- Email: [email protected]
Renal injury study in critical ill patients in accordance with the new definition given by the Acute Kidney Injury Network
Journal of Critical Care (2011) 26, 206–212
Renal Injury
Renal injury study in critical ill patients in accordance with the new definition given by the Acute Kidney Injury Network☆,☆☆ Nelson Javier Fonseca Ruiz MD, MSc a,b,⁎, Diana Paola Cuesta Castro MD, MSc c , Ana Milena Mesa Guerra MD c,d , Francisco Molina Saldarriaga MD, MSc a,b , Juan Diego Montejo Hernández MD e a
Intensive Care Unit, University Bolivariana Clinic, Medellin, Colombia Intensive Care Unit, CORBIC Institute, Medellin, Colombia c Pontificia Bolivariana University, Medellin, Colombia d Seedbed Investigation Medicine Faculty (SIFAM), Pontificia Bolivariana University, Medellin, Colombia e Pontificia Bolivariana University, Medellin Clinic, Colombia b
Keywords: Acute renal injury; Consensus definition; Epidemiology; Critical illness; Mortality
Abstract Objective: This research aims to apply the definition proposed by the Acute Kidney Injury Network (AKIN) research group to assess the incidence, risk factors, and outcomes in acute kidney injury (AKI) patients admitted at the intensive care unit (ICU). Design: This is a retrospective cohort study. Patients who were admitted to the ICU from January 1, 2003 to December 31, 2004 were studied. Interventions: Medical records of all patients were reviewed. Demographic information, diagnoses, risk factors for AKI, laboratory data, urinary output, frequency and days of exposure to mechanical ventilation, ICU and hospital stay, and outcomes were recorded. Measurements and Main Results: A total of 794 patients were studied. There were 39.8% of patients who presented AKI (stage 1: 13.9%, stage 2: 12%, stage 3: 13.9%). The variables that were associated with the presence of AKI in the multivariable analysis were as follows: sepsis (odds ratio [OR], 5.29; 95% confidence interval [CI], 3.36-8.33), heart failure (OR, 3.01; 95% CI, 1.59-5.67), vasopressor use (OR, 1.89; 95% CI, 1.26-2.83), and age (β = 1.02; 95% CI, 1.01-1.03). The mean hospital stay increased with renal commitment: patients without AKI, 10.9 days; AKIN stage 1, 17.8; AKIN stage 2, 21.1; and AKIN stage 3, 22.1 days (P b .0001). Mortality rate increased as more advanced the AKI stage was (no AKI, 7.3%; AKI 1, 16.4%; AKI 2, 34.7%; and AKIN 3, 45.5%; P b .0001).
☆
Financial support: Integrated Center for Research Development, Universidad Pontificia Bolivariana. The authors have not disclosed any potential conflicts of interest in relation to this manuscript. ⁎ Corresponding author. E-mail address: [email protected] (N.J. Fonseca Ruiz). ☆☆
0883-9441/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jcrc.2010.06.011
Renal injury study in critical ill patients
207
Conclusions: All of the result indicators—stay days in ICU, hospital stay days, frequency and days of mechanical ventilation, and mortality—considerably increased with more acute AKI stage. The most important risk factor of AKI was the sepsis. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Although the term “acute renal failure” was introduced since 1951, there has never been a standard biochemical definition and there has not been consensus as to the diagnosis criterion, clinical definition, or disease stages for it. This has resulted in the appearance of multiple definitions used for research. Kellum et al [1] have described more than 35 different definitions. The prevalence and mortality reported in patients of intensive care unit (ICU) due to the lack of uniformity on the definition is wide, ranging from 1% to 25% and 15% to 60%, respectively [2-4]. There has been little progress on research on preventive measurements and efficient treatments because there has not been a common language among clinicians and researchers that would allow the comparison of epidemiologic, therapeutic, and diagnosis variables. In recent years, several efforts have been made to achieve a unique definition [5,6], and in 2004, Bellomo et al [7] proposed a new definition for acute renal failure based on the stratification of the disease in several stages—risk, injury, failure, loss, and end-stage renal disease—so that pattern analysis would allow progress. This proposal also included entities like sepsis and acute respiratory distress syndrome. However, such a definition still named this entity “Acute Renal Failure,” decreasing the importance of the broad spectrum and extent the disease has since the initial disorder stage. It has been demonstrated that small elevations in serum creatinine levels increase the risk of complications and death. Levy et al [8] showed that a 25% increase on basal creatinine levels after the administration of contrast medium was associated with 5 times higher risk of intrahospital mortality. Lassnigg et al [9] published similar data in patients submitted to heart surgery. Also, Chertow et al [10] demonstrated that increments of just 0.3 mg/dL in serum creatinine level were related to cost, morbidity, and mortality increase in hospitalized patients. The stage called “Failure” also created confusion because it was named the same as the syndrome. In September 2005, a multidisciplinary group made up of several medical societies named Acute Kidney Injury Network (AKIN) gathered together in Amsterdam and defined “Acute Kidney Injury” (AKI) as any abrupt decrease (alteration in less than 48 hours) in renal function produced by any damage that could cause functional or structural changes in kidneys, setting aside the terms “Acute Renal Failure,” which refers to the final phases of the disease, and “Acute Tubular Necrosis,” which gave a physiopathologic predominant role to the ischemia so that it overlooked the
importance of the inflammatory phenomenon [11]. Besides, following the Risk, Injury, Failure, Loss, and End-stage Kidney (RIFLE) classification, they proposed stratification based on the urinary expenditure and changes the serum creatinine levels, naming the phases as stages 1, 2, and 3 of the initial phase and not “Risk,” “Injury,” “Failure” any more. Also, the categories “Loss” and “End-stage renal disease” disappeared because they were considered as results. This definition was issued in several medical journals, and it is proposed to be used to carry out clinical trials [12,13]. The aims of this study are to evaluate AKI incidence, to assess the risk factors associated with it, and to evaluate the relationship between hospital stay and hospital mortality with AKI presence and AKI stages in a group of general critically ill patients under the approach of the new definition proposed by the AKIN group.
2. Materials and methods A study of a retrospective cohort was carried out in patients treated in the ICU at University Bolivariana Clinic (UBC) in Medellín, Colombia, in the period from January 1, 2003 to December 31, 2004. The UBC is a third-level care center that has 200 hospitalization beds and 12 ICU beds. Patients older than 16 years who entered the ICU for any indication were included in the study. Patients referred from or to a different ICU were excluded due to the difficulties presented on data recollection. The information was gathered in a medical chart review where sociodemographic data as age and sex, clinical data as well as personal antecedents, and the reason for ICU entrance according to a diagnostic group list were recorded. Medical antecedents such as heart failure, hypertension, chronic obstructive pulmonary disease (COPD), tobacco smoking, coronary disease, diabetes, sepsis, vasopressor use, and the use of nephrotoxic medications were taken into account. Definitions were not standardized, but just recorded as they were found if they were going to be registered in the clinical profile. Other data considered were the severity of the condition measured by Acute Physiology and Chronic Health Evaluation scoring system (APACHE) version II and Simplified Acute Physiology Score (SAPS) version II, the severity of the clinical condition at the beginning of the renal injury measured by Sequential Organ Failure Assessment score (SOFA), the use of vasopressors, the starting and ending dates of the ventilatory mechanical support, ICU and
208
N.J. Fonseca Ruiz et al.
