Establishment And Validation Of A Risk Prediction Model For Acute Kidney Injury In Patients With Acute Heart Failure

Objective:Through analyzing the risk factors of Acute Kidney Injury(AKI)in patients with Acute Heart Failure(AHF),established a risk prediction model,in order to identify AHF patients with high risk of AKI early,intervention early to reduce the occurrence of AKI and improving the prognosis of patients with AHF.Methods:386 patients with AHF admitted to the first affiliated hospital of soochow university from January 2016 to December 2017 were selected as the study subjects,The subjects were randomly divided into model group(2/3,257 cases)and validation group(1/3,129 cases).According to whether AKI occurred or not,model group patients were divided into AKI group and N-AKI group.Application of SPSS 22.0 to statistical analysis,the Logistic regression method to analyze the risk factors of AKI in patients with AHF,according to ? to score of various factors,establish a risk prediction model,using the area under receiver operating characteristic curve(AUC)and Hosmer-Lemeshow goodness-of-fit test to evaluate the resolution and calibration of the model.The validation group was used to evaluate the stability of the model.Results:1.Among the 257 patients with AHF in the model group,there were 186 males and 71 females,with a male-female ratio of 2.62:1.The mean age was 68.76±13.56 years.The top three causes of AHF were coronary heart disease(68.5%),hypertensive heart disease(10.9%)and dilated cardiomyopathy(7.8%).Chronic heart failure with acute decompensation accounted for 63.0%,and new acute heart failure accounted for 37.0%.2.The incidence of AKI in AHF patients was 38.9%,with stage 1 of AKI accounting for 29.2%,stage 2 of AKI for 5.4%,and stage 3 of AKI for 4.3%.The hospitalization days and inpatient mortality in AKI group were higher than those in N-AKI group(P<0.05).3.Univariate logistic regression analysis showed that diabetes,pulmonary infection,age?72 years,diastolic blood pressure?78mmHg,hematocrit?0.4L/L,fasting plasma glucose?6.0mmol/L,eGFR<60ml/min/1.73m2,uric acid?430?mol/L and hypersensitive c-reactive protein?12.37mg/L were associated with the occurrence of AKI in AHF patients(P<0.05).4.Multivariate logistic regression analysis showed that eGFR<60ml/min/1.73m2,age?72 years,diastolic blood pressure?78mmHg,fasting plasma glucose?6.0mmol/L and uric acid?>430?mol/L were independent risk factors for AKI in AHF patients(P<0.05).5.According to P to score the above risk factors,and established a risk prediction model;The score of eGFR<60ml/min/1.73m2,age?72 years,diastolic blood pressure<78mmHg,fasting plasma glucose?6.0mmol/L and uric acid?430?mol/L were 5,2,2,2 and 3,respectively.The AUC of the model was 0.773(P<0.001,95%CI:0.713-0.834).Hosmer-Lemeshow goodness-of-fit test results P=0.806>0.05,indicating that the model has good discrimination ability and calibration.The optimal diagnostic threshold of the model was 6 points,the sensitivity was 72.0%,and the specificity was 71.3%.Patients with?6 points had a higher incidence of AKI than those with<6 points,and the difference was statistically significant(62.1%:19.9%,P<0.001).6.The incidence of AKI was positively correlated with the total risk score(r=0.479,P<0.001).According to the total risk score,patients were divided into five groups:extremely low risk group(0-2 points),low-risk group(3-5 points),medium-risk group(6-8 points),high-risk group(9-11 points)and extremely high risk group(?12 points).The incidence of AKI in each group was 14.8%,23.5%,45.8%,64.5%and 89.3%,respectively.With the increase of risk stratification,the inpatient mortality of AHF patients was gradually increased,and the very high risk group significantly higher than the very low risk group and the low risk group(P<0.05).7.The AUC of the validation group was 0.758(P<0.001,95%CI:0.672-0.844),and the Hosmer-Lemeshow goodness-of-fit test P=0.785>0.05,indicating that the risk prediction model has good resolution and calibration in the validation group.Conclusions:1.The majority of patients with AHF are elderly,and the top three causes of heart failure are coronary heart disease,hypertensive heart disease and dilated cardiomyopathy.2.The incidence of AKI in patients with AHF is 38.9%,among which stage 1 of AKI is mainly,and stage 2 and 3 are less.The hospitalization days of patients with AKI was significantly prolonged,and the inpatient mortality rate was higher.3.The eGFR<60ml/min/1.73m2,age?72 years,diastolic blood pressure?78mmHg,fasting plasma glucose?6.0mmol/L and uric acid?430?mol/L were independent risk factors for the occurrence of AKI in AHF patients.4.According to ? to score the above risk factors,and established a risk prediction model;The model was verified to have good resolution and calibration,and it also has good prediction ability for the risk of inpatient mortality of AHF patients.