Physicians' Academy for Cardiovascular Education

Causal AI improves accuracy of prediction of CV risk and expected benefit

News - Aug. 28, 2022

Causal AI substantially improves the validity of estimating cardiovascular risk and benefit

Presented at the ESC congress 2022 by: Brian Ference – Cambridge, UK

Introduction and methods

Causal effects are not included in current risk estimating algorithms, and therefore these algorithms do not accurately estimate CV risk caused by the two main modifiable causes of ASCVD events – LDL-c and SBP. Also, these algorithms do not accurately estimate the benefit of lowering LDL-c and SBP.

The incorporation of causal effects into algorithms leads to accurate estimates of baseline CV risk caused by LDL-c and SBP and of the benefit of lowering LDL-c, SBP or both. This would provide the essential information that is needed to inform patients for individual treatment decisions about timing, duration and intensity of lowering LDL-c and SBP for the optimal prevention of CV events.

AI algorithms can learn the effect of modifiable causes of disease that are targets of intervention, leading to causal AI. Causal AI results in AI algorithms, that can accurately predict risk and benefit.

Thus, the aim of this study was to create a new generation of AI algorithms that can be used for personalized medicine to guide individual treatment decisions.

Main results

Conclusion

The incorporation of causal effects of LDL-c and SBP by AI algorithms into risk estimating algorithms improved accuracy of estimates of baseline CV risk caused by LDL-c and SBP and the benefit of lowering LDL-c, SBP or both beginning at any age and extending for any duration. This helps to guide individuals to select optimal timing, duration and intensity of LDL-c and SBP lowering. Also, it provides the true clinical and economic benefit of preventing CVD to better inform clinical practice guidelines and health policies.

Share this page with your colleagues and friends: