Dynamic Prognosis Prediction for Patients on DAPT After Drug‐Eluting Stent Implantation: Model Development and Validation

Q: What are the key findings of this study?

The AI-DAPT model achieved peak accuracy in the 30 to 36 months prediction window, yielding an AUROC of 90% for ischemic events and 84% for bleeding events.

Q: What is the clinical evidence from this study?

Study design: Cohort. Population: Post-drug-eluting stent implantation on dual antiplatelet therapy (n=108214). Intervention: AI-DAPT dynamic prediction model vs. DAPT score. Primary outcome: Discriminative accuracy (AUROC) for ischemia prediction in the 30 to 36 months window (AUROC 90%, 95% CI 88-92).

Fang Li; Laila Rasmy; Yang Xiang; Jingna Feng; Ahmed Abdelhameed; Xinyue Hu; Zenan Sun; David Aguilar; Abhijeet Dhoble; Jingcheng Du; Qing Wang; Shuteng Niu; Yifang Dang; Xinyuan Zhang; Ziqian Xie; Yi Nian; Jianping He; Yujia Zhou; Jianfu Li; Mattia Prosperi; Jiang Bian; Degui Zhi; Cui Tao

doi:10.1161/jaha.123.029900

Key result

AI-DAPT model predicts ischemic events with ~90% accuracy at 30 to 36 months.

AUROC 90%
95% CI 88-92
n=108,214

Figures

Why the study?

Recent updates in DAPT guidelines and the evolution of AI highlighted the need for innovative dynamic models to predict ischemic or bleeding events after drug-eluting stent implantation.

Does the AI-DAPT model improve the dynamic prediction of ischemic and bleeding events in patients on DAPT after drug-eluting stent implantation compared to standard clinical scores?

Key Points

To develop and validate a dynamic artificial intelligence pipeline capable of predicting ischemic and bleeding risks over time in patients receiving dual antiplatelet therapy after drug-eluting stent implantation.
Retrospective cohort analysis included adult patients undergoing drug-eluting stent implantation from the Cerner Health Facts data set (n=98 236) and Optum's Clinformatics Data Mart Database (n=9978).
Assessed five artificial intelligence algorithms across six sequential prediction windows spanning a 36-month follow-up interval after stenting.
The top-performing weighted light gradient boosting machine model (AI-DAPT) was calibrated and tested against traditional risk-scoring methods.
The AI-DAPT model achieved its highest discriminative accuracy in the 30 to 36 months prediction window, with an AUROC of 90% [95% CI, 88%-92%] for ischemia and 84% [95% CI, 82%-87%] for bleeding.
In the 12 to 30 months window, AI-DAPT outperformed the standard DAPT score for both bleeding (AUROC 77% versus 68%, P<0.0001) and ischemic risk prediction (AUROC 82% versus 50%, P<0.0001).
Across all six windows, AI-DAPT maintained superior mean AUROC compared with the DAPT score for ischemia (87% versus 49%) and bleeding (81% versus 69%).

Study Design

Type

Cohort (n=108,214)

Multicenter

Yes

Structured PICO

Does the AI-DAPT model improve the dynamic prediction of ischemic and bleeding events in patients on DAPT after drug-eluting stent implantation compared to standard clinical scores?

P

Population

108,214 adult patients from two large US retrospective databases who underwent drug-eluting stent implantation, followed for up to 36 months to train and validate a dynamic prediction model for ischemic and bleeding events.

E

Exposure

AI-DAPT, a dynamic artificial intelligence risk prediction model based on a weighted light gradient boosting machine (LGBM) algorithm, utilizing continuous clinical profile updates to predict events in sequential 6-month windows

C

Comparator

Standard DAPT score and a newly trained Cox regression-based score (c-DAPT)

O

Outcome

Dynamic prediction of ischemia (composite of acute ischemic heart disease, ischemic stroke, and coronary revascularization) and bleeding (minor to severe spontaneous bleeding and blood transfusion) evaluated in 6 sequential 6-month prediction windows from 0 to 36 monthscomposite

Main Result

Effect estimate: AUROC 90% (95% CI 88-92)

Limitations

Inability to confirm certain information (e.g., stent characteristics, severity of bleeding, cause of death) using structured EHR data alone
Reliability of ICD codes-based outcome definitions without complete validation
Follow-up information gaps if patients did not return to the same hospital
Discrepancy between prescribed medications and actual patient adherence

Implication

May support dynamic post-DES risk stratification; leaves open prospective validation against DAPT score before practice change.

Abstract

BACKGROUND: The rapid evolution of artificial intelligence (AI) in conjunction with recent updates in dual antiplatelet therapy (DAPT) management guidelines emphasizes the necessity for innovative models to predict ischemic or bleeding events after drug-eluting stent implantation. Leveraging AI for dynamic prediction has the potential to revolutionize risk stratification and provide personalized decision support for DAPT management. METHODS AND RESULTS: We developed and validated a new AI-based pipeline using retrospective data of drug-eluting stent-treated patients, sourced from the Cerner Health Facts data set (n=98 236) and Optum's de-identified Clinformatics Data Mart Database (n=9978). The 36 months following drug-eluting stent implantation were designated as our primary forecasting interval, further segmented into 6 sequential prediction windows. We evaluated 5 distinct AI algorithms for their precision in predicting ischemic and bleeding risks. Model discriminative accuracy was assessed using the area under the receiver operating characteristic curve, among other metrics. The weighted light gradient boosting machine stood out as the preeminent model, thus earning its place as our AI-DAPT model. The AI-DAPT demonstrated peak accuracy in the 30 to 36 months window, charting an area under the receiver operating characteristic curve of 90% [95% CI, 88%-92%] for ischemia and 84% [95% CI, 82%-87%] for bleeding predictions. CONCLUSIONS: Our AI-DAPT excels in formulating iterative, refined dynamic predictions by assimilating ongoing updates from patients' clinical profiles, holding value as a novel smart clinical tool to facilitate optimal DAPT duration management with high accuracy and adaptability.

Discussion

No takes yet. Share an insight, caveat, or question.

Cite This Study

Li et al. (2024) conducted a cohort in Post-drug-eluting stent implantation on dual antiplatelet therapy (n=108,214). AI-DAPT dynamic prediction model vs. DAPT score was evaluated on Discriminative accuracy (AUROC) for ischemia prediction in the 30 to 36 months window (AUROC 90%, 95% CI 88-92). The AI-DAPT model achieved peak accuracy in the 30 to 36 months prediction window, yielding an AUROC of 90% for ischemic events and 84% for bleeding events.

synapsesocial.com/papers/6a20a465aa8e57945c6d9927 https://doi.org/10.1161/jaha.123.029900

View Full Paper

Ask AI

Bookmark

Share

Dynamic Prognosis Prediction for Patients on DAPT After Drug‐Eluting Stent Implantation: Model Development and Validation

Figures

Key Points

Study Design

Structured PICO

Main Result

Limitations

Implication

Abstract

Discussion

Cite This Study

Dynamic Prognosis Prediction for Patients on DAPT After Drug‐Eluting Stent Implantation: Model Development and Validation

Figures

Key Points

Study Design

Structured PICO

Main Result

Limitations

Implication

Abstract

Discussion

Cite This Study

Keep reading

Keep reading