC, Chiang WC, Wang JL, Sun HY, Chen SC, Chen WJ, Hsueh PR: Prognostic value of mortality in emergency department sepsis score, procalcitonin, and C-reactive protein in patients with sepsis at the emergency department. Shock 29:322-327, 2008.

Acute renal failure; intensive care unit; scoring system; outcome

©2010The Shock Society