Original Report: Laboratory Investigation American
Journal of
Nephrology
Am J Nephrol 2015;41:81–88 DOI: 10.1159/000371748
Received: July 26, 2014 Accepted: December 28, 2014 Published online: February 10, 2015
Variation in Risk and Mortality of Acute Kidney Injury in Critically Ill Patients: A Multicenter Study Nattachai Srisawat a, b Florentina E. Sileanu a, c Raghavan Murugan a Rinaldo Bellomo d Paolo Calzavacca d Rodrigo Cartin-Ceba e Dinna Cruz f Judith Finn g Eric A. Hoste h Kianoush Kashani e Claudio Ronco i Steve Webb j John A. Kellum a on behalf of the Acute Kidney Injury-6 Study Group a
Key Words Acute kidney injury · KDIGO · RIFLE · Mortality
Abstract Background: Despite standardized definitions of acute kidney injury (AKI), there is wide variation in the reported rates of AKI and hospital mortality for patients with AKI. Variation could be due to actual differences in disease incidence, clinical course, or a function of data ascertainment and application of diagnostic criteria. Using standard criteria may help
© 2015 S. Karger AG, Basel 0250–8095/15/0411–0081$39.50/0 E-Mail [email protected] www.karger.com/ajn
determine and compare the risk and outcomes of AKI across centers. Methods: In this cohort study of critically ill patients admitted to the intensive care units at six hospitals in four countries, we used KDIGO criteria to define AKI. The main outcomes were the occurrence of AKI and hospital mortality. Results: Of the 15,132 critically ill patients, 32% developed AKI based on serum creatinine criteria. After adjusting for differences in age, sex, and severity of illness, the odds ratio for AKI continued to vary across centers (odds ratio (OR), 2.57– 6.04, p < 0.001). The overall, crude hospital mortality of patients with AKI was 27%, which also varied across centers
John A. Kellum, MD, MCCM Department of Critical Care Medicine, University of Pittsburg 604, Scaife Hall, 3550 Terrace Street Pittsburgh, PA 15261 (USA) E-Mail kellumja @ upmc.edu
Downloaded by: UCSF Library & CKM 169.230.243.252 - 3/24/2015 8:13:17 PM
The Center for Critical Care Nephrology, CRISMA, Department of Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pa., USA; b Excellence Center for Critical Care Nephrology, King Chulalongkorn Memorial Hospital, Thai Red Cross and Division of Nephrology, Department of Medicine, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; c Department of Biostatistics, University of Pittsburgh Graduate School of Public Health, Pittsburgh, Pa., USA; d Department of Intensive Care and Department of Medicine, Austin Hospital and University of Melbourne, Melbourne, Australia; e Division of Pulmonary and Critical Care Medicine, Department of Medicine, Mayo Clinic, Rochester, Minn., f Division of Nephrology-Hypertension, University of California, San Diego, Calif., USA; g Prehospital, Resuscitation and Emergency Care Research Unit (PRECRU), Curtin University, Perth, Australia; h Department of Intensive Care Medicine, University Hospital, Ghent University, Ghent, and Research Foundation-Flanders, Brussels, Belgium; i International Renal Research Institute of Vicenza, Department of Nephrology, Dialysis and Transplantation, San Bortolo Hospital, Vicenza, Italy; j Intensive Care Unit, Royal Perth Hospital, Wellington Street, Perth, Australia
after adjusting for KDIGO stage, differences in age, sex, and severity of illness (OR, 1.13–2.20, p < 0.001). The severity of AKI was associated with incremental mortality risk across centers. Conclusions: In this study, the absolute and severity-adjusted rates of AKI and hospital mortality rates for AKI varied across centers. Future studies should examine whether variation in the risk of AKI among centers is due to differences in clinical practice or process of care or residual confounding due to unmeasured factors. © 2015 S. Karger AG, Basel
Introduction
Acute kidney injury (AKI) is a common clinical syndrome defined as a sudden onset of reduced kidney function manifested by acute decline in the glomerular filtration rate [1]. Despite standardized definitions for diagnosis of AKI [2–4], the occurrence of AKI in the ICU continues to vary across centers [4–10]. Hospital mortality and dialysis rates have also varied [5–10]. Whether variation is due to the actual differences in disease incidence or is simply a function of data ascertainment and application of diagnostic criteria is not known. Recently, the Kidney Disease Improving Global Outcomes (KDIGO), a multi-disciplinary international group, proposed modifications to the Risk, Injury, Failure, Loss and End-stage renal disease (RIFLE) criteria in order to increase the sensitivity of early AKI detection [4]. Whether the KDIGO definition improves the reclassification of diagnosis of AKI and reduces variation compared to previous definitions of AKI (RIFLE) is not known. Using a multinational ICU dataset, the aims of our study were to first examine whether there is variation in risk-adjusted rates of AKI and hospital mortality associated with AKI across centers using the KDIGO definitions. Second, we examined whether the KDIGO definition augments reclassification of risk of AKI and hospital mortality associated with AKI compared to RIFLE classification.
who were older than 15 years and who were admitted to one of the participating ICUs during the observational period were considered for inclusion in the study. The study protocol was reviewed by the ethics committee, or the institutional review board at each participating site and the need for informed consent was waived. Diagnosis of Acute Kidney Injury The diagnosis of AKI was based on the validated RIFLE and KDIGO criteria (online suppl. table S1; for all online suppl. material, see www.karger.com/doi/10.1159/000371748) [2, 4]. The two criteria mainly differ in that for KDIGO criteria, if there is an increase in serum creatinine ≥0.3 mg/dl during a 48 h window period, AKI exists even if a 50% increase from baseline does not occurs. Definitions of the baseline and reference serum creatinine values are (1) baseline creatinine = lowest value between (a) most recent pre-hospital creatinine value up to one year prior to the index hospital admission and (b) first creatinine recorded on the current admission; (2) reference creatinine = the baseline creatinine when available; otherwise, the lowest between (a) first creatinine available and (b) (for patients without history of CKD) derived from the Modification on Diet in Renal Disease (MDRD) equation for equation for serum creatinine using a GFR of 75 ml/min/1.73 m2 as recommend and previously applied [2, 7, 11]. For analysis, patients were assigned to their worst KDIGO category according to either serum creatinine or urine output (when used – see below). We excluded patients who had pre-existing end stage renal disease (ESRD) before ICU admission. For urine output data analysis, we used body weight at ICU admission for calculation of the rate of urine flow per kilogram per hour. If cumulative urine output was available only for 24 h, we modified the urine output criteria from standard criteria, into ≥0.5 ml/kg/h as no AKI or stage 0, 0.3–0.5 ml/kg/h as Injury or stage 2,
Journal of
Nephrology
Am J Nephrol 2015;41:81–88 DOI: 10.1159/000371748
Received: July 26, 2014 Accepted: December 28, 2014 Published online: February 10, 2015
Variation in Risk and Mortality of Acute Kidney Injury in Critically Ill Patients: A Multicenter Study Nattachai Srisawat a, b Florentina E. Sileanu a, c Raghavan Murugan a Rinaldo Bellomo d Paolo Calzavacca d Rodrigo Cartin-Ceba e Dinna Cruz f Judith Finn g Eric A. Hoste h Kianoush Kashani e Claudio Ronco i Steve Webb j John A. Kellum a on behalf of the Acute Kidney Injury-6 Study Group a
Key Words Acute kidney injury · KDIGO · RIFLE · Mortality
Abstract Background: Despite standardized definitions of acute kidney injury (AKI), there is wide variation in the reported rates of AKI and hospital mortality for patients with AKI. Variation could be due to actual differences in disease incidence, clinical course, or a function of data ascertainment and application of diagnostic criteria. Using standard criteria may help
© 2015 S. Karger AG, Basel 0250–8095/15/0411–0081$39.50/0 E-Mail [email protected] www.karger.com/ajn
determine and compare the risk and outcomes of AKI across centers. Methods: In this cohort study of critically ill patients admitted to the intensive care units at six hospitals in four countries, we used KDIGO criteria to define AKI. The main outcomes were the occurrence of AKI and hospital mortality. Results: Of the 15,132 critically ill patients, 32% developed AKI based on serum creatinine criteria. After adjusting for differences in age, sex, and severity of illness, the odds ratio for AKI continued to vary across centers (odds ratio (OR), 2.57– 6.04, p < 0.001). The overall, crude hospital mortality of patients with AKI was 27%, which also varied across centers
John A. Kellum, MD, MCCM Department of Critical Care Medicine, University of Pittsburg 604, Scaife Hall, 3550 Terrace Street Pittsburgh, PA 15261 (USA) E-Mail kellumja @ upmc.edu
Downloaded by: UCSF Library & CKM 169.230.243.252 - 3/24/2015 8:13:17 PM
