Risk Prediction Scores for Mortality, Cerebrovascular, and Heart Disease Among Chinese People With Type 2 Diabetes

Abstract

Context Risk scores for cardiovascular and mortality outcomes have not been commonly applied in Chinese populations. Objective To develop and externally validate a set of parsimonious risk scores [University of Hong Kong-Singapore (HKU-SG)] to predict the risk of mortality, cerebrovascular disease, and ischemic heart disease among Chinese people with type 2 diabetes and compare HKU-SG risk scores to other existing ones. Design Retrospective population-based cohorts drawn from Hong Kong Hospital Authority health records from 2006 to 2014 for development and Singapore Ministry of Health records from 2008 to 2016 for validation. Separate five-year risk scores were derived using Cox proportional hazards models for each outcome. Setting Study participants were adults with type 2 diabetes aged 20 years or over, consisting of 678,750 participants from Hong Kong and 386,425 participants from Singapore. Main Outcome Measures Performance was evaluated by discrimination (Harrell C-index), and calibration plots comparing predicted against observed risks. Results All models had fair external discrimination. Among the risk scores for the diabetes population, ethnic-specific risk scores (HKU-SG and Joint Asia Diabetes Evaluation) performed better than UK Prospective Diabetes Study and Risk Equations for Complications Of type 2 Diabetes models. External validation of the HKU-SG risk scores for mortality, cerebrovascular disease, and ischemic heart disease had corresponding C-indices of 0.778, 0.695, and 0.644. The HKU-SG models appeared well calibrated on visual plots, with predicted risks closely matching observed risks. Conclusions The HKU-SG risk scores were developed and externally validated in two large Chinese population-based cohorts. The parsimonious use of clinical predictors compared with previous risk scores could allow wider implementation of risk estimation in diverse Chinese settings.

Development of a predictive risk model for all-cause mortality in patients with diabetes in Hong Kong

“…BMJ Open Diab Res Care 2021;9:e001950. doi:10.1136/bmjdrc-2020-001950 available, [12][13][14][15] these have generally not incorporated temporal measures of variability for longitudinal data or machine learning approaches, both of which can enhance risk prediction. 16 17 Indeed, with the rapid development of big data analytics, it has become easier to improve discrimination by analyzing complex interactions among variables.…”

Development and validation of the CHIME simulation model to assess lifetime health outcomes of prediabetes and type 2 diabetes in Chinese populations: A modeling study

“…The J-EDIT trial enrolled substantially older participants of age 65 to 84 years (mean age 72) compared to the CMS dataset (mean age 61.9). This older age group with higher risk may be closer to the UKPDS-OM2, which tends to overestimate risks in trials and cohorts [ 4 , 6 , 7 , 24 ]. CHIME also performed well for the ADVANCE trial, which may be due to the diverse geographical/ethnic mix of the trial participants with almost 40% of participants from Asia.…”

“…Yet differences in epidemiology and outcomes among East Asian populations with diabetes render application of existing models that were developed and validated in European and North American populations problematic [ 3 ]. The most widely used model, United Kingdom Prospective Diabetes Study Outcomes Model 2 (UKPDS-OM2), is underpinned by risk equations from a 1970s UK cohort and overestimates the absolute risks of coronary heart disease and stroke among East Asians [ 4 , 5 ]. The more recent Risk Equations for Complications Of type 2 Diabetes (RECODe) model for 10-year risks was developed from a trial in the United States/Canada and has been validated in both North American trials and cohorts [ 6 , 7 ].…”

Cardiovascular risk in patients with type 2 diabetes: A systematic review of prediction models

“…The results of the 19 models developed in patients with type 2 diabetes are given in Table 1. [13][14][15][16][17][18][19][20][21][22][23][24][25][26][27][28][29][30][31] Fourteen of them used observational cohort data, four used registry data and the latter used clinical trial data. Eight were developed in the European population, six in the Asian population, three in the North American population and the other two in the Oceanian population.…”

“…Fifteen referred to a general CVD outcome and the last four to a specific subtype of CVD: two to stroke [18,23], one to cerebrovascular disease [21] and one to heart failure [26]. Among those referred to general CVD, three were very restrictive and included only CVD death 10 [13,14,30], five included only fatal and nonfatal acute myocardial infarction (AMI), stroke or CVD death [16,20,22,25,29], and the last seven were inclusive and accounted for a variety of fatal and nonfatal events [15,17,19,24,27,28,31], one of which included even peripheral arterial disease (PAD) [27].…”