肝硬化伴肝功能失代偿失代偿的预后意义：一项回顾性病例对照研究

Acute-on-chronic liver failure (ACLF), particularly when associated with hepatitis B virus (HBV) infection, is characterized by rapid deterioration of liver function, systemic inflammation, and high short-term mortality. Artificial liver support systems, such as the double plasma molecular adsorption system combined with plasma exchange, are increasingly used as bridging therapies; however, reliable prognostic indicators remain limited. This study evaluated the prognostic value of systemic immune-inflammatory markers - systemic immune-inflammation index (SII), neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and C-reactive protein - in 100 patients with HBV-ACLF undergoing artificial liver therapy. Patients were categorized into improved and ineffective outcome groups, and inflammatory indices were measured before and after treatment. Posttreatment reductions in neutrophil-to-lymphocyte ratio, platelet-to-lymphocyte ratio, and C-reactive protein levels were significantly greater in the improved group (P < .05). While model for end-stage liver disease (MELD) scores differed significantly between the groups, SII alone showed no statistical difference. However, the combined use of the SII and MELD produced the highest prognostic accuracy (area under the curve = 0.854), outperforming either metric alone. In conclusion, our study suggests that the integration of inflammatory markers with traditional liver function scoring enhances predictive performance. Combining the SII with MELD provides a cost-effective, accessible, and dynamic prognostic tool for clinicians managing HBV-ACLF. This dual-marker approach enables better early risk stratification, informs clinical decision-making, and may guide the optimal timing of artificial liver therapy. The use of both indices may be particularly valuable in resource-limited or time-sensitive settings, where rapid treatment decisions are critical. Future prospective studies are warranted to validate these results and refine their clinical application.