Linear regression model for continuous non-invasive blood pressure estimation for an ICU setting

Abstract

Continuous blood pressure (BP) monitoring in critically ill patients is crucially important for clinicians. The gold standard intra-arterial method is complicated and contraindicated. Current pulse transit time (PTT)-based BP estimation methods are validated on healthy subjects using intermittent cuff-based BP measurements. Research indicates significant differences in invasive and noninvasive BP recordings in critically ill patients. Research also indicates BP variations due to clinical condition induced heart rate changes and gender and age differences. Comprehensively addressing all these issues, a novel, continuous non-invasive BP estimation model with age, gender and heart rate as covariates along with PTT to estimate ICU patient invasive BP data is presented in this pilot-study. The proposed novel model is tested for four common ICU clinical conditions. The results indicate compliance with the Association for Advancement for Medical Instrumentation standards, thus providing a potential solution for the underlying risk-factors associated with the conventional method.