Acute kidney injury in an internal medicine ward in a

EJINME-02606; No of Pages 4 European Journal of Internal Medicine xxx (2013) xxx–xxx

Contents lists available at ScienceDirect

European Journal of Internal Medicine journal homepage: www.elsevier.com/locate/ejim

Original article

Acute kidney injury in an internal medicine ward in a Portuguese quaternary hospital Marta Neves a,⁎, Paula Fidalgo b, Carla Gonçalves c, Sara Leitão c, Rui Marques Santos c, Armando Carvalho c, José Manuel Nascimento Costa c a b c

Department of Nephrology, Centro Hospitalar e Universitário de Coimbra — Hospitais da Universidade de Coimbra, Portugal Instituto Português de Oncologia de Coimbra Francisco Gentil, E.P.E., Coimbra, Portugal Department of Internal Medicine, Centro Hospitalar e Universitário de Coimbra — Hospitais da Universidade de Coimbra, Portugal

a r t i c l e

i n f o

Article history: Received 27 October 2012 Received in revised form 10 January 2013 Accepted 13 September 2013 Available online xxxx Keywords: Acute kidney injury Classification Correlation

a b s t r a c t Background: The term acute kidney injury (AKI) was proposed to reflect the wide spectrum of traditional acute renal failure. RIFLE classification stratifies AKI into three classes of severity and two classes of outcome. AKIN classification proposes an improvement regarding RIFLE in the stratification of AKI, while recently published KDIGO guidelines comprise characteristics of both RIFLE and AKIN. There are no published studies on the utility and measure of agreement between classifications in patients admitted to internal medicine wards. Methods: Prospective study undertaken in two internal medicine wards in a Portuguese hospital. Patients admitted for a minimum of 72 h, with a diagnosis of AKI or acute-on-chronic kidney disease at admission or during hospitalisation, were included. RIFLE, AKIN and KDIGO criteria were applied for identification of AKI and stratification into risk groups. Results: Sixty-nine patients were included, with a mean age of 79.7 ± 10.0 years and mean GFR of 21.7 ± 8.8 mL/min/1.73 m2. Hypovolaemia due to dehydration was the main cause of AKI (53.6%) and, thereby, RIFLE classification identified a higher number of patients as having AKI, compared to AKIN (94.2% vs. 84.1%). Most patients (69.6%) recovered to their baseline renal function, however fifteen patients (21.7%) died, 53.3% presenting more severe kidney disease. Conclusions: Our results demonstrate good concordance and correlation between RIFLE, AKIN and KDIGO criteria for the diagnosis of AKI (p b 0.001 at initial and final assessment). The authors support the need for further improvement of the classification, ultimately through the use of new biomarkers capable of earlier identification of patients at risk. © 2013 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved.

1. Introduction Acute renal failure is a complex clinical condition, traditionally classified as pre-renal, renal or post-renal, which may occur in patients with a previously normal renal function or with pre-existing renal disease. It determines a fast and typically reversible decline of glomerular filtration rate (GFR) [1]. There is, however, no consensus on a quantitative definition of this failure, which has contributed to a great heterogeneity of data in several studies performed in this area [2]. The term acute kidney injury (AKI) was proposed by a group of nephrologists and intensivists, who integrated the Acute Kidney Injury Network (AKIN) group, in order to better reflect the broad spectrum of traditional acute renal failure [3] and the term is now recognized as the preferred nomenclature for this complex clinical syndrome [4]. ⁎ Corresponding author at: Centro Hospitalar e Universitário de Coimbra — Hospitais da Universidade de Coimbra, Praceta Professor Mota Pinto, 3000-075 Coimbra, Portugal. Tel.: +351 965352588. E-mail address: [email protected] (M. Neves).

The Acute Dialysis Quality Initiative (ADQI) group published in 2004 the RIFLE classification, thereby stratifying AKI according to two criteria: (1) the decrease in glomerular filtration rate or increase in serum creatinine, in function of each patient's baseline value, and (2) the decrease in urine output per kilogramme of body weight over a pre-determined period of time [5]. The stratification of AKI for each patient should be based on the least favourable criteria. The classification contemplates three classes of diagnosis and severity: the patient at risk (R), the patient presenting with acute renal injury (I) and the patient with acute renal failure (F). It also describes two classes of prognosis: lack of recovery of renal function over a four week period (L) and end-stage renal disease if there is no recovery of renal function over a period longer than three months (E) (Table 1). In 2007, a modification of the RIFLE criteria was proposed by the AKIN group [3], with the intention of improving diagnostic accuracy. This new classification stratifies AKI into three stages, from 1 to 3, corresponding to the first three groups in the RIFLE classification. The reference to change in GFR is not included in the AKIN classification

0953-6205/$ – see front matter © 2013 European Federation of Internal Medicine. Published by Elsevier B.V. All rights reserved. http://dx.doi.org/10.1016/j.ejim.2013.09.007

Please cite this article as: Neves M, et al, Acute kidney injury in an internal medicine ward in a Portuguese quaternary hospital, Eur J Intern Med (2013), http://dx.doi.org/10.1016/j.ejim.2013.09.007

2

M. Neves et al. / European Journal of Internal Medicine xxx (2013) xxx–xxx

Table 1 RIFLE classification. Stratification according to the RIFLE criteria for the classification of AKI [5] Class

Creatinine or GFR criteria

Urinary output criteria

Risk

Increase in ≥1.5 times of baseline serum creatinine or reduction in ≥25% of GFR Increase in ≥2 times of baseline serum creatinine or reduction in ≥50% of GFR Increase in ≥3 times of baseline serum creatinine (or serum creatinine N 4 mg/dL (353.6 μmol/L)) or reduction in ≥75% of GFR Complete loss of kidney function for more than 4 weeks Complete loss of kidney function for more than 3 months

b0.5 mL/kg/h during ≥6 h b0.5 mL/kg/h during ≥12 h b0.3 mL/kg/h during ≥24 h or anuria for ≥12 h

Injury Failure

Loss End-stage renal disease

AKI — acute kidney injury, GFR — glomerular filtration rate. The RIFLE classification has some peculiarities: the change must occur in one to seven days; it must be maintained for more than 24 h and the baseline serum creatinine of the patient must be known. In the absence of data, the estimated creatinine value is calculated corresponding to a baseline GFR of 75 mL/min/1.73 m2 taken from the Modified Diet in Renal Disease (MDRD) formula.

system, since the change in serum creatinine does not correlate with the percentage decrease in GFR cited in the RIFLE classification [6]. Apart from this, the differences with regard to RIFLE criteria emerge essentially in Stage 1. They advocate a more restricted range of time (up to 48 h) for the development of renal dysfunction and consider an increase of 0.3 mg/dL (26.5 μmol/L) in serum creatinine level from baseline as a diagnostic criterion. This new finding is supported by epidemiological studies that showed an 80% increase in overall mortality in the presence of mild to moderate changes in creatinine, in particular from 0.3 to 0.5 mg/dL (26.5 to 44.2 μmol/L) [7], although the explanations for this are not entirely clear. The classification of patients into the various groups is carried out only after the hydration status has been optimized and after excluding obstructive causes for renal dysfunction. The need to institute renal replacement therapy immediately stratifies patients into Stage 3. The AKIN classification does not consider medium and long term consequences into the initial stratification, as happens with the RIFLE classification. In March 2012, the Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group published the KDIGO clinical practice guideline for acute kidney injury [8] with the purpose of creating evidence-based recommendations for the assessment and management of patients with AKI. With regard to diagnosis or staging of AKI, the Work Group chose to provide ungraded statements which stemmed from unanimous consensus of the Work Group. They defined AKI as any of the following: (1) increase in serum creatinine by ≥0.3 mg/dL (≥26.5 μmol/L) within 48 h (retained from the AKIN definition), (2) increase in serum creatinine to ≥1.5 times from baseline, which is known or presumed to have occurred within the prior 7 days (retained from the RIFLE criteria), or (3) urine volume b 0.5 mL/kg/h for 6 h. The criteria used for staging severity of AKI are the same as those used in the AKIN classification system and patients should be staged according to the criteria that bestow upon them the most severe stage of injury: Stage 1 as an increase in serum creatinine from 1.5 to 1.9 times baseline or an increase of ≥0.3 mg/dL (≥26.5 μmol/L) from baseline or urinary output b 0.5 mL/kg/h for 6 to 12 h; Stage 2 as an increase in serum creatinine from 2.0 to 2.9 times baseline or urinary output b0.5 mL/kg/h for ≥12 h; and Stage 3 as an increase in serum creatinine from ≥3.0 times baseline or to ≥4.0 mg/dL (≥353.6 μmol/L) (provided that they first satisfy the criteria for the definition of AKI) or need for renal replacement therapy or urinary output b0.3 mL/kg/h for ≥24 h or anuria for ≥12 h. Since serum creatinine is measured and GFR can only be estimated, creatinine and not GFR should be used along with urine output for diagnosis and staging of AKI, according to the KDIGO guidelines, with

the exception of children under 18 years of age, for whom an acute decrease in estimated GFR below 35 mL/min/1.73 m2 is considered as a criterion for Stage 3 AKI. AKI is a universal issue and transpires not only in intensive care units and surgical wards, but also in medical and paediatric wards, and deserves the utmost consideration as a predictor of immediate and longterm adverse consequences. The present study aimed to determine the spectrum of AKI according to RIFLE and AKIN criteria, in a population of patients admitted to an internal medicine department, and the comparison between the two classification systems. In light of the recently published KDIGO guidelines, these were also compared to the RIFLE and AKIN criteria. 2. Materials and methods We conducted a prospective and descriptive study, between 20 July and 31 December 2009 (164 days) in two wards (A and B — for a total of 66 inpatient beds) of the Internal Medicine Department of the Coimbra University Hospitals. The study included all patients admitted for a minimum period of 72 h, with the diagnosis of AKI or exacerbation of chronic kidney disease (CKD) on admission or as a complication developed during hospitalisation. All patients with Stage 5 CKD and patients with a functioning renal transplant were excluded. A previously developed questionnaire was completed, in which anamnestic data (age, gender, race, length of hospital stay, pathological personal antecedents, history of kidney disease or previous episodes of AKI, outpatient medication, contrast studies performed or administration of nephrotoxic drugs), physical examination (state of hydration and diuresis) and laboratory data (serum creatinine and estimated GFR by the four variable MDRD formula [9], in which GFR = 175 × (serum creatinine)−1.154 × (age)−0.203 × (0.742 if female gender) × (1.212 if black)) were collected and considered. Data were recorded on admission, at the time of identification of renal dysfunction, at 48 h after hospitalisation and at discharge. Subsequently, patients were stratified according to the RIFLE and AKIN criteria, as well as the KDIGO criteria [3,5,8]. Data analysis was performed using the commercially available statistical software programme SPSS® (version 17.0, SPSS, Chicago, IL, USA) for Windows®. Continuous variables are expressed as the mean ± standard deviation, and categorical variables as a percentage of the number of studied cases. The measure of agreement and correlation between the RIFLE, AKIN and KDIGO classifications was performed using Cohen's kappa coefficient and Spearman correlation, respectively. Measurements of agreement between 0.6 and 0.8 were considered substantial, and between 0.8 and 1.0 as almost perfect. A p value of less than 0.05 was considered statistically significant. 3. Results During the study period a total of 1045 patients were admitted to wards A and B of the Internal Medicine Department. The final sample comprised 69 patients (6.6% of all hospitalised patients), with 44.9% (n = 31) male patients and 55.1% (n = 38) females, all Caucasian, with an overall mean age of 79.7 ± 10.0 years (minimum 56 and maximum 97 years). The average length of hospital stay was 15 ± 12 days (minimum three and maximum 59 days). Most patients (53.6%, n = 37) had a history of acute kidney dysfunction in previous hospital admissions. Important comorbidities were ascertained in 59 patients, including arterial hypertension in 50.7% (n = 35), diabetes mellitus in 36.2% (n = 25), congestive heart failure in 33.3% (n = 23) and neoplastic diseases of variable aetiologies (with no documented renal tumours) in 15.9% (n = 11). The initial mean serum creatinine was 2.70 ± 1.21 mg/dL (238.7 ± 107.0 μmol/L), corresponding to a mean GFR of 21.7 ± 8.8 mL/min/ 1.73 m2. The average urine output in the first 24 h of admission was 1486 ± 794 mL (minimum 51 mL and maximum 3100 mL). AKI or exacerbated CKD was responsible for hospital admissions in 20.3%

Please cite this article as: Neves M, et al, Acute kidney injury in an internal medicine ward in a Portuguese quaternary hospital, Eur J Intern Med (2013), http://dx.doi.org/10.1016/j.ejim.2013.09.007

M. Neves et al. / European Journal of Internal Medicine xxx (2013) xxx–xxx

(n = 14) of patients and the overall incidence of AKI in hospitalised patients during this period in both wards was 6.6%. The primary cause of AKI was hypovolaemia due to dehydration (in 53.6%, n = 37). Other causes included decompensated congestive heart failure in 13% (n = 9), sepsis or septic shock in 7.2% (n = 5), varied infections in 7.2% (n = 5), nephrotoxic medication or contrast nephropathy in 7.2% (n = 5) and other unspecified causes in four patients (5.8%). The cause of AKI was unknown in four of the 69 patients (5.8%). The RIFLE classification identified a higher frequency of patients with some degree of AKI than the AKIN or KDIGO classifications (94.2% vs. 84.1% vs. 84.1%), with the RIFLE and AKIN classifications presenting a measure of agreement by Cohen's kappa coefficient of 0.64 and a statistically significant positive correlation (p b 0.001 and r = 0.57). RIFLE and KDIGO classifications also presented a measure of agreement by Cohen's kappa coefficient of 0.64 and a statistically significant positive correlation (p b 0.001 and r = 0.57), while stratifications with the AKIN and KDIGO criteria presented a measure of agreement by Cohen's kappa coefficient of 1.00 and a very strong statistically significant positive correlation (p b 0.001 and r = 1.00), since the staging systems are identical. The percentage of patients in the first stages (Risk vs. AKIN Stage 1 vs. KDIGO Stage 1, and Injury vs. AKIN Stage 2 vs. KDIGO Stage 2) was similar (26.1% vs. 29.0% vs. 29.0% and 23.2% vs. 26.1% vs. 26.1%, respectively). However, the RIFLE classification identified a greater number of patients with more severe disease than the AKIN or KDIGO classifications (Failure vs. AKIN Stage 3 vs. KDIGO Stage 3 in 45% vs. 29% vs. 29%). Four patients (5.8%) with diagnosis of AKI or exacerbated CKD on admission did not meet RIFLE criteria for AKI and 11 patients (15.9%) did not meet the diagnostic criteria of the AKIN or KDIGO classification for AKI. (Table 2) At the final assessment, results were similar across all degrees of dysfunction (risk vs. AKIN Stage 1 vs. KDIGO Stage 1 in 7.2% vs. 8.7% vs. 8.7%; Injury vs. AKIN Stage 2 vs. KDIGO Stage 2 in 10.1% vs. 10.1% vs. 10.1%; Failure vs. AKIN Stage 3 vs. KDIGO Stage 3 in 18.8% vs. 18.8% vs. 18.8%), with most patients not meeting the necessary criteria for the diagnosis of AKI in any one of the classifications (63.8% vs. 62.3% vs. 62.3%). The RIFLE and AKIN classifications presented, at the time of final assessment, a measure of agreement by Cohen's kappa coefficient of 0.97 and a very strong statistically significant positive correlation (p b 0.001 and r = 0.99). RIFLE and KDIGO classifications also presented a measure of agreement by Cohen's kappa coefficient of 0.97 and a very strong statistically significant positive correlation (p b 0.001 and r = 0.99), while, as expected, stratifications with the AKIN and KDIGO criteria presented a measure of agreement by Cohen's kappa coefficient of 1.00 and a very strong statistically significant positive correlation (p b 0.001 and r = 1.00). Three patients (4.3%) required renal replacement therapy, all of whom died during hospitalisation due to underlying severe medical conditions. In the studied population, fifteen patients (21.7%) died, most of which had more severe renal disease (Failure by RIFLE classification and Stage 3 by the AKIN and KDIGO classifications in 53.3% of deceased patients), with 20% of patients not presenting diagnostic criteria for AKI with either classification. At the time of discharge, the majority of patients (69.6%, n = 48) recovered to baseline creatinine values,

Table 2 Incidence of AKI at the time of diagnosis, stratified according to the RIFLE, AKIN and KDIGO classifications. RIFLE classification

AKIN Initial Assessment classification (n, %)

KDIGO Initial assessment classification (n, %)

Initial assessment (n, %)

Risk Injury Failure Without AKI Any category

18 (26.1%) 16 (23.2%) 31 (44.9%) 4 (5.8%) 65 (94.2%)

20 (29%) 18 (26.1%) 20 (29%) 11 (15.9%) 58 (84.1%)

20 (29%) 18 (26.1%) 20 (29%) 11 (15.9%) 58 (84.1%)

AKI — acute kidney injury.

Stage 1 Stage 2 Stage 3 Without AKI Any category

Stage 1 Stage 2 Stage 3 Without AKI Any category

3

with six patients (8.7%) maintaining some degree of renal dysfunction, without the need for chronic renal replacement therapy. 4. Discussion AKI, sometimes undervalued, presents with a broad spectrum — from a simple mild or moderate increase in serum creatinine from baseline to anuric renal failure and/or eventual need for renal replacement therapy [10–13]. It can entail severe clinical consequences, with high morbidity and mortality, particularly in patients with severe sepsis and multiorgan failure, and it can also prove to be economically costly due to the fact that the severity of the acute disorder leads to increased mortality, an increased need for renal replacement therapy and a prolonged hospitalisation period [7]. A systematic review of 13 publications showed a step increase in the relative risk of death in patients who met RIFLE criteria for the various stages of AKI compared with patients who did not suffer from AKI. Stratification into the risk category increases mortality by 2.4 times, the Injury group by 4.2 times and 6.2 times in the Failure group, with p b 0.0001 for all three groups [14]. Researchers in 2010 reported that transient azotaemia (defined as recovery of renal function within 72 h of diagnosis of AKI) has an independent correlation with a significantly increased risk of death (p b 0.0001) [15]. Epidemiological studies indicate the pre-renal causes as more often associated with the development of AKI. Our experience corroborates this fact since over half of the cases of documented AKI resulted from hypovolaemia secondary to dehydration. The incidence of AKI during the study period was 6.6%. In the literature, there are great variations in the reported incidence (1–31%), probably due to a variety of definitions previously in use [16]. Four patients (5.8%) with diagnosis of AKI or exacerbated CKD on admission did not meet RIFLE criteria for AKI and 11 patients (15.9%) did not meet the diagnostic criteria of the AKIN classification for AKI. This difference is, in part, due to the AKIN classification excluding initially dehydrated patients in whom the hydration state is optimized through medical therapy and who consequently show an improvement in their analytical results less than 48 h after admission. The RIFLE classification does not provide guidance for the discrimination between pre-renal and renal causes of AKI. This discrimination is crucial, since it has been wellestablished that prolonged renal hypoperfusion can result in tubular necrosis, while the timely restoration of circulating blood volume can prevent a negative outcome [17]. The mortality rate was high in the studied population — 21.7% during the study period, most with more severe kidney disease as classified by RIFLE, AKIN and KDIGO. The mortality reported in the literature varies between 19 and 83% [16]. Most patients (69.6%, n = 48) were discharged with improved renal function and returned to usual values of serum creatinine, but 4.3% (n = 3) came to require renal replacement therapy during the hospital stay. It is unknown whether there is significant capacity to discern between one system and another, but there seems to be a significant improvement in terms of predictive ability of AKI and sensitivity of the AKIN criteria in relation to the RIFLE classification [18]. The AKIN classification meets the proposal of “elimination” of dehydrated patients and identifying patients at risk, with changes in renal function deemed minor, that could, however, evolve unfavourably. It omits the prognostic classes that were considered in the RIFLE staging system, considering them as outcomes and not an integral part of staging of AKI, and also eliminates the reference to change in GFR as a criterion to stratify patients with AKI, as is the routine amongst most investigators and clinicians. In our study, the RIFLE classification identified 10.1% more patients as having some degree of AKI at the initial assessment (as expected by the inclusion of patients whose cause of AKI was hypovolaemia due to dehydration — these patients were excluded in the AKIN classification), but at the date of the final evaluation, results were similar in all degrees of renal dysfunction between the three classification systems. There are

Please cite this article as: Neves M, et al, Acute kidney injury in an internal medicine ward in a Portuguese quaternary hospital, Eur J Intern Med (2013), http://dx.doi.org/10.1016/j.ejim.2013.09.007

4

M. Neves et al. / European Journal of Internal Medicine xxx (2013) xxx–xxx

no published studies on the utility and measure of agreement between the classifications in patients admitted to internal medicine wards; therefore it was not possible to compare the results with those of other studies. However, in the present study, the ratings showed a good measure of agreement and a statistically significant positive correlation (p b 0.001), both at the initial and final assessments. An improved classification may eventually resort to new biomarkers that experience an earlier increase in renal dysfunction than creatinine, such as tubular and inflammatory proteins and other markers of tubular injury currently under investigation, such as kidney injury molecule 1, neutrophil gelatinase-associated lipocalin, interleukin 18 e N-acetyl-βD-glucosaminidase [19,20]. In conclusion, and as expected, our results showed a good agreement and correlation between the RIFLE, AKIN and KDIGO criteria for diagnosis of AKI in internal medicine wards, but by their characteristics, no classification seems to be clearly superior to the other. We propose the use of the term AKI in all patients with acute changes of creatinine or urinary output — de novo or superimposed on a pre-existing alteration, and the use of criteria, validated in numerous studies for definition and stratification of AKI in patients admitted to internal medicine wards, in order to enable early identification of changes that can be addressed in a timely manner, thereby preventing progression to chronic kidney disease or increased morbidity and mortality. The KDIGO guidelines seem to have acknowledged the need for implementation of an improved classification that comprises characteristics of the RIFLE and AKIN criteria and considers an interval longer than 48 h for the onset of AKI, as well as the inclusion of a minimum increase of 0.3 mg/dL (26.5 μmol/L) compared to baseline serum creatinine as a criterion for stratification. The importance of an aetiology variable [21] (such as the hydration status contemplated in the AKIN classification) and whether it is incorporated in or external to the definition of AKI is, however, not clearly defined within these guidelines. Just as the chronic kidney disease stages have been linked to specific treatment recommendations, showing to be of great use in the management of this disease and its complications, the KDIGO guidelines developed recommendations for evaluation and management of patients with AKI of varying origins using this stepwise approach and encompassing prevention techniques, pharmacological therapy and, when necessary, renal replacement therapy. The KDIGO guidelines now require validation through numerous studies, as occurs with the RIFLE classification [16,22–25], but the future course ultimately lies in the use of new biomarkers capable of greater sensitivity and earlier identification of patients at risk. Learning points • The term acute kidney injury was proposed in order to better reflect the broad spectrum of traditional acute renal failure and the term is now recognized as the preferred nomenclature for this complex clinical syndrome. • Results showed a good measure of agreement and a statistically significant positive correlation (p b 0.001), both at the initial and final assessments, between RIFLE, AKIN and KDIGO classifications in internal medicine wards. • An improved classification may eventually resort to new biomarkers that experience an earlier increase in renal dysfunction than creatinine, such as tubular and inflammatory proteins and other markers of tubular injury.

Conflict of interests This work had no financial or grant support. The authors declare no conflict of interest.

