Psychological Factors Do Not Improve SMART2 Prediction of Recurrent CV Events
Overview
In a study of 20,050 patients with established atherosclerotic cardiovascular disease, adding psychological factors such as depression, anxiety, and insomnia to the SMART2 risk model did not significantly improve prediction of 10-year recurrent cardiovascular events. The SMART2 model reliably estimated risk regardless of psychological comorbidities, with minimal changes in discrimination and calibration.
Background
Patients with established atherosclerotic cardiovascular disease are at high risk for recurrent events, and risk prediction models like SMART2 guide treatment decisions. Psychological factors including depression, anxiety, and insomnia are common in this population and have been associated with increased cardiovascular risk and lower treatment adherence. Prior evidence suggested that incorporating psychological factors might improve risk prediction, potentially preventing undertreatment. This study evaluated whether adding psychological variables to the SMART2 model enhances its predictive performance for recurrent cardiovascular events.
Data Highlights
Parameter
Prevalence (%)
ΔC-statistic (95% CI)
Diagnosis of psychological factors
3–9
Minimal, not significant
Self-reported symptoms
3–45
−0.0003 (−0.0005 to 0.0001)
Prescribed treatment
10–14
0.0011 (0.0011 to 0.0011)
Key Findings
Among 20,050 patients from five European cohorts, 2,987 experienced recurrent cardiovascular events over 10 years.
Psychological factors were present in 3–45% depending on assessment method (diagnosis, symptoms, treatment).
No psychological factor significantly improved prediction of recurrent events beyond the SMART2 model.
Changes in C-statistic with addition of psychological factors ranged from −0.0003 to 0.0011, indicating negligible impact.
Calibration of the SMART2 model remained adequate in patients with psychological conditions.
SMART2 reliably estimates recurrent cardiovascular risk regardless of depression, anxiety, or insomnia status.
Clinical Implications
Clinicians can confidently use the SMART2 risk model for 10-year recurrent cardiovascular event prediction in patients with established atherosclerotic disease, irrespective of coexisting psychological conditions. Incorporating depression, anxiety, or insomnia into risk calculations does not meaningfully enhance predictive accuracy and is therefore not necessary for clinical decision-making. This supports continued reliance on SMART2 without modification for psychological factors.
Conclusion
The SMART2 model provides robust and reliable risk estimates for recurrent cardiovascular events in patients with atherosclerotic disease, even in the presence of psychological comorbidities. Adding psychological factors does not improve predictive performance, affirming SMART2's utility across diverse patient profiles.
References
Authors/Source/2024 -- Exploring Psychological Influences Beyond the SMART2 Model in Predicting Recurrence of Events in Patients with Atherosclerotic Cardiovascular Disease
by Joris Holtrop, Mari Nordbø Gynnild, Toril Dammen, Håvard Dalen, Ingvild Saltvedt, Hanne Ellekjær, Steven H J Hageman, Frank L J Visseren, John William McEvoy, Jannick A N Dorresteijn, John Munkhaugen, On behalf of the UCC-SMART study group
