Evaluating The Diagnostic Value Of RIFLE And AKIN Criteria In Critically Ill Patients Of Acute Kidney Injury In Intensive Care Unit

Objective:1.To compare the difference between RIFLE criteria andAKIN criteria in the diagnosis of acute kidney injury in critically illpatients.2.To evaluate the relationship between the RIFLE or AKIN criteriaand hospital mortality.3.To compare the difference between RIFLE criteria andAKIN criteria to predict the outcome of patients with AKI.Methods:We retrospectively analyzed patients admitted to theemergency ICU in the Second Hospital of Hebei Medical University fromJanuary1st,2011to October31th,2012.1.Screening conditions:(1).ICUlength of stay≥24hours,(2).age≥18years,(3).We excluded the patients withthe past history of chronic kidney disease, kidney transplantation, and patientsreceived renal replacement therapy(RRT).2.We retrieved the basic data andclinical data.Severity of illness was assessed using the APACHE II score.3.AKI was classified according to the RIFLE and AKIN criteriarespectively,and we compared the difference between two criterion in thediagnosis of acute kidney injury and the value to predict theoutcome.4.Statistical analysis was performed using SPSS13.0.Results: We included331critically ill patients,of that,107AKI patientswere diagnosed by AKIN criteria(1stage46,2stage22,3stage39). On theother hand,104AKI patients were diagnosed by RIFLE criteria(Risk30,Injury30,Failure44). There were no statistical differences (32.3%vs31.4%,P>0.05)between two criterion in the diagnosis of AKI. But we found the significantdifference(P<0.05) between the AKI incidence by AKIN stage1(16.7%) andRIFLE Risk stage(11.7%). A diagnosis of septic shock, serum potassium,APACHE II score were the independent risk factor for development of AKI.Crude hospital mortality was significantly higher for AKI defined by any of RIFLE criteria and AKIN criteria(AKIN:30.8%vs6.3%P<0.05,RIFLE：31.7%vs6.2%P<0.05).In all patients, the AUC of the ROC curve, the indexfor hospital mortality prediction for RIFLE criteria was0.740,95%CI0.657-0.822, P<0.05, and for AKIN was0.736,95%CI0.653-0.818, P<0.05.No significant difference was found between the two criterion in hospitalmortality(30.8%vs31.7%, P>0.05). Any stage of RIFLE and AKIN criteria,APACHE II score, CRRT and mechanical ventilation were the independentrisk factor for hospital mortality. In123AKI patients, hospital mortality ofpatients needing CRRT was significantly increased(17.9%vs13.0%, P<0.05).Conclusion:Compared to the RIFLE criteria1.the AKIN criteria can notmaterially improve the sensitivity of the definition of AKI in critically illpatients, except in AKI early diagnosis.2.Both AKIN and RIFLE criteria havethe important value of assessing the state of AKI and predicting theoutcome.3.AKIN criteria can not improve the predictive ability of theclassification of AKI in critically ill patients.4.Any stage of RIFLE and AKINcriteria, APACHE II score, CRRT and mechanical ventilation are theindependent risk factor for hospital mortality.5.A diagnosis of septic shock,serum potassium and APACHE II score are the independent risk factor fordevelopment of AKI.6.In all AKI patients, the patients needing CRRT stillhave a high hospital mortality.