Early Recognition and Dynamic Risk Warning System of Multiple Organ Dysfunction Syndrome Caused by Sepsis

Early Recognition and Dynamic Risk Warning System of Multiple Organ Dysfunction Syndrome Caused by Sepsis

Background Sepsis still the main challenge of ICU patients, because of its high morbidity and mortality. The proportion of sepsis, severe sepsis, and septic shock in china were 3.10%, 43.6%, and 53.3% with a 2.78%, 17.69%, and 51.94%, of 90-day mortality, respectively.

Besides, according to the latest definition of sepsis- "a life-threatening organ dysfunction caused by a dysregulated host response to infection. ", it is a disease with intrinsic heterogeneity. Sepsis as a syndrome with such great heterogeneity, there will be significant differences in the severity of sepsis. As a result, there will be significant differences in the treatment and monitoring intensity required by patients with severe sepsis and mild sepsis. No matter from the economic perspective or from the risk of treatment, a proper level of treatment will be the best chose of patient. However, the evaluation of the sepsis severity was not satisfied. Such of SOFA, the AUC of predict patients' mortality was only 69%. Weather these patients occurred multiple organ dysfunction syndrome (MODS) may had totally different outcome and needed totally different treatment. All these treatments need early interference, in order to achieve a good prognosis. Hence, early recognition of MODS caused by sepsis became an imperious demand.

Study design On the base of regional critical medicine clinical information platform, a multi-center, sepsis big data platform (including clinical information database and biological sample database) and a long-term follow-up database will be established. Thereafter, an early identification, risk classification and dynamic early warning system of sepsis induced MODS will be established. This system was based on the real-time dynamic vital signs and clinical information, combined with biomarker and multi-omics information. And this system was evaluated sepsis patients via artificial intelligence, machine learning, bioinformatics analysis techniques.

Finally, optimize the early diagnosis of sepsis induced MODS, standardized the treatment strategy, reduce the morbidity and mortality of MODS through this system.

Inclusion Criteria:

• Patients diagnosed with sepsis3.0

Exclusion Criteria:

• Patients' data missing is greater than 20%