The Center for Critical Care Nephrology, CRISMA, Department of Critical Care Medicine, University of Pittsburgh School of Medicine, Pittsburgh, Pa., USA; b Excellence Center for Critical Care Nephrology, King Chulalongkorn Memorial Hospital, Thai Red Cross and Division of Nephrology, Department of Medicine, Faculty of Medicine, Chulalongkorn University, Bangkok, Thailand; c Department of Biostatistics, University of Pittsburgh Graduate School of Public Health, Pittsburgh, Pa., USA; d Department of Intensive Care and Department of Medicine, Austin Hospital and University of Melbourne, Melbourne, Australia; e Division of Pulmonary and Critical Care Medicine, Department of Medicine, Mayo Clinic, Rochester, Minn., f Division of Nephrology-Hypertension, University of California, San Diego, Calif., USA; g Prehospital, Resuscitation and Emergency Care Research Unit (PRECRU), Curtin University, Perth, Australia; h Department of Intensive Care Medicine, University Hospital, Ghent University, Ghent, and Research Foundation-Flanders, Brussels, Belgium; i International Renal Research Institute of Vicenza, Department of Nephrology, Dialysis and Transplantation, San Bortolo Hospital, Vicenza, Italy; j Intensive Care Unit, Royal Perth Hospital, Wellington Street, Perth, Australia
after adjusting for KDIGO stage, differences in age, sex, and severity of illness (OR, 1.13–2.20, p < 0.001). The severity of AKI was associated with incremental mortality risk across centers. Conclusions: In this study, the absolute and severity-adjusted rates of AKI and hospital mortality rates for AKI varied across centers. Future studies should examine whether variation in the risk of AKI among centers is due to differences in clinical practice or process of care or residual confounding due to unmeasured factors. © 2015 S. Karger AG, Basel
Introduction
Acute kidney injury (AKI) is a common clinical syndrome defined as a sudden onset of reduced kidney function manifested by acute decline in the glomerular filtration rate [1]. Despite standardized definitions for diagnosis of AKI [2–4], the occurrence of AKI in the ICU continues to vary across centers [4–10]. Hospital mortality and dialysis rates have also varied [5–10]. Whether variation is due to the actual differences in disease incidence or is simply a function of data ascertainment and application of diagnostic criteria is not known. Recently, the Kidney Disease Improving Global Outcomes (KDIGO), a multi-disciplinary international group, proposed modifications to the Risk, Injury, Failure, Loss and End-stage renal disease (RIFLE) criteria in order to increase the sensitivity of early AKI detection [4]. Whether the KDIGO definition improves the reclassification of diagnosis of AKI and reduces variation compared to previous definitions of AKI (RIFLE) is not known. Using a multinational ICU dataset, the aims of our study were to first examine whether there is variation in risk-adjusted rates of AKI and hospital mortality associated with AKI across centers using the KDIGO definitions. Second, we examined whether the KDIGO definition augments reclassification of risk of AKI and hospital mortality associated with AKI compared to RIFLE classification.
who were older than 15 years and who were admitted to one of the participating ICUs during the observational period were considered for inclusion in the study. The study protocol was reviewed by the ethics committee, or the institutional review board at each participating site and the need for informed consent was waived. Diagnosis of Acute Kidney Injury The diagnosis of AKI was based on the validated RIFLE and KDIGO criteria (online suppl. table S1; for all online suppl. material, see www.karger.com/doi/10.1159/000371748) [2, 4]. The two criteria mainly differ in that for KDIGO criteria, if there is an increase in serum creatinine ≥0.3 mg/dl during a 48 h window period, AKI exists even if a 50% increase from baseline does not occurs. Definitions of the baseline and reference serum creatinine values are (1) baseline creatinine = lowest value between (a) most recent pre-hospital creatinine value up to one year prior to the index hospital admission and (b) first creatinine recorded on the current admission; (2) reference creatinine = the baseline creatinine when available; otherwise, the lowest between (a) first creatinine available and (b) (for patients without history of CKD) derived from the Modification on Diet in Renal Disease (MDRD) equation for equation for serum creatinine using a GFR of 75 ml/min/1.73 m2 as recommend and previously applied [2, 7, 11]. For analysis, patients were assigned to their worst KDIGO category according to either serum creatinine or urine output (when used – see below). We excluded patients who had pre-existing end stage renal disease (ESRD) before ICU admission. For urine output data analysis, we used body weight at ICU admission for calculation of the rate of urine flow per kilogram per hour. If cumulative urine output was available only for 24 h, we modified the urine output criteria from standard criteria, into ≥0.5 ml/kg/h as no AKI or stage 0, 0.3–0.5 ml/kg/h as Injury or stage 2,
