Establishment and validation of a new predictive equation with multiple risk factors for the development of cardiorenal syndrome type 1 in patients with acute myocardial infarction.
10.3760/cma.j.cn112148-20201118-00916
- VernacularTitle:联合预测因子对急性心肌梗死患者院内并发1型心肾综合征风险的预测价值
- Author:
Cai Hong HE
1
;
Jin Wu LIU
1
;
Zhao He ZHU
2
;
Hong Wei PAN
1
;
Zhao Fen ZHENG
1
;
Jin HE
1
;
Zheng Yu LIU
1
;
Yu ZHANG
1
;
Chang Lu WANG
1
;
Jing Jing RONG
1
;
Yi TANG
1
;
Qing Hai ZHANG
1
Author Information
1. Department of Cardiology, Hunan Provincial People's Hospital, First Affiliated Hospital of Hunan Normal University, Changsha 410000, China.
2. Department of Cardiology, Luohe Central Hospital, Luohe 462000, China.
- Publication Type:Journal Article
- From:
Chinese Journal of Cardiology
2021;49(8):802-808
- CountryChina
- Language:Chinese
-
Abstract:
Objective: To investigate the independent risk factors of cardiorenal syndrome type 1 (CRS1) in patients with acute myocardial infarction (AMI) and to build a predictive equation for the development of CRS1 in these patients. Method: Consecutive inpatients with AMI, who hospitalized from January 2017 to December 2018 in the Hunan Provincial People's Hospital, were enrolled in this case-control study. Patients were divided into CRS1 group and non-CRS1 group according to the presence or absence of CRS1.The clinical data were collected through the electronic medical record system of Hunan Provincial People's Hospital. The matching process was conducted with a minimum-distance scoring method and a 1∶1 match between the CRS1 group and the no-CRS1 group, the propensity score was calculated through the logistic regression model. Factors with statistically significant differences in univariate analysis were included in the multivariate logistic regression model to analyze the risk factors of AMI patients with CRS1, then the independent risk factors were used to establish a predicting equation for CRS1 by logistic regression function for model building. Area under the curve (AUC) value and the best cut-off value of the combined predictors was determined according to the ROC curve. Python 3.8 software was used to perform 10-fold cross-validation on modeling samples. Results: A total of 942 patients were included, there were 113 cases in CRS1 group and 829 cases in non-CRS1 group. Ultimately, 99 CRS1 patients were successfully matched to 99 non-CRS1 patient using 1∶1 matching. After propensity score matching, the baseline age and sex along with heart rate, mean arterial pressure, percentage of people with a history of diabetes, hypertension, ST-segment elevation myocardial infarction, myocardial ischemia time, angiotensin converting enzyme inhibitors or angiotensin Ⅱ receptor blockers use, and β receptor blocker use were similar between the two groups(all P>0.05). The contrast agent dosage was also similar between the two groups (P=0.266). The peak cardiac troponin I (cTnI), N-terminal pro-brain natriuretic peptide(NT-proBNP), white blood cell count, base estimated glomerular filtration rate (eGFR), albumin and hemoglobin levels were statistically significant between the two groups (all P<0.05). Multivariate logistic regression analysis showed that decreased baseline eGFR, increased NT-proBNP, peak cTnI concentrations and white blood cell count were independent risk factors of CRS1 in AMI patients (all P<0.01).The predicting equation of the combined predictor was established by transforming the logistic model equation, L=0.031×cTnI+0.000 2×NT-proBNP-0.024×eGFR+0.254×white blood cell count, where L represented the combined predictor. ROC curve analysis indicated that the AUC of the peak cTnI, NT-proBNP, baseline eGFR, white blood cell count, and combined predictor were 0.76, 0.85, 0.79, 0.81, and 0.92 respectively (all P<0.05), and the cutoff value of combined predictor was 2.6. The AUC of ROC curve after the model's ten-fold cross validation was 0.89. Conclusions: Decreased baseline eGFR, increased NT-proBNP, peak cTnI concentrations and white blood cell count are the independent risk factors for CRS1 in AMI patients. The combined predictor equation based on the above 4 biomarkers presents a good predictive value for CRS1 in AMI patients.
