Scientific rationale: The burden of stroke is a global problem, which is increasing despite some positive trends in incidence and mortality in high-income countries. Recently developed digital technologies in stroke prevention, such as the internationally endorsed, validated, and award-winning Stroke Riskometer app for laypeople and the PreventS-MD web app for health professionals, are promising solutions to address the gap in individual stroke prevention and have already shown their efficacy in multiple trials. However, their algorithm is derived from the Framingham Study, established last century, and is not specific enough for multi-ethnic populations, which limits its adoption in various countries. The development of country-specific coefficients for risk prediction algorithms is a complex problem that requires significant time and resources. However, there is a successful example of the adoption of the NZ population-derived CVD risk assessment algorithm to the Australian population. We plan to test this approach for the stroke prediction algorithm in the UK population, where gold-standard prospective data are available for comparison with the new approach.
Research question: How can the Stroke Riskometer algorithm be recalibrated and validated to accurately predict first-ever strokes across diverse populations
Study aims: (1) to recalibrate and validate the Stroke Riskometer/PreventS-MD prediction algorithm for first-ever strokes in different populations;
(2) to test approaches for the development of adjustment coefficients for the existing stroke prediction algorithm based on the available data on stroke incidence in different age/sex subgroups.
