Predicting 24-Hour Blood Pressure Variability Post Thrombectomy Using Machine Learning for Patients with Ischemic...

Area of Science:

• Neurology
• Cardiology
• Data Science

Background:

• Mechanical thrombectomy is standard for ischemic stroke with large vessel occlusion.
• Elevated blood pressure variability (BPV) post-procedure correlates with poor functional outcomes.
• Identifying predictors of BPV is crucial for improving patient management.

Purpose of the Study:

• To identify predictive factors for increased 24-hour blood pressure variability (BPV) post-mechanical thrombectomy.
• To utilize machine learning for identifying key predictors of BPV.
• To enable proactive management of BPV in at-risk stroke patients.

Main Methods:

• Retrospective analysis of 395 patients undergoing mechanical thrombectomy (2016-2019).
• Primary outcome: BPV comparison between adequate (TICI 2b+) and inadequate reperfusion groups.
• Random forest analysis to identify BPV predictors; multivariable regression for secondary outcomes (functional status, reperfusion).

Main Results:

• Higher age, NIHSS, number of passes, and mechanical ventilation were linked to poorer 90-day functional outcomes (mRS ≤2).
• Machine learning identified time from last-known-well to groin puncture, age, and NIHSS as significant predictors of increased 24-hour BPV.
• 82% of patients achieved adequate reperfusion (TICI 2b+).

Conclusions:

• Time to groin puncture, age, and NIHSS are key predictors of post-thrombectomy BPV.
• Machine learning models can identify high-risk patients for BPV.
• Early identification allows for targeted blood pressure management to improve stroke outcomes.