Table 1
Classification/staging system for AKI
Stage Serum creatinine criteria
Urine output criteria
1
Less than 0.5 mL/kg per hour for more than 6 h
2
3
Increase in serum creatinine ≥0.3 mg/dL (≥26.4 μmol/L) or increase to more ≥150% to 200% (1.5- to 2-fold) from baseline Increase in serum creatinine to more than 200% to 300% (N2- to 3-fold) from baseline Increase in serum creatinine to more than 300% (N3-fold) from baseline (or serum creatinine ≥4.0 mg/dL [≥354 μmol/L] with an acute increase of at least 0.5 mg/dL [44 μmol/L])
Less than 0.5 mL/kg per hour for more than 12 h Less than 0.3 mL/kg per hour for 24 h or anuria for 12 h
From Mehta et al [12].
hospitalization entrance and discharge dates, and final clinical result. To rank the patients according to the maximum AKIN category, the urine output and the daily changes in the serum creatinine level were assessed. Acute kidney injury diagnosis was analyzed in all the patients from entrance to ICU until exit from ICU based on AKI definition proposed by AKIN group, that is, an abrupt (within 48 hours) reduction in kidney function currently defined as an absolute increase in serum creatinine of more than or equal to 0.3 mg/dL, a percentage increase in serum creatinine of more than or equal to 50% (1.5-fold from baseline), or a reduction in urine output (documented oliguria of less than 0.5 mL/kg per hour for more than 6 hours). To sort out the patients in the different stages of AKI, all of those who had previously received any way of chronic renal support therapy (dialysis) before entrance to ICU were excluded. Patients were stratified according to the definitions of the AKIN group using both criteria: creatinine level increase and urinary output (Table 1).
Table 2
Sex Masculine (%) Feminine (%) Age (y), mean (SD) Personal history Hypertension (%) COPD (%) Tobacco use (%) Diabetes mellitus (%) Coronary artery disease (%) Heart failure (%) Other (%) Diagnostic group Sepsis (%) Abdominal surgery (%) Thoracic surgery (%) Lung disease (%) Cardiovascular disease (%) Neurosurgery (%) Trauma Neurology Obstetrics Gastrointestinal Other SAPS II, median (IQR) APACHE II, median (IQR) SOFA, median (IQR)
Mean and median of continuous variables were calculated with their respective standard deviation or interquartile range (IQR) depending on the case. Categorical variables were
Study flow chart.
44.6 15.8 27.1 18.2 8.5 9.6 5.1 19.6 15 14.6 10.3 9.6 8.6 4.2 4.0 2.8 2.8 8.5 27 (17-42) 13 (8-19) 4 (2-7)
AKI associated risk factors—univariate analysis
Risk factor
OR
95% CI
P
Sex: masculine sex Personal history Hypertension a COPD Tobacco use Diabetes mellitus a Coronary artery disease Heart failure a Sepsis a Nephrotoxic medications Aminoglycoside use a Vancomycin use a Contrast medium a Vasopressor use a
1.33
0.97-1.81
.064
1.90 1.50 1.22 1.82 1.57 3.14 7.08
1.39-2.59 1.00-2.25 0.87-1.70 1.21-2.73 0.90-2.75 1.74-5.65 4.72-10. 63
.0001 .063 .28 .005 .143 .0001 .0001
2.20 3.60 1.85 3.89
1.21-4.03 1.91-6.79 1.18-2.93 2.76-5.48
.013 .0001 .010 .0001
a
Fig. 1
53.4 46.6 52.6 (19.5)
described by means of proportions. A univariate analysis was used to establish a relationship between risk factors (sex, personal history, sepsis, nephrotoxic drugs, and vasopressor use) and AKI. Whenever appropriate, results are presented as relative risk (RR) and odds ratio (OR) with 95% confidence interval. Acute Kidney Injury Network stages were sorted out in 2 groups: no AKI (patients without AKIN criteria) and
Table 3
3. Statistical analysis
Description of the studied population
Statistically significant association. AKI: includes patients with AKI 1, 2, or 3 stages.
Renal injury study in critical ill patients Table 4
209
AKI associated risk factors—multivariable analysis
Risk factor
OR
95% CI
Heart failure Sepsis Vasopressor use Age
3.01 5.29 1.89 1.02
1.59-5.67 3.36-8.3 1.26-2.83 1.01-1.03
R2 Nagelkerke: 0.27.
AKI patients (patients under AKI 1, 2, or 3 stage). Hypothesis tests used in the study were Chi-square test for categorical variables and Student's t test for continuous variables. A P value of .05 or less was considered statistically significant. The interaction and confusion between nephrotoxic medications and vasoppressors with sepsis were evaluated by stratified analysis and multivariate analysis. Age, SAPS II score, APACHE II score II, SOFA score, ICU stay, and hospital stay variables were analyzed by Kolmogorov-Smirnov test with Lilliefors correction and homocedasticity by Levene test. We compared mortality with Chi-square test and RR with CI of 95%. The multivariable model applied to the study to establish prognosis factors for AKI was logistic regression analysis. Variables found as statistically significantly different (P b .05) in the univariate analysis were selected for further analysis using Wald method. We used Hosmer-Lemeshow to assess the goodness of fit of the model and Nagelkerke R2 to assess the percentage of variations in the resulting variable explained by the model. No AKI and AKI stages 1, 2, and 3 were compared. Oneway analysis of variance with Bonferroni correction was used to compare age. Chi-square test was used to compare sex, sepsis presence, and mortality. Kruskal-Wallis was used to compare APACHE II, SAPS II and SOFA scores, ICU stay days, and total in-hospital days among the different AKI stages for more than 2 groups. A P value less than .05 was considered statistically significant. The statistical programs SPSS version 15.0 y EPI-INFO 6.0 (SPSS, Chicago, Ill) was used for result analysis.
Table 5
4. Results During the study, 830 patients were identified to fulfill the proposed entrance criteria. Of these, 794 (95.6%) were studied. For 21 patients (2.5%), the required information was not found suitable for the study (Fig. 1). Majority (53.4%) were men, and 46.6% were women, with an average (SD) age of 52.6 (19.6) years. The main causes for entrance to ICU were sepsis, abdominal surgery, and thoracic surgery (Table 2). SAPS II severity median was 27 (IQR, 17-42), with an expected mortality of 7.9%, APACHE II median was 13 (IQR, 8-19), with an expected mortality of 16.5%, and SOFA evaluation from the acute renal injury day had a median of 4 (IQR, 2-7). Previous significant antecedents of the studied population are shown in Table 2. Acute renal injury was detected in 39.8% (315 patients), of which 13.9% was sorted out in stage 1, 12% in stage 2, and 13.9% in stage 3, and 12.2% had end-stage renal failure with replacement renal therapy. For risk factors analysis, end-stage renal failure patients with replacement renal therapy were excluded (97). The use of some types of nephrotoxic medications was identified in 23.5%, vancomycin in 7.5%, aminoglycosides in 7.0%, and contrast medium in 12.5% of the studied population. Univariate analysis was used to no-AKI patients (382) with AKI patients (315). The following risk factors were found associated to AKI: hypertension, heart failure, diabetes mellitus, sepsis, use of aminoglycosides, use of vancomycin, use of contrast medium, and vasopressor use (Table 3). Statistically significant risk factors were analyzed using the multivariable analysis with multiple logical regression, finding the following as AKI associated: sepsis (OR, 5.29; 95% CI, 3.36-8.33), heart failure (OR, 3.01; 95% CI, 1.595.67), vasopressor use (OR, 1.89; 95% CI, 1.26-2.83), and age (β = 1.02; 95% CI, 1.01-1.03; Table 4). Renal replacement therapy was used in 12.4% (39 patients) with AKI, of which 69.2% were subjected to intermittent hemodialysis, and 30.8% to continuous slow therapy. From the entire group of patients with AKI, 67.9%
AKI patient comparisons
AKI stage
No AKI (n = 382)
Stage 1 (n = 110)
Stage 2 (n = 95)
Stage 3 (n = 110)
P
Baseline serum creatine (mg/dL), mean (SD) Peak serum creatine (mg/dL), mean (SD) Sex: masculine (%) Age (y), mean (SD) Sepsis (%) SAPS II, median (IQR) APACHE II, Median (IQR) SOFA, median (IQR) ICU stay days, mean (SD) Total in-hospital days, mean (SD) General mortality (%)
0.75 (0.4) 0.79 (0.4) 41.9 49.5 (18.9) 9.7 20 (12-28) 9 (5-14) 2 (1-4) 3.04 (3.01) 10.9 (8.7) 7.3
0.8 (0.5) 1.17 (0.9) 47.3 55.7 (18.8) 32.7 30 (19-41) 14 (9-19) 4 (2-7) 8.0 (9.7) 17.8 (16.1) 16.4
0.8 (0.4) 1.4 (0.9) 50.5 59.7 (19.9) 45.3 35 (25-53) 15 (10-22) 6 (3-9) 10.2 (14.9) 21.1 (23.5) 34.7
1.08 (1.15) 3.85 (3.3) 49.1 60.6 (17.8) 51.8 43 (31-54) 20 (14-26) 8 (5-12) 12.41 (11.9) 22.1 (16.7) 45.5
.373 b.0001 .278 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001
210
N.J. Fonseca Ruiz et al.
