Development of a risk prediction model for acute cerebral infarction in patients with type 2 diabetes mellitus
10.3760/cma.j.cn115624-20240401-00251
- VernacularTitle:2型糖尿病患者伴发急性脑梗死的风险预测模型的构建
- Author:
Rong YANG
1
;
Jie ZHENG
;
Jiaojiao GUO
;
Caiyun GUO
;
Shiwei LIU
Author Information
1. 山西医科大学第三医院（山西白求恩医院 山西医学科学院 同济山西医院）内分泌科，太原 030032
- Keywords:
Diabetes mellitus, type 2;
Cerebral infarction;
Risk factors;
Forecasting;
Nomogram
- From:
Chinese Journal of Health Management
2024;18(12):886-893
- CountryChina
- Language:Chinese
-
Abstract:
Objective:To develop a risk prediction model of acute cerebral infarction (ACI) in patients with type 2 diabetes mellitus (T2DM).Methods:It was a cross-sectional study. The clinical data of 798 patients with T2DM hospitalized in the Department of Endocrinology and Neurology of Shanxi Bethune Hospital from August 2021 to October 2023 were collected. Based on whether they had concurrent ACI, the patients were divided into T2DM with ACI group (case group) and pure T2DM group (control group). The patients were then allocated to a training set ( n=558) and a validation set ( n=240) in a 7∶3 ratio by the sample functions in R software. LASSO regression was employed to screen and optimize variables, and a multivariate logistic regression analysis was used to establish the nomogram prediction model. The discriminative ability, calibration, and clinical usefulness of the risk prediction model were assessed with receiver operating characteristic (ROC) curves, calibration curves, and decision curve analysis, respectively. Results:LASSO regression identified gender, age, systolic blood pressure, fasting plasma glucose (FPG), albumin (ALB), and carotid vascular condition as the variables for prediction. The multivariable logistic regression analysis showed that female ( OR=0.489, 95% CI: 0.308-0.778) and ALB ( OR=0.846, 95% CI: 0.795-0.901) were protective factors for ACI occurrence in T2DM patients, while age ( OR=1.051, 95% CI:1.025-1.077), systolic blood pressure ( OR=1.047, 95% CI: 1.034-1.059), FPG ( OR=1.185, 95% CI: 1.089-1.288), and carotid plaque ( OR=7.359, 95% CI: 3.050-17.756) were risk factors. The area under the ROC curve (AUC) for risk of ACI in the training set was 0.863(95% CI: 0.833-0.893), and it was 0.846(95% CI: 0.797-0.896) for the validation set. Calibration curves and the Hosmer-Lemeshow goodness-of-fit test indicated good model fit (training set χ2=8.311, P=0.404; validation set χ2=3.957, P=0.861). Decision curve analysis showed that the clinical effectiveness of the model was higher when the threshold probabilities of the training set and the validation set was 0.02-0.93 and 0.12-0.99, respectively. Conclusion:In this study, a prediction model of ACI risk in T2DM patients was successfully established.
