| Summary: | Hui Zhang,1 Dandan Chen,1 Jing Shao,1 Ping Zou,2 Nianqi Cui,3 Leiwen Tang,1 Dan Wang,1 Zhihong Ye1 1Zhejiang University, School of Medicine, Sir Run Run Shaw Hospital, Hangzhou, Zhejiang, People’s Republic of China; 2School of Nursing, Nipissing University, Toronto, Ontario, Canada; 3Department of Nursing, The Second Affiliated Hospital, Zhejiang University School of Medicine, Hangzhou, Zhejiang, People’s Republic of ChinaCorrespondence: Zhihong YeZhejiang University, School of Medicine, Sir Run Run Shaw Hospital, Qingchun Dong Road, Jianggan Strict, Hangzhou, Zhejiang, People’s Republic of ChinaTel +86 13606612119Email 3192005@zju.edu.cnPurpose: A prognostic prediction model for metabolic syndrome can help nurses or physicians evaluate the future individual absolute risk of MetS in order to develop personalized care strategies. We aimed to derive and internally validate a prognostic prediction model for 4-year risk of metabolic syndrome in adults.Patients and Methods: This was a retrospective cohort study conducted in a tertiary care setting, and the dataset was obtained from the Healthcare Information and Management Systems of a tertiary hospital. The cohort included Chinese adults attending health examination from 1 January 2011 to 31 December 2014. A total of 6793 participants without metabolic syndrome were included in the cohort and were followed up for 4 years. Available candidate predictors in the dataset were weight, MCV, MCH, AST, ALT, BMI, NGC, TC, serum uric acid, gender, smoking, WBC, LC, Hb, HCT, and age. A logistic regression model was adopted to build the risk equation, and bootstrapping was used when considering internal validation. Calibration, discrimination, and the clinical utility were calculated for the model’s performance.Results: Of the 6793 participants, 1750 participants were diagnosed with metabolic syndrome within 4 years. The developed prediction model contained 5 predictors (body mass index, age, total cholesterol, alanine transaminase, and serum uric acid). After internal validation, the C-statistic was 0.783 (95% CI, 0.772– 0.795). Additionally, the current model had good calibration. Calibration slope was 0.995 (95% CI, 0.934– 1.058), and calibration intercept was − 0.008 (95% CI, − 0.088– 0.073). The Brier score was 0.156. The decision-curve analysis indicated that the prediction model provided greater net benefit than the default strategies of providing treatment or not providing treatment for all patients.Conclusion: A prognostic risk prediction model for determining 4-year risk of metabolic syndrome onset in adults was developed and internally validated. This model was based on routine clinical measurements that quantified individual future risk of metabolic syndrome.Keywords: prediction model, prognosis, metabolic syndrome, algorithms, calibration, discrimination
|
|---|