Fig. 4
Fig. 2
Length of stay in ICU according AKI stage.
(214) patients had totally recovered from renal function at the end of hospitalization, 22.9% (72 patients) did not recover renal function, and 3.5% (11 patients) had partial recovery (lowering their AKI stage but not normalizing their creatinine values). Information was not obtained from 5.7% (18 patients). The groups of patients were compared according to their AKI stage. Baseline characteristics are shown in Table 5. There are significant differences in age, presence of sepsis, and the severity expressed by APACHE II, SAPS II, and SOFA. There was a statistically significant different length of stay in ICU (Fig. 2) and total in-hospitalization (Fig. 3). Mechanical ventilation requirement was 22.8% for patients without AKI, 46.4% for patients with AKI stage 1, 62.1% for stage 2, and 69.1% for those in stage 3 (Fig. 4). Comparison of mechanical ventilation necessity between noAKI patients with AKI patients was statistically different (P b .0001). Days of mechanical ventilation averaged (SD) 4.58
Fig. 3
Length of stay in-hospital depending on AKI stage.
Mechanical ventilation use (percentage).
(7.94) days for patients without AKI, 9.31 (11.3) days for patients with AKI in stage 1, 11.6 (15.2) days for stage 2, and 12.7 (11.6) days for stage 3. Mortality analysis in ICU is shown in (Fig. 5). A significant difference was demonstrated statistically for mortality between patients with some degree of AKI and those without AKI (25.4% versus 5.0%; RR, 5.1; 95% CI, 3.17-8.23; P b .0001). Patients with AKI had higher hospital mortality rate than patients with AKI (32.1% versus 7.3%; RR, 5.96; 95% CI, 3.8-9.4; P b .0001). Hospital mortality is higher at higher AKIN stage (Table 5).
5. Discussion We developed a retrospective study under the new definition of AKI based on the new terminology, definition, and sorting of the AKIN group [12,13]. A cohort of critically ill hospitalized patients was characterized in ICU at UBC, a 200-bed community teaching hospital. A higher incidence of the disease was obtained (39.8%) in comparison with the previously reported articles based on RIFLE [14-20] classification. This is probably due to the higher sensitivity that the new definition provides for earlier diagnosis. Patients who previously would not be diagnosed with AKI were included. Although this is a retrospective work, we used both creatinine and urine output criteria because we gathered the data prospectively as part of routine clinical practice. In 42.5% of the patients, the diagnosis was done only for creatinine value; in 28.6%, only for alteration of urinary
Fig. 5
Intensive care unit and total in-hospital mortality.
Renal injury study in critical ill patients output; and in 28.9%, for alteration of both parameters. Ricci and colleagues [21] recently published a systematic review of 24 AKI studies; in only 12% of the analyzed population were creatinine and urine output criteria used together. It is important to use the urinary output criterion in the AKI definition because if we do not, it would alter the findings [18]. Although this article is not appropriate for RIFLE versus AKIN comparison and the design is only descriptive in relation to AKIN scale, most articles about epidemiology of AKI are made based only on RIFLE scale. Under RIFLE definition, ICU patients are reported to have a broad range of AKI incidence, between 10.8% and 86.9%. Our data are closer to the one described by Ostermann and Chang [17], who reported an incidence of 35%, but it is higher than the one reported by Cruz et al [18] (10.8%), and it is lower than the one reported by Ahlström et al [19] (52%) and Hoste et al [20] (67%). This may reflect the differences in the criteria used. Some used the criteria based only on changes in creatinine levels, others used the glomerular filtration rate, and still others added the changes in urinary expenditure. Besides, differences in times of data gathering and the population studied may have also had an impact on the differences. In the retrospective work of Barrantes et al [22], an incidence of 31.5% was reported, which is very close to the value found by our group. Our results with AKI stage 1 group agree with previous findings referring to the fact that small changes in creatinine levels have a significant weight over the patients' prognosis [8-10]. It has also been demonstrated that the development of AKI stage 1 is correlated to a higher number of stay in ICU, hospital stay, and a greater need for mechanical ventilation for longer periods. It has been found as well that ICU and intrahospital mortality was greater for the group of patients with a statistically significant difference. The fact of having included patients with just an increase in 0.3 gm/dL in creatinine level in stage 1 group, which had not been included in “R”(Risk) group based on “RIFLE” classification in one of our previous articles [23], did not vary the significance of the differences with respect to the group of patients without AKI. In that article, we found AKI in 38% of the patients. Acute kidney injury was classified as “Risk” in 29.9%, “Injury” in 27.2%, and “Failure” in 28.2% of patients. The data showed a stepwise increase in mortality with increasing AKI severity (Risk, 10%; Injury, 28%; Failure, 36.5%) [23]. The large retrospective analysis of the Australian and New Zealander Intensive Care Society database [24] compared RIFLE and AKIN on the first ICU day only. They found that prevalence and crude mortality were very similar between the 2 classification schemes. In the present study, all indicators—result, days of stay in ICU, days of hospital stay, days of mechanical ventilation, and mortality rate—increased significantly as AKI stage acuteness increased. However, it is important to take into account that age standards, measured gravity by APACHE II, SAPS II, and SOFA, and the percentage of patients with
211 sepsis increased with a rise in AKI stage, which may well explain the harmful result at higher stages. SAPS II score in AKI patients was 18 with a predicted mortality of 21%; APACHE II was 18 with a predicted mortality of 29.1%. The observed mortality in this population was 25.4%. We did not find an observed mortality higher than the one predicted by Barrantes et al [22]. Based on the use of the new definition and classification, and using a logistic multiple regressions for the risk factors that had been statistically significant in univariable analysis, we found that the main factors associated with presence of AKI in ICU are sepsis, heart failure, use of vasoconstrictors, and age, but further multivariable analysis leave heart failure, sepsis, and the use of vasoconstrictors as the main factors. Our results reinforce the importance of sepsis as a risk factor for AKI [25] appearance. In our study, 62.7% of the septic patients presented AKI. Sepsis increased 5 times the risk to present AKI in our population of general ICU patients. It called our attention that usage of nephrotoxic medication such aminoglycosides and vancomycin did not display statistically significant association with AKI in the multivariable analysis. This probably means that to prevent the appearance or progression of AKI, an appropriate treatment than to avoid the use of nefrotoxic antibiotics is more important. Because we do not have a great number of patients and the study was not designed for this purpose, we cannot know if the use of these 2 nefrotoxics can increase the risk of renal failure in septic patients. Definition and stratification of RIFLE for renal failure were published more than 3 years ago and have been tried on several types of populations [26-31]. In our population, we have found that the use of the definition and classification of AKIN helps to foretell the patient diagnosis, as others have shown using the “RIFLE” classification [32-34], and also confirmed what has been postulated by Barrantes et al [22] with AKIN definition. Replacement renal therapy was more commonly used in our study (12.4%) than Hoste et al [20] (4.1%), Ostermann and Chang [17] (7%), and less frequent than the study of Cruz et al [18] (30.3%), possibly denoting differences in the initial criteria of renal support. Hospital mortality of patients with AKI was 32.1%, similar to Ostermann and Chang [17] (36.1%), lower than Cruz et al [18] (49.5%), and lower than Hoste et al [20] (13.3%), reflecting differences in the severity of patients and possibly in therapies. Acute Kidney Injury Network criteria came about to improve some issues of RIFLE definition; for example, elimination of the glomerular filtration rate criterion, inclusion of patients with smaller changes in creatinine, and exclusion of easily reversible causes of azotemia. Although AKIN definition has limitations, it is imperative to recognize that no single definition will be perfect [35]. It is also important to consider the limitations of our study. It has a retrospective design and was carried out at a single center referent in our city for the management of patients with acute
212 or chronic renal failure who require therapy of renal support. Acute kidney injury definition in our studied population allowed us to predict prognosis, which must be further validated with new prospective studies and studies with more methodological weight.
References [1] Kellum JA, Levin N, Bouman C, Lameire N. Developing a consensus classification system for acute renal failure. Curr Opin Crit Care 2002; 8(6):509-14. [2] Uchino S, Kellum JA, Bellomo R, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA 2005;294(7): 813-8. [3] Liano F, Pascual J. Madrid Acute Renal Failure Study Group. Epidemiology of acute renal failure: a prospective, multicenter, community based study. Kidney Int 1996;50(3):811-8. [4] Brivet FG, Kleinknecht DJ, Loirat P. the French Study Group on Acute Renal Failure. Acute renal failure in intensive care units: causes, outcome, and prognostic factors of hospital mortality. A prospective, multicenter study. Crit Care Med 1996;24(2):192-8. [5] Bellomo R, Kellum JA, Ronco C. Acute renal failure: time for consensus. Intensive Care Med 2001;27(11):1685-8. [6] Kellum JA, Mehta RL, Ronco C. Acute Dialysis Quality Initiative (ADQI). Contrib Nephrol 2001;132:258-65. [7] Bellomo R, Ronco C, Kellum JA, Mehta RL, Paleusky 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 2004;8: R204-12. [8] Levy EM, Viscoli CM, Horwitz RI. The effect of acute renal failure on mortality. A cohort analysis. JAMA 1996;275(19):1489-94. [9] Lassnigg A, Schmidlin D, Mouhieddine 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): 1597-605. [10] Chertow GM, Burdick E, Honour M, Bonventre JV, Bates DW. Acute kidney injury, mortality, length of stay, and costs in hospitalized patients. J Am Soc Nephrol 2005;16(11):3149-70. [11] Kellum JA, Bellomo R, Ronco C. The concept of acute kidney injury and the RIFLE criteria. Contrib Nephrol 2007;156:10-6. [12] Mehta RL, Kellum JA, Shah SV, et al. the Acute Kidney Injury Network. Report of an initiative to improve outcomes in acute kidney injury. Critical Care 2007;11(2):R31. [13] Molitoris BA, Levin A, Warnock DG, Joannidis M, Mehta RL, et al, Acute Kidney Injury Network working group. Improving outcomes of acute kidney injury: report of an initiative. Nat Clin Pract Nephrol 2007;3(8):439-42. [14] Chertow GM, Levy EM, Hammermeister KE, Grover F, Daley J. Independent association between acute renal failure and mortality following cardiac surgery. Am J Med 1998;104(4):343-8. [15] Guerin C, Girard R, Selli JM, Perdrix JP, Ayzac L. Rhone-Alpes Area Study Group on Acute Renal Failure. Initial versus delayed acute renal failure in the intensive care unit. A multicenter prospective epidemiological study. Am J Respir Crit Care Med 2000;161(3 Pt 1): 872-9.
N.J. Fonseca Ruiz et al. [16] Liano F, Junco E, Pascual J, Madero R, Verde E. the Madrid Acute Renal Failure Study Group. The spectrum of acute renal failure in the intensive care unit compared with that seen in other settings. Kidney Int Suppl 1998;66:S16-24. [17] Ostermann M, Chang RW. Acute kidney injury in the intensive care unit according to RIFLE. Crit Care Med 2007;35(8):1837-43. [18] Cruz DN, Bolgan I, Perazella MA, Bonello M, de Cal M, North East Italian Prospective Hospital Renal Outcome Survey on Acute Kidney Injury (NEiPHROS-AKI) Investigators, et al. Targeting the problem with the RIFLE criteria. Clin J Am Soc Nephrol 2007;2(3):418-25. [19] Ahlström A, Kuitunen A, Peltonen S, et al. Comparison of 2 acute renal failure severity scores to general scoring systems in the critically ill. Am J Kidney Dis 2006;48(2):262-8. [20] Hoste EA, Clermont G, Kersten A, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Critical Care 2006;10(3):R73-83. [21] Ricci Z, Cruz D, Ronco C. The RIFLE criteria and mortality in acute kidney injury: a systematic review. Kidney Int 2008;73:538-46. [22] Barrantes F, Tian J, Vazquez R, Amoateng-Adjepong Y, Manthous CA. Acute kidney injury criteria predict outcomes of critically ill patients. Crit Care Med 2008;36(5):1397-403. [23] Fonseca N, Mesa A, Molina F, Montejo J, Cuesta D. Epidemiología de la falla renal aguda en pacientes críticos bajo la nueva definición. Cambian las cosas? Med Intensiva 2007;31(Supl 2):A395. [24] Bagshaw SM, George C, Bellomo R. ANZICS Database Management Committee: a comparison of the RIFLE and AKIN criteria for acute kidney injury in critically ill patients. Nephrol Dial Transplant 2008; 23:1569-74. [25] Schrier R, Wang W. Acute renal failure and sepsis. N Engl J Med 2004;351(2):159-69. [26] Jenq CC, Tsai MH, Tian YC, et al. RIFLE classification can predict short-term prognosis in critically ill cirrhotic patients. Intensive Care Med 2007;33(11):1921-30. [27] Askenazi DJ, Bunchman TE. Pediatric acute kidney injury: the use of the RIFLE criteria. Kidney Int 2007;71(10):963-4. [28] 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 2006;34(7):1913-7. [29] Bell M, Liljestam E, Granath F, Fryckstedt J, Ekbom A, Martting CR. Optimal follow-up time after continuous renal replacement therapy in actual renal failure patients stratified with the RIFLE criteria. Nephrol Dial Transplant 2005;20(2):354-60. [30] Maccariello E, Soares M, Valente C, et al. RIFLE classification in patients with acute kidney injury in need of renal replacement therapy. Intensive Care Med 2007;33(4):597-605. [31] Lopes JA, Jorge S, Neves FC, et al. An assessment of the RIFLE criteria for acute renal failure in severely burned patients. LetterNephrol Dial Transplant 2007;22(1):285. [32] Lopes JA, Jorge S, Resina C, et al. Prognostic utility of RIFLE for acute renal failure in patients with sepsis. Letter. Crit Care 2007; 11:408. [33] Abosaif NY, Tolba YA, Heap M, Russell J, El Nahas AM. The outcome of acute renal failure in the intensive care unit according to RIFLE: model applicability, sensitivity, and predictability. Am J Kidney Dis 2005;46(6):1038-48. [34] Kuitunen A, Vento A, Suojaranta-Ylinen R, Pettila V. Acute renal failure after cardiac surgery: evaluation of the RIFLE classification. Ann Thorac Surg 2006;81(2):542-6. [35] Cruz D, Ricci Z, Ronco C. RIFLE and AKIN—time for reappraisal. Crit Care 2009;13:211.
Renal Injury
Renal injury study in critical ill patients in accordance with the new definition given by the Acute Kidney Injury Network☆,☆☆ Nelson Javier Fonseca Ruiz MD, MSc a,b,⁎, Diana Paola Cuesta Castro MD, MSc c , Ana Milena Mesa Guerra MD c,d , Francisco Molina Saldarriaga MD, MSc a,b , Juan Diego Montejo Hernández MD e a
Intensive Care Unit, University Bolivariana Clinic, Medellin, Colombia Intensive Care Unit, CORBIC Institute, Medellin, Colombia c Pontificia Bolivariana University, Medellin, Colombia d Seedbed Investigation Medicine Faculty (SIFAM), Pontificia Bolivariana University, Medellin, Colombia e Pontificia Bolivariana University, Medellin Clinic, Colombia b
Keywords: Acute renal injury; Consensus definition; Epidemiology; Critical illness; Mortality
Abstract Objective: This research aims to apply the definition proposed by the Acute Kidney Injury Network (AKIN) research group to assess the incidence, risk factors, and outcomes in acute kidney injury (AKI) patients admitted at the intensive care unit (ICU). Design: This is a retrospective cohort study. Patients who were admitted to the ICU from January 1, 2003 to December 31, 2004 were studied. Interventions: Medical records of all patients were reviewed. Demographic information, diagnoses, risk factors for AKI, laboratory data, urinary output, frequency and days of exposure to mechanical ventilation, ICU and hospital stay, and outcomes were recorded. Measurements and Main Results: A total of 794 patients were studied. There were 39.8% of patients who presented AKI (stage 1: 13.9%, stage 2: 12%, stage 3: 13.9%). The variables that were associated with the presence of AKI in the multivariable analysis were as follows: sepsis (odds ratio [OR], 5.29; 95% confidence interval [CI], 3.36-8.33), heart failure (OR, 3.01; 95% CI, 1.59-5.67), vasopressor use (OR, 1.89; 95% CI, 1.26-2.83), and age (β = 1.02; 95% CI, 1.01-1.03). The mean hospital stay increased with renal commitment: patients without AKI, 10.9 days; AKIN stage 1, 17.8; AKIN stage 2, 21.1; and AKIN stage 3, 22.1 days (P b .0001). Mortality rate increased as more advanced the AKI stage was (no AKI, 7.3%; AKI 1, 16.4%; AKI 2, 34.7%; and AKIN 3, 45.5%; P b .0001).
☆
Financial support: Integrated Center for Research Development, Universidad Pontificia Bolivariana. The authors have not disclosed any potential conflicts of interest in relation to this manuscript. ⁎ Corresponding author. E-mail address: [email protected] (N.J. Fonseca Ruiz). ☆☆
0883-9441/$ – see front matter © 2011 Elsevier Inc. All rights reserved. doi:10.1016/j.jcrc.2010.06.011
Renal injury study in critical ill patients
207
Conclusions: All of the result indicators—stay days in ICU, hospital stay days, frequency and days of mechanical ventilation, and mortality—considerably increased with more acute AKI stage. The most important risk factor of AKI was the sepsis. © 2011 Elsevier Inc. All rights reserved.
1. Introduction Although the term “acute renal failure” was introduced since 1951, there has never been a standard biochemical definition and there has not been consensus as to the diagnosis criterion, clinical definition, or disease stages for it. This has resulted in the appearance of multiple definitions used for research. Kellum et al [1] have described more than 35 different definitions. The prevalence and mortality reported in patients of intensive care unit (ICU) due to the lack of uniformity on the definition is wide, ranging from 1% to 25% and 15% to 60%, respectively [2-4]. There has been little progress on research on preventive measurements and efficient treatments because there has not been a common language among clinicians and researchers that would allow the comparison of epidemiologic, therapeutic, and diagnosis variables. In recent years, several efforts have been made to achieve a unique definition [5,6], and in 2004, Bellomo et al [7] proposed a new definition for acute renal failure based on the stratification of the disease in several stages—risk, injury, failure, loss, and end-stage renal disease—so that pattern analysis would allow progress. This proposal also included entities like sepsis and acute respiratory distress syndrome. However, such a definition still named this entity “Acute Renal Failure,” decreasing the importance of the broad spectrum and extent the disease has since the initial disorder stage. It has been demonstrated that small elevations in serum creatinine levels increase the risk of complications and death. Levy et al [8] showed that a 25% increase on basal creatinine levels after the administration of contrast medium was associated with 5 times higher risk of intrahospital mortality. Lassnigg et al [9] published similar data in patients submitted to heart surgery. Also, Chertow et al [10] demonstrated that increments of just 0.3 mg/dL in serum creatinine level were related to cost, morbidity, and mortality increase in hospitalized patients. The stage called “Failure” also created confusion because it was named the same as the syndrome. In September 2005, a multidisciplinary group made up of several medical societies named Acute Kidney Injury Network (AKIN) gathered together in Amsterdam and defined “Acute Kidney Injury” (AKI) as any abrupt decrease (alteration in less than 48 hours) in renal function produced by any damage that could cause functional or structural changes in kidneys, setting aside the terms “Acute Renal Failure,” which refers to the final phases of the disease, and “Acute Tubular Necrosis,” which gave a physiopathologic predominant role to the ischemia so that it overlooked the
importance of the inflammatory phenomenon [11]. Besides, following the Risk, Injury, Failure, Loss, and End-stage Kidney (RIFLE) classification, they proposed stratification based on the urinary expenditure and changes the serum creatinine levels, naming the phases as stages 1, 2, and 3 of the initial phase and not “Risk,” “Injury,” “Failure” any more. Also, the categories “Loss” and “End-stage renal disease” disappeared because they were considered as results. This definition was issued in several medical journals, and it is proposed to be used to carry out clinical trials [12,13]. The aims of this study are to evaluate AKI incidence, to assess the risk factors associated with it, and to evaluate the relationship between hospital stay and hospital mortality with AKI presence and AKI stages in a group of general critically ill patients under the approach of the new definition proposed by the AKIN group.
2. Materials and methods A study of a retrospective cohort was carried out in patients treated in the ICU at University Bolivariana Clinic (UBC) in Medellín, Colombia, in the period from January 1, 2003 to December 31, 2004. The UBC is a third-level care center that has 200 hospitalization beds and 12 ICU beds. Patients older than 16 years who entered the ICU for any indication were included in the study. Patients referred from or to a different ICU were excluded due to the difficulties presented on data recollection. The information was gathered in a medical chart review where sociodemographic data as age and sex, clinical data as well as personal antecedents, and the reason for ICU entrance according to a diagnostic group list were recorded. Medical antecedents such as heart failure, hypertension, chronic obstructive pulmonary disease (COPD), tobacco smoking, coronary disease, diabetes, sepsis, vasopressor use, and the use of nephrotoxic medications were taken into account. Definitions were not standardized, but just recorded as they were found if they were going to be registered in the clinical profile. Other data considered were the severity of the condition measured by Acute Physiology and Chronic Health Evaluation scoring system (APACHE) version II and Simplified Acute Physiology Score (SAPS) version II, the severity of the clinical condition at the beginning of the renal injury measured by Sequential Organ Failure Assessment score (SOFA), the use of vasopressors, the starting and ending dates of the ventilatory mechanical support, ICU and
208
N.J. Fonseca Ruiz et al.
