Abstract:OBJECTIVE:To evaluate the effectiveness of different screening strategies for type 2 diabetes to prevent cardiovascular disease in a community-based Chinese population from economically developed areas based on the Chinese electronic health records research in Yinzhou (CHERRY) study.METHODS:A Markov model was used to simulate different systematic diabetes screening strategies, including: (1) screening among Chinese adults aged 40-70 years recommended by the 2020 Chinese Guideline for the prevention and Treatment of Type 2 Diabetes (Strategy 1); (2) screening among Chinese adults aged 35 to 70 years recommended by the 2022 American Diabetes Association Standard of Medical Care in Diabetes (Strategy 2); and (3) screening among Chinese adults aged 35-70 years with overweight or obesity recommended by the 2021 United States Preventive Services Task Force Recommendation Statement on Screening for Prediabetes and Type 2 Diabetes (Strategy 3). According to the guidelines, individuals who were screened positively (fasting plasma glucose ≥ 7.0 mmol/L) would be introduced to intensive glycemic targets management (glycated hemoglobin < 7.0%).The Markov model simulated different screening scenarios for ten years (cycles) with parameters mainly from the CHERRY study or published literature. Number of cardiovascular disease events or deaths could be prevented and number needed to screen (NNS) were calculated to compare the effectiveness of the different strategies. One-way sensitivity analysis on the sensitivity of screening methods and probabilistic sensitivity analysis on uncertainties of diabetes incidence, the sensitivity of screening methods, and intensive glycemic management effects were conducted.RESULTS:Totally 289 245 Chinese adults aged 35-70 years without cardiovascular diseases or diagnosed diabetes at baseline were enrolled. In terms of the number of cardiovascular disease events could be prevented, Strategy 1 for systematic diabetes screening among the adults aged 35-70 years was 222 (95%UI: 180-264), Strategy 2 for systematic diabetes screening among the adults aged 40-70 years was 227 (95%UI: 185-271), and Strategy 3 for systematic diabetes screening among the adults aged 35-70 years with obesity or overweight (body mass index ≥ 24 kg/m2) was 131 (95%UI: 98-164), compared with opportunistic screening. NNS per cardiovascular disease event for the strategies 1, 2 and 3 were 1 184 (95%UI: 994-1 456), 1 274 (95%UI: 1 067-1 564) and 814 (95%UI: 649-1 091), respectively. Compared with Strategy 1, NNS per cardiovascular disease event for Strategy 2 increased by 90 (95%UI: -197-381) with similar effectiveness of cardiovascular prevention; however, NNS per cardiovascular disease event for Strategy 3 was reduced by 460 (95%UI: 185-724) in contrast to the Strategy 2, suggesting that the Strategy 3 was more efficient. The results were consistent in multiple sensitivity analyses.CONCLUSION:Systematic screening for diabetes based on the latest guidelines in economically developed areas of China can reduce cardiovascular events and deaths. However, merely lowering the starting age of screening from 40 to 35 years seems ineffective for preventing cardiovascular disease, while screening strategy for Chinese adults aged 35-70 years with overweight or obesity is recommended to improve efficiency.

To Set Up Model As Screening Tools for Estimating the Risk of DM in Community by Use of Decision Tree and Data of Health Records

Study on the Application of Classification Tree Model in Screening the Risk Factors of Malignant Tumor

Application of a Decision Tree Model for Predicting Diabetic Retinopathy

Diabetes risk prediction model based on community follow-up data using machine learning

Diabetes Risk Data Mining Method Based on Electronic Medical Record Analysis

Diabetes prediction model for unbalanced community follow-up data set based on optimal feature selection and scorecard

Diabetes prediction models based on health check-up population in a hospital

A Risk Assessment Model for Type 2 Diabetes in Chinese

A risk prediction model for type 2 diabetes mellitus complicated with retinopathy based on machine learning and its application in health management

Cardiovascular disease risk prediction models for type 2 diabetes:verification in Chinese patients

Research on Diabetes Risk Prediction Model at Early Stage Based on Machine Learning

Using machine learning models to improve stroke risk level classification methods of China national stroke screening

Establishment and health management application of a prediction model for high-risk complication combination of type 2 diabetes mellitus based on data mining

A Dynamic Risk-Based Early Warning Monitoring System for Population-Based Management of Cardiovascular Disease

Comparison of three data mining models for predicting diabetes or prediabetes by risk factors

Risk prediction of diabetes and pre-diabetes based on physical examination data

[Effectiveness of Different Screening Strategies for Type 2 Diabete on Preventing Cardiovascular Diseases in a Community-Based Chinese Population Using a Decision-Analytic Markov Model].

Setting up a risk prediction model on metabolic syndrome among 35-74 year-olds based on the Taiwan MJ Health-checkup Database

Risk Prediction of Diabetes: Big data mining with fusion of multifarious physical examination indicators

A Data-experience intelligent model to integrate human judging behavior and statistics for predicting diabetes complications

Research on diabetes prediction method based on electronic medical record data analysis