# Clinical Characteristics and Prognostic Factors in Patients With Sepsis: A Retrospective Study

## TL;DR

This study identifies key factors affecting sepsis patient survival and shows how combining traditional Chinese medicine and Western medical indicators can improve prognosis predictions.

## Contribution

The study introduces a combined model integrating TCM and Western medical indicators for predicting sepsis mortality.

## Key findings

- CRP, TT, disease severity, CA, and ARDS were identified as independent risk factors for sepsis mortality.
- The combined model of indicators achieved the highest predictive accuracy (AUC > 0.5) and demonstrated good stability.
- Blood stasis syndrome showed significant prognostic value despite being excluded from the final model due to collinearity.

## Abstract

This retrospective study aimed to investigate the risk factors influencing the 28‐day clinical prognosis of sepsis patients and evaluate their predictive efficacy. Clinical data of patients diagnosed with sepsis between January 1, 2019, and December 31, 2023, were collected from the Hospital Information System (HIS) of Beijing Hospital of Traditional Chinese Medicine, Capital Medical University. Based on 28‐day outcomes, patients were divided into survival (n = 146) and death (n = 81) groups. Statistical analysis was performed using SPSS 20, employing univariate and multivariate logistic regression to identify prognostic risk factors. Receiver operating characteristic (ROC) curve analysis was conducted to assess the predictive performance of these factors, with the area under the curve (AUC) calculated for evaluation. Although blood stasis syndrome was not included in the final model due to collinearity with critical indicators, univariate analysis demonstrated its significant prognostic value (OR = 2.49, 95% CI 1.199–5.17, p = 0.014), and ROC curve analysis confirmed its fundamental discriminatory capacity (AUC > 0.5). Multivariable logistic regression identified CRP, TT, disease severity, CA and ARDS as independent risk factors for sepsis mortality. ROC analysis showed all individual indicators and the combined model had AUC > 0.5, with the combined model achieving the highest AUC. The combined model demonstrated good stability via Hosmer‐Lemeshow testing (p = 0.067). This study established CRP, TT, disease severity, CA and ARDS as independent mortality risk factors in sepsis, with the combined model showing optimal performance. It demonstrated consistency between TCM macro‐pattern differentiation and Western medical indicators, providing a framework for integrated prognostic models that combines both medical approaches.