Establishment and evaluation of a multivariate Cox proportional-hazards prediction model for mortality during short-term hospitalization in patients with liver cirrhosis and sepsis / 临床肝胆病杂志

ABSTRACT

Objective To establish a Cox proportional-hazards prediction model for mortality during short-term hospitalization in patients with liver cirrhosis and sepsis. Methods A retrospective analysis was performed for the clinical data of 336 patients with liver cirrhosis and sepsis who were admitted to The Third People's Hospital of Kunming from January 2012 to August 2022, and according to whether the patient died during short-term hospitalization, they were divided into death group with 40 patients and survival group with 296 patients. Demographic data, comorbidities, and clinical biochemical parameters were collected and compared between the two groups. The independent-samples t test was used for comparison of normally distributed continuous data between two groups, and the Mann-Whitney U test was used for comparison of non-normally distributed continuous data between two groups; the chi-square test was used for comparison of categorical data between two groups. The multivariate Cox analysis was used for screening of variables, then a Cox proportional-hazards prediction model was established, and hazard ratio ( HR ) and its 95% confidence interval [ CI ] were calculated; C-index index was used to evaluate the prediction accuracy of the model. The Cox proportional-hazards prediction model was visualized by a nomogram, and calibration curve was plotted to evaluate the consistency between the prediction results of the model and the actual condition. Results Among the 336 patients, there were 261 male patients (77.7%) and 75 female patients (22.3%), with a mean age of 50.0±10.6 years, and 40 patients died, with a mean hospital stay of 16.8±11.3 days (range 8.2-23.0 days). Compared with the survival group, the death group had a significantly higher proportion of patients with an age of ≥60 years, a history of invasive operation within the past two weeks, gastrointestinal bleeding, hepatic encephalopathy (HE) or hepatorenal syndrome (HRS), a significantly higher Modified Early Warning Score (MEWS) score, and significantly higher levels of prothrombin time (PT), activated partial thromboplastin time, international normalized ratio, D-dimer, CD4/CD8 ratio, lactate, white blood cell count, norepinephrine, total bilirubin, interleukin-6, procalcitonin, high-sensitivity C-reactive protein (hsCRP), blood urea nitrogen, and creatinine (all P < 0.05), as well as significantly lower levels of red blood cell count, hemoglobin, albumin, total cholesterol, low-density lipoprotein, and high-density lipoprotein (all P < 0.05). The multivariate Cox regression analysis showed that age ( HR =2.602, 95% CI 1.277-5.303, P =0.008), HE ( HR =2.516, 95% CI 1.258-5.033, P =0.009), HRS ( HR =2.324, 95% CI 1.010-5.349, P =0.047), hsCRP ( HR =1.008, 95% CI 1.003-1.013, P =0.004), MEWS score ( HR =1.205, 95% CI 1.022-1.422, P =0.027), and PT ( HR =1.076, 95% CI 1.030-1.124, P =0.027) were independent influencing factors for death in patients with liver cirrhosis and sepsis. The model showed a C-index of 0.857 (95% CI 0.815-0.920), suggesting that the model had relatively high prediction accuracy, and the calibration curve showed good consistency between the predicted risk and the actual risk. Conclusion The Cox proportional-hazards prediction model established for death during short-term hospitalization in patients with liver cirrhosis and sepsis can be used to predict the risk of death during short-term hospitalization in patients with liver cirrhosis and sepsis, thereby guiding clinical medical staff to take targeted intervention measures to avoid or reduce the possibility of death in patients.