Authors: (including presenting author): :
Zhang XD(1), Xiao CT (), Zhang YL(1), Lo LYN(2), Wai AKC(1), Teo KC(3), Lam RPK(1), Lau GKK(3), Ho JWK(2)(4), Rainer TH(1)
Affiliation: :
(1) Department of Emergency Medicine, School of Clinical Medicine, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong SAR, China. (2) School of Biomedical Sciences, Li Ka Shing Faculty of Medicine, The University of Hong Kong, Pokfulam, Hong Kong SAR, China. (3) Division of Neurology, Department of Medicine, School of Clinical Medicine, LKS Faculty of Medicine, The University of Hong Kong, Hong Kong SAR, China. (4) School of Biomedical Engineering, The University of Hong Kong, Pokfulam, Hong Kong SAR, China.
Keyword 1: :
Cardiovascular disease
Keyword 2: :
Primary prevention
Keyword 3: :
SCORE2-Diabetes
Keyword 4: :
External validation
Keyword 5: :
Older person
Introduction: :
The SCORE2-Diabetes model estimates 10-year cardiovascular disease (CVD) risk in patients with type 2 diabetes (T2DM). However, its performance may vary across ethnic and demographic groups.
Objectives: :
This study aimed to validate SCORE2-Diabetes in a representative Hong Kong T2DM cohort and to assess its generalizability across age and renal function subgroups.
Methodology: :
We utilised data from a Hong Kong representative T2DM cohort (2009–2011), comprising patients aged 40–89 years with T2DM who had no prior history of CVD. We assessed the performance of the original SCORE2-Diabetes model and four region-specific recalibrated versions in predicting 10-year CVD risk. Cardiovascular endpoints were defined in strict conformity with the SCORE2-Diabetes criteria, encompassing non-fatal myocardial infarction (MI), non-fatal ischaemic stroke (IS), and fatal CVD. Subgroup analyses were stratified by age decade (40–49, 50–59, 60–69, 70–79, 80–89 years) and chronic kidney disease (CKD) staging (G1–G5). We used the C-index to assess the discrimination performance of SCORE2-Diabetes. The study followed the Transparent Reporting of a multivariable prediction model for Individual Prognosis or Diagnosis (TRIPOD) guidelines.
Result & Outcome: :
Among 108,726 patients with T2DM in Hong Kong (mean follow-up: 8.6 years), 14,644 (13.5%) experienced CVD events, and 13,970 (12.8%) died from non-CVD causes. The SCORE2-Diabetes model yielded a C-Index of 0.734 (95% confidence interval [CI]: 0.728–0.739) for females, 0.704 (95% CI: 0.698–0.710) for males in predicting 10-year CVD risk. The original model showed the best calibration, underestimating 10-year CVD risk by only 0.6% in women and 0.1% in men. The Hong Kong regional recalibrated model significantly underestimated risk by 11.6% in females and 17.3% in males. Discrimination was moderate for patients aged 40-69 years and with CKD stage G1-G2, with a C-index ranging from 0.619 to 0.694. Model performance diminished (C-index < 0.6) among the elderly(≥70) and those with advanced CKD (G3-G4). In conclusion, SCORE2-Diabetes is generalizable in Hong Kong; the original model provides optimal calibration. Given the limited performance of the current models in elderly and advanced CKD populations, dedicated risk models for these groups should be developed in the future.
