Studying the Efficacy of and Developing Data-Driven Real-Time Clinical Decision Support Systems for Hypotension Detection

Abstract

Critically ill patients admitted into intensive care units are prone to reoccurring episodes of sustained hypotension. Prolonged durations of hypotension are correlated to, and potentially cause, permanent body-wide damage to patients if not properly treated, which may result in death. Currently, typical care for the management of hypotension in the critically ill is reactive and delayed, perhaps due to clinical inertia. The purpose of this study is to describe the current problem that is faced in critical care through a retrospective analysis and introduce candidate models that may be used as clinical informatics systems for preemptive hypotension detection to aid clinicians and nurses providing care in the fast-paced clinical environment. The clinical performance of the models is quantified and the efficacy of implementation of these models is discussed.