Development and Validation of an Early Warning System for Septic Shock in Intensive Care Unit

In the Department of Critical Care Medicine, septic shock is a typical critical illness, with the characteristics of difficult discovery and high mortality. What’s more, doctors cannot conduct real-time observation and diagnosis for each patient due to high work pressure and heavy work burden. In order to solve this problem, we developed a septic shock prediction model and implemented it in a septic shock early warning system for the intensive care unit. The system has been applied and evaluated in clinical practice. The results of the evaluation show that the specificity of the early warning model at the optimal threshold reaches 0.95, which effectively helps doctors identify septic shock, intervene for the treatment of patients in advance, and reduce the damage caused by septic shock.