Key result
The authors discuss the calibration of the SCORE2 Asia-Pacific model, emphasizing that age group-based plots best reflect their recalibration method for cardiovascular disease risk prediction.
Captured external expert commentary on this paper, strongest first. Original sources are linked where available.
“Rather than direct estimates of total CVD incidence, CVD mortality data were converted with 'multiplication factors', [given that] CVD morbidity data are very heterogeneous across countries [and that] data quality in most countries are not sufficient for recalibration. These estimates of CVD incidence were [then] used in combination with mean risk factor levels to recalibrate to every region.”
“previous work has demonstrated that SCORE2 could be successfully recalibrated for use in the Asia Pacific region, including China, by categorizing countries according to their WHO-reported CVD mortality rates. China was classified as a high-risk region. Remarkably, there was minimal heterogeneity between the European SCORE2 coefficients and risk factor effects in Asian populations. Following recalibration, SCORE2 performed well across 13 studies comprising over 9.5 million participants and 350,550 incident CVD events in Asia-Pacific cohorts.”
The authors defend their choice of age group-based calibration plots over objective tests like the Nam-D'Agostino test for the SCORE2 Asia-Pacific model, arguing that large datasets make such tests overly sensitive to clinically insignificant differences.
This commentary refers to ‘Risk prediction of cardiovascular disease in the Asia-Pacific region: the SCORE2 Asia-Pacific model’, by S.H.J. Hageman et al., https://doi.org/10.1093/eurheartj/ehae609 and the discussion piece ‘Risk prediction for cardiovascular diseases in Asia-Pacific: to separate subtypes or not, that is a new question’, by S. Yang and J. Lv, https://doi.org/10.1093/eurheartj/ehae854. In the commentary from Yang and Lv,1 the differences in stroke subtype incidence in Asia are discussed, as well as their consequences for risk prediction and subsequent treatment. We appreciate their engagement with our work and their valuable insights, particularly regarding the model’s performance and the considerations relevant to the Asia-Pacific region’s distinct cardiovascular disease (CVD) characteristics. The authors suggest to provide additional calibration metrics to assess the SCORE2 Asia-Pacific model’s performance. While we completely agree with the authors that the calibration is the most relevant measure for the clinical practice, it is our view that more objective tests, such as the Nam–D’Agostino test, may not provide additional value given our extensive data sources.2 With large datasets, these tests will nearly always indicate statistically significant differences, even when those differences lack practical or clinical significance. A plot in deciles is an alternative we had considered during the development of our model. However, given the strong relation between age and CVD incidence, these plots generally show a broadly similar result (see example Figure 1). We had chosen the age group-based plot as this best reflects our recalibration method, which was also based on 5-year age groups. This way, readers cannot only judge whether predicted risks match the observed incidence but also whether the recalibration efforts have succeeded over the whole age range in the respective risk region.
No takes yet. Share an insight, caveat, or question.
Hageman et al. (2025) conducted a letter in Cardiovascular disease. SCORE2 Asia-Pacific model was evaluated. The authors discuss the calibration of the SCORE2 Asia-Pacific model, emphasizing that age group-based plots best reflect their recalibration method for cardiovascular disease risk prediction.
Synapse has enriched 5 closely related papers on similar clinical questions. Consider them for comparative context: