Abstract: Objective: To investigate the risk factors on prognostic survival of patients with deep neck infection (DNI). Methods: Clinical materials from 196 patients that were admitted to the two principal hospitals in Xinjiang Uygur Autonomous Region between January 2010 and August 2024 due to DNIs were retrospectively analyzed. In addition to assessing patient’s demographic characteristics, the independent risk factors that could predict clinical prognosis for patients due to serious complications and long-term hospitalization were identified by performing analyses of the univariate and multivariate logistic regression model. Results: According to multivariate logistic regression analyses, the presence of comorbidities (P₁=0.013,P₂<0.001) and neutrophil-to-lymphocyte ratio (NLR) (P₁=0.005,P₂=0.008) were independent risk factors for severe complications and long-term hospitalization in patients with DNI. The presence of comorbidities had good diagnostic value for severe complications and long-term hospitalization in DNI patients, with area under curve (AUC) of 0.77 and 0.82, respectively. The NLR level also showed good evaluating performance, with AUC=0.81 and AUC=0.82, respectively. In addition, the AUC value for diagnosing severe complications based on white blood cell count (WBC) was 0.90; other AUC values for predicting long-term hospitalization (>10 days) of patients based on different origin of infection, C-reactive protein (CRP), and procalcitonin (PCT) levels were 0.69, 0.73, and 0.71, respectively. Conclusion: This present study found that the combination of variables for both complications and long-term hospitalization are promising for prognostic purposes.

Key words: deep neck infection, Logistic regression analysis, vacuum sealing drainage, prognostic variable