Pharmacokinetic models improve prediction of intraoperative blood pressure drops
Pharmacokinetic State Space Models for Unbiased Prediction of Haemodynamic Collapse
Machine Learning
Summary
Low blood pressure during surgery under general anesthesia can cause serious problems, but current tools only react after it happens. The authors developed a new predictive model that uses information about how anesthesia drugs are infused to better forecast upcoming blood pressure drops. Their model performs better than random guessing and shows that drug levels give early warning signs before actual blood pressure changes occur. The model also maintains stable computer memory use, making it practical for continuous monitoring during surgery.
What this means in practice
- •For anesthesiologists: Predict upcoming dangerous drops in patient blood pressure using pharmacokinetic data during surgery to enable earlier intervention.
- •For clinical device manufacturers: Develop continuous monitoring systems that use drug infusion histories to provide real-time warnings of intraoperative hypotension.$Commercial implications: Enables creation of predictive anesthesia monitoring devices that improve patient safety and could be sold to hospitals and surgical centers.
Authors
Rithin Nagaraj, Sudiksha Chindula, Bhaskarjyoti Das
Abstract
An Intraoperative Hypotension (IOH) event is a frequent complication during administration of general anaesthesia with serious downstream consequences, yet clinical management remains reactive and not predictive. Existing predictive models, however, ignore drug infusion history as a valuable signal for prediction despite its direct pharmacological relevance. Our model achieves an Area Under the Receiver Operating Characteristic curve (AUROC) of 0.7360 and an Area Under the Precision-Recall Curve (AUPRC) of 0.1794, representing a 2.73-fold lift over the random guessing AUPRC baseline (0.0657), with the removal of propofol and remifentanil effect-site concentrations resulting in a 13.9% AUPRC drop compared to the full model. This is consistent with the hypothesis that pharmacokinetic trajectories encode impending haemodynamic changes before they manifest in the Mean Arterial Pressure (MAP). Additionally, this paper shows that training without lead-gap filtering degraded AUROC by 16.7%, empirically confirming that unfiltered models learn to detect ongoing hypotension rather than predict future events. Finally, a Mamba-based architecture achieves the aforementioned high prediction performance while maintaining a constant memory footprint across a range of sequence lengths, unlike the quadratic VRAM overhead typical of vanilla Transformers, making it the more practical choice for continuous intraoperative deployment.