Early piperacillin/tazobactam administration and sepsis-induced cardiomyopathy in emergency department patients: An observational study and exploratory prediction model

Abstract: Background: Sepsis-induced cardiomyopathy (SICM), a significant cardiovascular complication of sepsis, necessitates prompt and effective management. While timely antimicrobial treatment forms the bedrock of early sepsis management, the relationship between the timing of piperacillin/tazobactam administration and the development of SICM remains incompletely characterized. Objectives: The primary objective of this study was to investigate the association between early piperacillin/tazobactam administration and SICM incidence and short-term outcomes. A secondary objective was to identify factors associated with SICM and to develop an exploratory model for its prediction. Methods: This single-center, retrospective observational study included 149 adults with sepsis admitted between January 2023 and January 2026. Patients were categorized into two groups: Early treatment (≤1 h; n=68) or delayed treatment (>1 h; n=81). SICM was assessed using echocardiographic criteria. Clinical outcomes were compared using univariate and multivariable logistic regression. Internal validation was performed using 1,000 bootstrap resamples. Results: SICM occurred in 15/68 (22.06%) early-treatment patients and 33/81 (40.74%) delayed-treatment patients (P=0.014). Twenty-eight-day mortality was 14.71% versus 29.63% (P=0.029) and ICU and total hospital stays were shorter in the early-treatment group (both P<0.001). Delayed treatment (OR=2.354, 95% CI 1.044–5.310), SOFA ≥8 (OR=2.762, 95% CI 1.215–6.279), NT-proBNP ≥1500 pg/mL (OR=3.019, 95% CI 1.320–6.906) and each 0.1-ng/mL elevation in cTnI (OR=1.208, 95% CI 1.037–1.408) were independently linked to SICM. The model AUC was 0.842 (95% CI 0.776–0.908); Hosmer–Lemeshow P=0.409. Conclusion: Early administration of piperacillin/tazobactam was associated with lower SICM incidence and improved short-term outcomes. However, because this study is retrospective, a causal relationship cannot be definitively established. Furthermore, the exploratory prediction model developed warrants external validation to confirm its generalizability and accuracy.