Construction of death early-warning model for patients with septic myocardial depression: a retrospective analysis of 129 patients / 中华危重病急救医学

ABSTRACT

Objective To explore the death risk factors of septic myocardial depression (SMD) and their predictive effect, and to set up a death early-warning model. Methods A retrospective analysis was conducted. The patients with SMD admitted to emergency department and rescue room of Beilun Branch of the First Affiliated Hospital of Zhejiang University Medical College from January 2015 to November 2017 were enrolled. The patients were divided into survival group and non-survival group according to 28-day outcome, and the gender, age, and the initial examination parameters [white blood cell (WBC) count, neutrophil (Neut) count, activated partial thromboplastin time (APTT), procalcitonin (PCT), D-dimer, C-reactive protein (CRP), cardiac troponin I (cTnI), N-terminal pro-brain natriuretic peptide (NT-proBNP), left ventricular ejection fraction (LVEF), left ventricular end-diastolic diameter (LVEDD), and left atrium diameter (LAD)] of both groups were compared. Binary logistic regression analysis was conducted on the factors with statistically significant difference analyzed in univariate analysis, and death early-warning model was set up subsequently. For parameters in early-warning model after variable screening, receiver operating characteristic curve (ROC) was applied to evaluate the predictive effect of death. Results A total of 129 patients were enrolled, 34 patients died within 28 days with the mortality of 26.4%. Univariate analysis showed that the PCT, cTnI and NT-proBNP in non-survival group were significantly higher than those of the survival group. However, there was no statistical difference in gender, age, WBC, Neut, APTT, D-dimer, CRP, LVEF, LVEDD or LAD between the two groups. Logistic stepwise regression analysis showed that PCT and cTnI were the independent factors influencing the death of patients with SMD [PCT odds ratio (OR) =1.495, 95% confidence interval (95%CI) = 1.192-1.876, P = 0.001; cTnI OR = 11.154, 95%CI = 5.709-17.264, P = 0.004], and the death early-warning model was logP =-3.737+0.402×PCT+2.412×cTnI. According to the statistics of Homser-Lemeshow, the effect of this model was good (χ2= 6.258, P = 0.617). The analysis of ROC displayed that the area under ROC curve (AUC) of the combination of PCT and cTnI for predicting the prognosis of SMD patients was 0.851, and it was significantly higher than that of PCT and cTnI alone (0.738 and 0.719, respectively, both P < 0.05). When the combination of PCT and cTnI was 0.26, the sensitivity was 79.97%, the specificity was 87.01%, the positive predictive value was 71.3%, and the negative predictive value was 91.7%. Conclusions PCT and cTnI are independent factors influencing the death of SMD patients. The combination of PCT and cTnI has predictive value for the prognosis of SMD patients. The death early-warning model of SMD patients can be used to predict the prognosis of SMD patients.