Mimic-Ⅲ-based Analysis Of Clinical Characteristics And Prognostic Correlation In Patients With Sepsis

OBJECTIVE: To analyze the prognostic differences in sepsis patients with different anatomical sources of infection,tissue perfusion,and immune status and to explore their relevance to prognosis.METHODS: Retrospective analysis of medical record information from the Medical Information in Intensive Care database(MIMIC-III)of adult(≥18 years)patients with primary sepsis admitted to the ICU for the first time.A total of 1540 patients were included in this study,and sepsis patients were divided into different anatomical sources of infection(lung,gallbladder,pancreas,kidney,etc.),different tissue perfusion statuses(with or without hyperlactatemia,refractory hypotension,refractory hypotension with hyperlactatemia,normal tissue perfusion)and different immune states(solid organ transplantation(SOT),hematopoietic stem cell transplantation,etc.)according to different subgroups.Mortality rates between subgroups of different anatomical sources were analyzed descriptively,using multifactorial Cox regression analysis to determine the correlation between mortality rates at 28 days of admission for different anatomical sources of infection.Kaplan-Meier survival curves were used to show 28-day mortality rates for sepsis patients admitted to hospitals in different immune states and tissue perfusion states.RESULTS:1.The overall 28-day mortality rate for sepsis patients included in the study was20.5%,with spontaneous peritonitis having the highest mortality rate of 61.9%,followed by pulmonary and visceral perforation with 26.4% and 25.6%,respectively,and pyelonephritis having the lowest mortality rate(7.7%),with significant differences in mortality rates between anatomical sources of infection(P<0.001),and the differences remaining significant after adjusting for confounding factors significant(P<0.001).2.Of the included immune status criteria [solid organ transplantation(SOT),hematopoietic stem cell transplantation,glucocorticoids,etc.],only SOT was statistically significant(P<0.05),with mortality rates of 9.9% and 21.0% for SOT and non-SOT sepsis patients,respectively.3.Mortality was 29.8% in sepsis patients with refractory hypotension and 14.2% in those without refractory hypotension,a significant difference(P<0.001);mortality was26.0% in sepsis patients with hyperlactatemia and 16.1% in sepsis patients without hyperlactatemia(P<0.001).When refractory hypotension and hyperlactatemia were present together,the patient’s morbidity and mortality rate was 37.6%(P<0.001).CONCLUSIONS:Short-term mortality in patients with sepsis varies markedly by the anatomical origin of infection,immune status,and tissue perfusion status.Exploring scientific clinical typing or stratification can help assess the clinical prognosis of sepsis and help advance precision medicine and scientific research.