Precision critical care management of blood pressure in stroke patients using dynamic linear models

For stroke patients in the ICU, optimal blood pressure management remains an open area of research. Numerous observational studies and clinical trials on both ischemic and hemorrhagic stroke have resulted in conflicting evidence for the benefit of lowering blood pressure. A major limitation of these studies is their inability to account for and distinguish the effects of physician-initiated blood pressure treatment versus a patient's spontaneous blood pressure time course. We address this problem with the Acute Intervention Model of Blood Pressure (AIM-BP) framework: an individualized, human interpretable model of blood pressure management in the acute care setting. The framework consists of two components: one, a model of blood pressure homeostasis and the various effects that perturb it; and two, a parameter estimator that can learn clinically important model parameters on a patient by patient basis. By estimating the parameters of the AIM-BP model for a given patient, the effectiveness of antihypertensive medication can be quantified separately from the patient's spontaneous blood pressure trends. We hypothesize that the AIM-BP is a sufficient framework for estimating parameters of a homeostasis perturbation model of a stroke patient's blood pressure time course and the AIM-BP parameter estimator can do so more accurately and consistently than a state-of-the-art maximum likelihood estimation method. The first part of this hypothesis is proved mathematically, while the second is demonstrated using simulated clinical scenarios modeled on stroke patients from two ICU datasets. Finally, the ability of the AIM-BP framework to model real world patients is demonstrated using several examples from a UPMC dataset.