Disease severity modifies the ceftazidime-avibactam exposure-response relationship in critically ill patients: a retrospective study

Background Ceftazidime-avibactam (CAZ-AVI) is a key agent for MDR-GNB, but significant pharmacokinetic (PK) variability in critically ill patients poses challenges to achieving optimal dosing. Objectives To investigate the association and influencing factors between CAZ-AVI serum concentration and clinical efficacy in critically ill patients, and specifically to evaluate the influence of disease severity on this relationship, with the aim of identifying a therapeutic target for critically ill patients with different disease severity. Methods A retrospective study was conducted of 168 critically ill patients receiving ceftazidime/avibactam between 2021 and 2025. The relationship between plasma drug concentration, clinical efficacy and disease severity was evaluated. Results 168 patients were included. C min of CAZ and AVI was significantly higher in the effective group than in the ineffective group (CAZ: 39.68 mg/L vs. 26.11 mg/L, P < 0.001; AVI: 8.29 mg/L vs. 4.92 mg/L, P = 0.003). Multivariate analysis identified that CAZ C min was an independent positive predictor (P < 0.05) and SOFA score on TDM day as an independent negative predictor (P < 0.05) of clinical efficacy. Stratification by SOFA score revealed that this exposure-response relationship was markedly modified by disease severity. In the subgroup with less severely ill patients (SOFA score<6), the positive association between CAZ concentration and clinical efficacy was markedly strengthened (P < 0.001), with ROC analysis identifying an optimal predictive CAZ threshold of 34.79 mg/L (AUC = 0.829). Regarding safety, the overall incidence of adverse drug reactions (ADRs) was 15.03%, primarily elevated transaminases, acute kidney injury, and hematological abnormalities. A concentration-dependent risk for ADRs was established, with a CAZ concentration of 54.84 mg/L identified as the optimal predictive cut-off. Conclusion This retrospective analysis provides evidence that optimizing CAZ-AVI plasma concentrations is crucial for improving outcomes in critically ill patients, and that this relationship is significantly modulated by disease severity, necessitating a severity-adapted therapeutic drug monitoring (TDM) strategy.