References [1] Lamiere N, Van Biesen W, Vanholder R. Acute renal failure. Lancet 2005;365:417–30. [2] Mehta RL, Chertow GM. Acute renal failure definitions and classification: time for change? J Am Soc Nephrol 2003;14:2178–87. [3] Mehta RL, Kellum JA, Shah SV, Molitoris BA, Ronco C, Warnock DG, et al. Acute kidney injury network: report of an initiative to improve outcomes in acute kidney injury. Crit Care 2007;11:R31. [4] Ricci Z, Cruz DN, Ronco C. Classification and staging of acute kidney injury: beyond the RIFLE and AKIN criteria. Nat Rev Nephrol 2011;7:201–8. [5] 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 2004;8: R204–12. [6] Pickering JW, Endre ZH. GFR shot by RIFLE: errors in staging acute kidney injury. Lancet 2009;373:1318–9. [7] 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;16:3365–70. [8] Kidney Disease: Improving Global Outcomes (KDIGO) Acute Kidney Injury Work Group. KDIGO clinical practice guideline for acute kidney injury. Kidney Int 2012:1–138. [9] Levey AS, Bosch JP, Lewis JB, Greene T, Rogers N, Roth D. A more accurate method to estimate glomerular filtration rate from serum creatinine: a new prediction equation. Modification of Diet in Renal Disease Study Group. Ann Intern Med 1999;130:461–70. [10] Mehta RL, Pascual MT, Soroko S, Savage BR, Himmelfarb J, Ikizler TA, et al. Spectrum of acute renal failure in the intensive care unit: the PICARD experience. Kidney Int 2004;66:1613–21. [11] Palevsky PM. Epidemiology of acute renal failure: the tip of the iceberg. Clin J Am Soc Nephrol 2006;1:6–7. [12] Ympa YP, Sakr Y, Reinhart K, Vincent JL. Has mortality from acute renal failure decreased? A systematic review of the literature. Am J Med 2005;118:827–32. [13] Metnitz PG, Krenn CG, Steltzer H, Lang T, Ploder J, Lenz K, et al. Effect of acute renal failure requiring renal replacement therapy on outcome in critically ill patients. Crit Care Med 2002;30:2051–8. [14] Ricci Z, Cruz D, Ronco C. The RIFLE criteria and mortality in acute kidney injury: a systematic review. Kidney Int 2008;73:538–46. [15] Uchino S, Bellomo R, Bagshaw SM, Goldsmith D. Transient azotaemia is associated with a high risk of death in hospitalized patients. Nephrol Dial Transplant 2010;25:1833–9. [16] Hoste EA, Clermont G, Kersten A, Venkataraman R, Angus DC, De Bacquer D, et al. RIFLE criteria for acute kidney injury are associated with hospital mortality in critically ill patients: a cohort analysis. Crit Care 2006;10:R73–83. [17] Schrier RW, Wang W, Poole B, Mitra A. Acute renal failure: definitions, diagnosis, pathogenesis, and therapy. J Clin Invest 2004;114:5–14. [18] 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. [19] Han WK, Bonventre JV. Biologic markers for the early detection of acute kidney injury. Curr Opin Crit Care 2004;10:476–82. [20] Zhou H, Hewitt SM, Yuen PS, Star RA. Acute kidney injury biomarkers: needs, present status, and future promise. NephSAP 2006;5:63–71. [21] Van Biesen W, Vanholder R, Lamiere N. Defining acute renal failure: RIFLE and beyond. Clin J Am Soc Nephrol 2006;1:1314–9. [22] 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 application, sensitivity, and predictability. Am J Kidney Dis 2005;46:1038–48. [23] Bell M, Liljestam E, Granath F, Fryckstedt J, Ekbom A, Martling CR. Optimal follow-up time after continuous renal replacement therapy in actual renal failure patients stratified with the RIFLE criteria. Nephrol Dial Transplant 2005;20:254–360. [24] Kuitunen A, Vento A, Suojaranta-Ylinen R, Pettilä V. Acute renal failure after cardiac surgery: evaluation of the RIFLE classification. Ann Thorac Surg 2006;81:542–6. [25] 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;34:1913–7.

Please cite this article as: Neves M, et al, Acute kidney injury in an internal medicine ward in a Portuguese quaternary hospital, Eur J Intern Med (2013), http://dx.doi.org/10.1016/j.ejim.2013.09.007