Vancomycin therapeutic drug monitoring is associated with reduced toxicity in ICU patients: a MIMIC-IV retrospective study

Vancomycin is a first-line treatment for methicillin-resistant Staphylococcus aureus (MRSA) infections but is associated with risks of nephrotoxicity (5–43%), hepatotoxicity, and hematotoxicity. Therapeutic drug monitoring (TDM) is recommended to optimize dosing, yet its impact on multi-organ toxicity and mortality in intensive care unit (ICU) patients remains controversial because of conflicting evidence and methodological limitations in prior studies. Data were extracted from the Medical Information Mart for Intensive Care IV (MIMIC-IV, v3.1) database for a retrospective cohort analysis of 28,451 ICU patients receiving intravenous vancomycin. The primary outcomes were vancomycin-associated nephrotoxicity (AKI according to the KDIGO criteria), hepatotoxicity (ALT/AST ≥ 120 U/L or bilirubin ≥ 2.5 mg/dL), and hematotoxicity (thrombocytopenia, anemia, or leukopenia); secondary outcomes included ICU / hospital mortality. Propensity score matching (PSM, 1:1 nearest neighbor with caliper = 0.1) balanced 32 baseline covariates, including demographics. The associations between TDM and outcomes were evaluated via multivariable logistic regression and Cox proportional hazards models, with the results validated through subgroup analyses (stratified by comorbidities and concomitant medications) and sensitivity analyses. Data from 28,451 ICU patients receiving intravenous vancomycin were extracted from the MIMIC-IV database, with 10,758 (37.8%) receiving TDM and 17,693 (62.2%) not receiving TDM. Before PSM, the TDM group presented higher baseline illness severity scores (e.g., SOFA, APS III) and more comorbidities. Unadjusted analyses revealed increased risks of adverse outcomes in the TDM group (AKI: OR = 2.98, 95% CI: 2.83–3.15; hematotoxicity: OR = 1.97, 95% CI: 1.88–2.07; hepatotoxicity: OR = 2.34, 95% CI: 2.19–2.50; all P < 0.001). However, with progressive adjustment for confounders, these associations attenuated significantly (Model 3: AKI OR = 1.93, hematotoxicity OR = 1.55, hepatotoxicity OR = 1.25; all P < 0.001). After PSM, the TDM group demonstrated significantly reduced risks of AKI (OR = 0.580, 95% CI: 0.540–0.610, P = 0.001), hematotoxicity (OR = 0.760, 95% CI: 0.710–0.800, P = 0.001), and hepatotoxicity (OR = 0.800, 95% CI: 0.750–0.860, P = 0.001). Secondary outcomes also favored TDM, with lower in-hospital mortality (OR = 0.672, 95% CI: 0.570–0.790, P = 0.001) and ICU mortality (OR = 0.691, 95% CI: 0.580–0.820, P = 0.001). Kaplan-Meier analysis further confirmed the survival benefits of TDM in both ICU and hospital settings (log-rank P < 0.001). Subgroup analyses revealed that hypertension, type 2 diabetes mellitus (T2DM), cancer, cerebral bleeding (CB), and concomitant use of aspirin or antibiotics were significant risk factors for nephrotoxicity, hematotoxicity and hepatotoxicity. This study demonstrated that vancomycin TDM is significantly associated with reduced toxicity risks (nephrotoxicity, hepatotoxicity, hematotoxicity) and mortality in intensive care unit (ICU) patients, supporting its routine use in critically ill populations.