Modeling and analysis of hospital inpatient rescue process: A Markov chain approach

Improving patient safety is the top priority for hospital management. On the hospital floor, an inpatient may experience clinical deterioration during his/her stay. Quick and appropriate treatment from the nurse, physician, and rapid response team (RRT) is essential to rescue the patient. In this paper, we introduce an analytical method to model and analyze the hospital inpatient rescue (HIR) process. A continuous time Markov chain model is presented to characterize the patient status and analyze the transitions between different patient states, such as risk, non-risk, intervention by the care provider, or elevation to intensive care, etc. Closed formulas to calculate the probability of the patient staying in different states are developed for single patient case. An approximation method, referred to as the shared resource iteration (SRI) approach, is proposed to study the multiple patients scenario. It is shown that such an iteration is convergent and results in a high accuracy estimation of patient state probability. This method provides a quantitative tool to analyze the HIR process and investigate strategies to improve patient safety.