Table 1
Classification/staging system for AKI
Stage Serum creatinine criteria
Urine output criteria
1
Less than 0.5 mL/kg per hour for more than 6 h
2
3
Increase in serum creatinine ≥0.3 mg/dL (≥26.4 μmol/L) or increase to more ≥150% to 200% (1.5- to 2-fold) from baseline Increase in serum creatinine to more than 200% to 300% (N2- to 3-fold) from baseline Increase in serum creatinine to more than 300% (N3-fold) from baseline (or serum creatinine ≥4.0 mg/dL [≥354 μmol/L] with an acute increase of at least 0.5 mg/dL [44 μmol/L])
Less than 0.5 mL/kg per hour for more than 12 h Less than 0.3 mL/kg per hour for 24 h or anuria for 12 h
From Mehta et al [12].
hospitalization entrance and discharge dates, and final clinical result. To rank the patients according to the maximum AKIN category, the urine output and the daily changes in the serum creatinine level were assessed. Acute kidney injury diagnosis was analyzed in all the patients from entrance to ICU until exit from ICU based on AKI definition proposed by AKIN group, that is, an abrupt (within 48 hours) reduction in kidney function currently defined as an absolute increase in serum creatinine of more than or equal to 0.3 mg/dL, a percentage increase in serum creatinine of more than or equal to 50% (1.5-fold from baseline), or a reduction in urine output (documented oliguria of less than 0.5 mL/kg per hour for more than 6 hours). To sort out the patients in the different stages of AKI, all of those who had previously received any way of chronic renal support therapy (dialysis) before entrance to ICU were excluded. Patients were stratified according to the definitions of the AKIN group using both criteria: creatinine level increase and urinary output (Table 1).
Table 2
Sex Masculine (%) Feminine (%) Age (y), mean (SD) Personal history Hypertension (%) COPD (%) Tobacco use (%) Diabetes mellitus (%) Coronary artery disease (%) Heart failure (%) Other (%) Diagnostic group Sepsis (%) Abdominal surgery (%) Thoracic surgery (%) Lung disease (%) Cardiovascular disease (%) Neurosurgery (%) Trauma Neurology Obstetrics Gastrointestinal Other SAPS II, median (IQR) APACHE II, median (IQR) SOFA, median (IQR)
Mean and median of continuous variables were calculated with their respective standard deviation or interquartile range (IQR) depending on the case. Categorical variables were
Study flow chart.
44.6 15.8 27.1 18.2 8.5 9.6 5.1 19.6 15 14.6 10.3 9.6 8.6 4.2 4.0 2.8 2.8 8.5 27 (17-42) 13 (8-19) 4 (2-7)
AKI associated risk factors—univariate analysis
Risk factor
OR
95% CI
P
Sex: masculine sex Personal history Hypertension a COPD Tobacco use Diabetes mellitus a Coronary artery disease Heart failure a Sepsis a Nephrotoxic medications Aminoglycoside use a Vancomycin use a Contrast medium a Vasopressor use a
1.33
0.97-1.81
.064
1.90 1.50 1.22 1.82 1.57 3.14 7.08
1.39-2.59 1.00-2.25 0.87-1.70 1.21-2.73 0.90-2.75 1.74-5.65 4.72-10. 63
.0001 .063 .28 .005 .143 .0001 .0001
2.20 3.60 1.85 3.89
1.21-4.03 1.91-6.79 1.18-2.93 2.76-5.48
.013 .0001 .010 .0001
a
Fig. 1
53.4 46.6 52.6 (19.5)
described by means of proportions. A univariate analysis was used to establish a relationship between risk factors (sex, personal history, sepsis, nephrotoxic drugs, and vasopressor use) and AKI. Whenever appropriate, results are presented as relative risk (RR) and odds ratio (OR) with 95% confidence interval. Acute Kidney Injury Network stages were sorted out in 2 groups: no AKI (patients without AKIN criteria) and
Table 3
3. Statistical analysis
Description of the studied population
Statistically significant association. AKI: includes patients with AKI 1, 2, or 3 stages.
Renal injury study in critical ill patients Table 4
209
AKI associated risk factors—multivariable analysis
Risk factor
OR
95% CI
Heart failure Sepsis Vasopressor use Age
3.01 5.29 1.89 1.02
1.59-5.67 3.36-8.3 1.26-2.83 1.01-1.03
R2 Nagelkerke: 0.27.
AKI patients (patients under AKI 1, 2, or 3 stage). Hypothesis tests used in the study were Chi-square test for categorical variables and Student's t test for continuous variables. A P value of .05 or less was considered statistically significant. The interaction and confusion between nephrotoxic medications and vasoppressors with sepsis were evaluated by stratified analysis and multivariate analysis. Age, SAPS II score, APACHE II score II, SOFA score, ICU stay, and hospital stay variables were analyzed by Kolmogorov-Smirnov test with Lilliefors correction and homocedasticity by Levene test. We compared mortality with Chi-square test and RR with CI of 95%. The multivariable model applied to the study to establish prognosis factors for AKI was logistic regression analysis. Variables found as statistically significantly different (P b .05) in the univariate analysis were selected for further analysis using Wald method. We used Hosmer-Lemeshow to assess the goodness of fit of the model and Nagelkerke R2 to assess the percentage of variations in the resulting variable explained by the model. No AKI and AKI stages 1, 2, and 3 were compared. Oneway analysis of variance with Bonferroni correction was used to compare age. Chi-square test was used to compare sex, sepsis presence, and mortality. Kruskal-Wallis was used to compare APACHE II, SAPS II and SOFA scores, ICU stay days, and total in-hospital days among the different AKI stages for more than 2 groups. A P value less than .05 was considered statistically significant. The statistical programs SPSS version 15.0 y EPI-INFO 6.0 (SPSS, Chicago, Ill) was used for result analysis.
Table 5
4. Results During the study, 830 patients were identified to fulfill the proposed entrance criteria. Of these, 794 (95.6%) were studied. For 21 patients (2.5%), the required information was not found suitable for the study (Fig. 1). Majority (53.4%) were men, and 46.6% were women, with an average (SD) age of 52.6 (19.6) years. The main causes for entrance to ICU were sepsis, abdominal surgery, and thoracic surgery (Table 2). SAPS II severity median was 27 (IQR, 17-42), with an expected mortality of 7.9%, APACHE II median was 13 (IQR, 8-19), with an expected mortality of 16.5%, and SOFA evaluation from the acute renal injury day had a median of 4 (IQR, 2-7). Previous significant antecedents of the studied population are shown in Table 2. Acute renal injury was detected in 39.8% (315 patients), of which 13.9% was sorted out in stage 1, 12% in stage 2, and 13.9% in stage 3, and 12.2% had end-stage renal failure with replacement renal therapy. For risk factors analysis, end-stage renal failure patients with replacement renal therapy were excluded (97). The use of some types of nephrotoxic medications was identified in 23.5%, vancomycin in 7.5%, aminoglycosides in 7.0%, and contrast medium in 12.5% of the studied population. Univariate analysis was used to no-AKI patients (382) with AKI patients (315). The following risk factors were found associated to AKI: hypertension, heart failure, diabetes mellitus, sepsis, use of aminoglycosides, use of vancomycin, use of contrast medium, and vasopressor use (Table 3). Statistically significant risk factors were analyzed using the multivariable analysis with multiple logical regression, finding the following as AKI associated: sepsis (OR, 5.29; 95% CI, 3.36-8.33), heart failure (OR, 3.01; 95% CI, 1.595.67), vasopressor use (OR, 1.89; 95% CI, 1.26-2.83), and age (β = 1.02; 95% CI, 1.01-1.03; Table 4). Renal replacement therapy was used in 12.4% (39 patients) with AKI, of which 69.2% were subjected to intermittent hemodialysis, and 30.8% to continuous slow therapy. From the entire group of patients with AKI, 67.9%
AKI patient comparisons
AKI stage
No AKI (n = 382)
Stage 1 (n = 110)
Stage 2 (n = 95)
Stage 3 (n = 110)
P
Baseline serum creatine (mg/dL), mean (SD) Peak serum creatine (mg/dL), mean (SD) Sex: masculine (%) Age (y), mean (SD) Sepsis (%) SAPS II, median (IQR) APACHE II, Median (IQR) SOFA, median (IQR) ICU stay days, mean (SD) Total in-hospital days, mean (SD) General mortality (%)
0.75 (0.4) 0.79 (0.4) 41.9 49.5 (18.9) 9.7 20 (12-28) 9 (5-14) 2 (1-4) 3.04 (3.01) 10.9 (8.7) 7.3
0.8 (0.5) 1.17 (0.9) 47.3 55.7 (18.8) 32.7 30 (19-41) 14 (9-19) 4 (2-7) 8.0 (9.7) 17.8 (16.1) 16.4
0.8 (0.4) 1.4 (0.9) 50.5 59.7 (19.9) 45.3 35 (25-53) 15 (10-22) 6 (3-9) 10.2 (14.9) 21.1 (23.5) 34.7
1.08 (1.15) 3.85 (3.3) 49.1 60.6 (17.8) 51.8 43 (31-54) 20 (14-26) 8 (5-12) 12.41 (11.9) 22.1 (16.7) 45.5
.373 b.0001 .278 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001 b.0001
210
N.J. Fonseca Ruiz et al.
Fig. 4
Fig. 2
Length of stay in ICU according AKI stage.
(214) patients had totally recovered from renal function at the end of hospitalization, 22.9% (72 patients) did not recover renal function, and 3.5% (11 patients) had partial recovery (lowering their AKI stage but not normalizing their creatinine values). Information was not obtained from 5.7% (18 patients). The groups of patients were compared according to their AKI stage. Baseline characteristics are shown in Table 5. There are significant differences in age, presence of sepsis, and the severity expressed by APACHE II, SAPS II, and SOFA. There was a statistically significant different length of stay in ICU (Fig. 2) and total in-hospitalization (Fig. 3). Mechanical ventilation requirement was 22.8% for patients without AKI, 46.4% for patients with AKI stage 1, 62.1% for stage 2, and 69.1% for those in stage 3 (Fig. 4). Comparison of mechanical ventilation necessity between noAKI patients with AKI patients was statistically different (P b .0001). Days of mechanical ventilation averaged (SD) 4.58
Fig. 3
Length of stay in-hospital depending on AKI stage.
Mechanical ventilation use (percentage).
(7.94) days for patients without AKI, 9.31 (11.3) days for patients with AKI in stage 1, 11.6 (15.2) days for stage 2, and 12.7 (11.6) days for stage 3. Mortality analysis in ICU is shown in (Fig. 5). A significant difference was demonstrated statistically for mortality between patients with some degree of AKI and those without AKI (25.4% versus 5.0%; RR, 5.1; 95% CI, 3.17-8.23; P b .0001). Patients with AKI had higher hospital mortality rate than patients with AKI (32.1% versus 7.3%; RR, 5.96; 95% CI, 3.8-9.4; P b .0001). Hospital mortality is higher at higher AKIN stage (Table 5).
5. Discussion We developed a retrospective study under the new definition of AKI based on the new terminology, definition, and sorting of the AKIN group [12,13]. A cohort of critically ill hospitalized patients was characterized in ICU at UBC, a 200-bed community teaching hospital. A higher incidence of the disease was obtained (39.8%) in comparison with the previously reported articles based on RIFLE [14-20] classification. This is probably due to the higher sensitivity that the new definition provides for earlier diagnosis. Patients who previously would not be diagnosed with AKI were included. Although this is a retrospective work, we used both creatinine and urine output criteria because we gathered the data prospectively as part of routine clinical practice. In 42.5% of the patients, the diagnosis was done only for creatinine value; in 28.6%, only for alteration of urinary
Fig. 5
Intensive care unit and total in-hospital mortality.
Renal injury study in critical ill patients output; and in 28.9%, for alteration of both parameters. Ricci and colleagues [21] recently published a systematic review of 24 AKI studies; in only 12% of the analyzed population were creatinine and urine output criteria used together. It is important to use the urinary output criterion in the AKI definition because if we do not, it would alter the findings [18]. Although this article is not appropriate for RIFLE versus AKIN comparison and the design is only descriptive in relation to AKIN scale, most articles about epidemiology of AKI are made based only on RIFLE scale. Under RIFLE definition, ICU patients are reported to have a broad range of AKI incidence, between 10.8% and 86.9%. Our data are closer to the one described by Ostermann and Chang [17], who reported an incidence of 35%, but it is higher than the one reported by Cruz et al [18] (10.8%), and it is lower than the one reported by Ahlström et al [19] (52%) and Hoste et al [20] (67%). This may reflect the differences in the criteria used. Some used the criteria based only on changes in creatinine levels, others used the glomerular filtration rate, and still others added the changes in urinary expenditure. Besides, differences in times of data gathering and the population studied may have also had an impact on the differences. In the retrospective work of Barrantes et al [22], an incidence of 31.5% was reported, which is very close to the value found by our group. Our results with AKI stage 1 group agree with previous findings referring to the fact that small changes in creatinine levels have a significant weight over the patients' prognosis [8-10]. It has also been demonstrated that the development of AKI stage 1 is correlated to a higher number of stay in ICU, hospital stay, and a greater need for mechanical ventilation for longer periods. It has been found as well that ICU and intrahospital mortality was greater for the group of patients with a statistically significant difference. The fact of having included patients with just an increase in 0.3 gm/dL in creatinine level in stage 1 group, which had not been included in “R”(Risk) group based on “RIFLE” classification in one of our previous articles [23], did not vary the significance of the differences with respect to the group of patients without AKI. In that article, we found AKI in 38% of the patients. Acute kidney injury was classified as “Risk” in 29.9%, “Injury” in 27.2%, and “Failure” in 28.2% of patients. The data showed a stepwise increase in mortality with increasing AKI severity (Risk, 10%; Injury, 28%; Failure, 36.5%) [23]. The large retrospective analysis of the Australian and New Zealander Intensive Care Society database [24] compared RIFLE and AKIN on the first ICU day only. They found that prevalence and crude mortality were very similar between the 2 classification schemes. In the present study, all indicators—result, days of stay in ICU, days of hospital stay, days of mechanical ventilation, and mortality rate—increased significantly as AKI stage acuteness increased. However, it is important to take into account that age standards, measured gravity by APACHE II, SAPS II, and SOFA, and the percentage of patients with
211 sepsis increased with a rise in AKI stage, which may well explain the harmful result at higher stages. SAPS II score in AKI patients was 18 with a predicted mortality of 21%; APACHE II was 18 with a predicted mortality of 29.1%. The observed mortality in this population was 25.4%. We did not find an observed mortality higher than the one predicted by Barrantes et al [22]. Based on the use of the new definition and classification, and using a logistic multiple regressions for the risk factors that had been statistically significant in univariable analysis, we found that the main factors associated with presence of AKI in ICU are sepsis, heart failure, use of vasoconstrictors, and age, but further multivariable analysis leave heart failure, sepsis, and the use of vasoconstrictors as the main factors. Our results reinforce the importance of sepsis as a risk factor for AKI [25] appearance. In our study, 62.7% of the septic patients presented AKI. Sepsis increased 5 times the risk to present AKI in our population of general ICU patients. It called our attention that usage of nephrotoxic medication such aminoglycosides and vancomycin did not display statistically significant association with AKI in the multivariable analysis. This probably means that to prevent the appearance or progression of AKI, an appropriate treatment than to avoid the use of nefrotoxic antibiotics is more important. Because we do not have a great number of patients and the study was not designed for this purpose, we cannot know if the use of these 2 nefrotoxics can increase the risk of renal failure in septic patients. Definition and stratification of RIFLE for renal failure were published more than 3 years ago and have been tried on several types of populations [26-31]. In our population, we have found that the use of the definition and classification of AKIN helps to foretell the patient diagnosis, as others have shown using the “RIFLE” classification [32-34], and also confirmed what has been postulated by Barrantes et al [22] with AKIN definition. Replacement renal therapy was more commonly used in our study (12.4%) than Hoste et al [20] (4.1%), Ostermann and Chang [17] (7%), and less frequent than the study of Cruz et al [18] (30.3%), possibly denoting differences in the initial criteria of renal support. Hospital mortality of patients with AKI was 32.1%, similar to Ostermann and Chang [17] (36.1%), lower than Cruz et al [18] (49.5%), and lower than Hoste et al [20] (13.3%), reflecting differences in the severity of patients and possibly in therapies. Acute Kidney Injury Network criteria came about to improve some issues of RIFLE definition; for example, elimination of the glomerular filtration rate criterion, inclusion of patients with smaller changes in creatinine, and exclusion of easily reversible causes of azotemia. Although AKIN definition has limitations, it is imperative to recognize that no single definition will be perfect [35]. It is also important to consider the limitations of our study. It has a retrospective design and was carried out at a single center referent in our city for the management of patients with acute
212 or chronic renal failure who require therapy of renal support. Acute kidney injury definition in our studied population allowed us to predict prognosis, which must be further validated with new prospective studies and studies with more methodological weight.
References [1] Kellum JA, Levin N, Bouman C, Lameire N. Developing a consensus classification system for acute renal failure. Curr Opin Crit Care 2002; 8(6):509-14. [2] Uchino S, Kellum JA, Bellomo R, et al. Acute renal failure in critically ill patients: a multinational, multicenter study. JAMA 2005;294(7): 813-8. [3] Liano F, Pascual J. Madrid Acute Renal Failure Study Group. Epidemiology of acute renal failure: a prospective, multicenter, community based study. Kidney Int 1996;50(3):811-8. [4] Brivet FG, Kleinknecht DJ, Loirat P. the French Study Group on Acute Renal Failure. Acute renal failure in intensive care units: causes, outcome, and prognostic factors of hospital mortality. A prospective, multicenter study. Crit Care Med 1996;24(2):192-8. [5] Bellomo R, Kellum JA, Ronco C. Acute renal failure: time for consensus. Intensive Care Med 2001;27(11):1685-8. [6] Kellum JA, Mehta RL, Ronco C. Acute Dialysis Quality Initiative (ADQI). Contrib Nephrol 2001;132:258-65. [7] Bellomo R, Ronco C, Kellum JA, Mehta RL, Paleusky 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 2004;8: R204-12. [8] Levy EM, Viscoli CM, Horwitz RI. The effect of acute renal failure on mortality. A cohort analysis. JAMA 1996;275(19):1489-94. [9] Lassnigg A, Schmidlin D, Mouhieddine 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): 1597-605. [10] Chertow GM, Burdick E, Honour M, Bonventre JV, Bates DW. Acute kidney injury, mortality, length of stay, and costs in hospitalized patients. J Am Soc Nephrol 2005;16(11):3149-70. [11] Kellum JA, Bellomo R, Ronco C. The concept of acute kidney injury and the RIFLE criteria. Contrib Nephrol 2007;156:10-6. [12] Mehta RL, Kellum JA, Shah SV, et al. the Acute Kidney Injury Network. Report of an initiative to improve outcomes in acute kidney injury. Critical Care 2007;11(2):R31. [13] Molitoris BA, Levin A, Warnock DG, Joannidis M, Mehta RL, et al, Acute Kidney Injury Network working group. Improving outcomes of acute kidney injury: report of an initiative. Nat Clin Pract Nephrol 2007;3(8):439-42. [14] Chertow GM, Levy EM, Hammermeister KE, Grover F, Daley J. Independent association between acute renal failure and mortality following cardiac surgery. Am J Med 1998;104(4):343-8. [15] Guerin C, Girard R, Selli JM, Perdrix JP, Ayzac L. Rhone-Alpes Area Study Group on Acute Renal Failure. Initial versus delayed acute renal failure in the intensive care unit. A multicenter prospective epidemiological study. Am J Respir Crit Care Med 2000;161(3 Pt 1): 872-9.
N.J. Fonseca Ruiz et al. [16] Liano F, Junco E, Pascual J, Madero R, Verde E. the Madrid Acute Renal Failure Study Group. The spectrum of acute renal failure in the intensive care unit compared with that seen in other settings. Kidney Int Suppl 1998;66:S16-24. [17] Ostermann M, Chang RW. Acute kidney injury in the intensive care unit according to RIFLE. Crit Care Med 2007;35(8):1837-43. [18] Cruz DN, Bolgan I, Perazella MA, Bonello M, de Cal M, North East Italian Prospective Hospital Renal Outcome Survey on Acute Kidney Injury (NEiPHROS-AKI) Investigators, et al. Targeting the problem with the RIFLE criteria. Clin J Am Soc Nephrol 2007;2(3):418-25. [19] Ahlström A, Kuitunen A, Peltonen S, et al. Comparison of 2 acute renal failure severity scores to general scoring systems in the critically ill. Am J Kidney Dis 2006;48(2):262-8. [20] Hoste EA, Clermont G, Kersten A, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Critical Care 2006;10(3):R73-83. [21] Ricci Z, Cruz D, Ronco C. The RIFLE criteria and mortality in acute kidney injury: a systematic review. Kidney Int 2008;73:538-46. [22] Barrantes F, Tian J, Vazquez R, Amoateng-Adjepong Y, Manthous CA. Acute kidney injury criteria predict outcomes of critically ill patients. Crit Care Med 2008;36(5):1397-403. [23] Fonseca N, Mesa A, Molina F, Montejo J, Cuesta D. Epidemiología de la falla renal aguda en pacientes críticos bajo la nueva definición. Cambian las cosas? Med Intensiva 2007;31(Supl 2):A395. [24] Bagshaw SM, George C, Bellomo R. ANZICS Database Management Committee: a comparison of the RIFLE and AKIN criteria for acute kidney injury in critically ill patients. Nephrol Dial Transplant 2008; 23:1569-74. [25] Schrier R, Wang W. Acute renal failure and sepsis. N Engl J Med 2004;351(2):159-69. [26] Jenq CC, Tsai MH, Tian YC, et al. RIFLE classification can predict short-term prognosis in critically ill cirrhotic patients. Intensive Care Med 2007;33(11):1921-30. [27] Askenazi DJ, Bunchman TE. Pediatric acute kidney injury: the use of the RIFLE criteria. Kidney Int 2007;71(10):963-4. [28] 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 2006;34(7):1913-7. [29] Bell M, Liljestam E, Granath F, Fryckstedt J, Ekbom A, Martting CR. Optimal follow-up time after continuous renal replacement therapy in actual renal failure patients stratified with the RIFLE criteria. Nephrol Dial Transplant 2005;20(2):354-60. [30] Maccariello E, Soares M, Valente C, et al. RIFLE classification in patients with acute kidney injury in need of renal replacement therapy. Intensive Care Med 2007;33(4):597-605. [31] Lopes JA, Jorge S, Neves FC, et al. An assessment of the RIFLE criteria for acute renal failure in severely burned patients. LetterNephrol Dial Transplant 2007;22(1):285. [32] Lopes JA, Jorge S, Resina C, et al. Prognostic utility of RIFLE for acute renal failure in patients with sepsis. Letter. Crit Care 2007; 11:408. [33] Abosaif NY, Tolba YA, Heap M, Russell J, El Nahas AM. The outcome of acute renal failure in the intensive care unit according to RIFLE: model applicability, sensitivity, and predictability. Am J Kidney Dis 2005;46(6):1038-48. [34] Kuitunen A, Vento A, Suojaranta-Ylinen R, Pettila V. Acute renal failure after cardiac surgery: evaluation of the RIFLE classification. Ann Thorac Surg 2006;81(2):542-6. [35] Cruz D, Ricci Z, Ronco C. RIFLE and AKIN—time for reappraisal. Crit Care 2009;13:211.