This study aims to evaluate the clinical effectiveness of the oXiris® hemofilter in patients with septic shock and acute kidney injury (AKI) requiring continuous renal replacement therapy (CRRT). While some studies suggest that oXiris® may improve hemodynamic stability by removing endotoxins and cytokines, its impact on survival remains inconsistent.To provide more reliable causal evidence, this multicenter retrospective cohort study utilizes a Target Trial Emulation (TTE) approach. The study will compare patients treated with the oXiris® filter to those treated with standard CRRT filters, focusing on in-hospital mortality and other clinical outcomes such as fluid balance and vasopressor requirements.

Background:

Septic shock is characterized by a dysregulated host response to infection, leading to a "cytokine storm" that causes organ dysfunction and hemodynamic instability. Extracorporeal blood purification (EBP) using the oXiris® filter is designed to restore immune homeostasis by adsorbing both endotoxins and cytokines while performing standard CRRT functions. However, existing observational data often suffer from selection and immortal time biases.

Study Design and Methodology:

This study employs the Target Trial Emulation (TTE) framework to mimic the design of a randomized controlled trial using retrospective observational data.

• Target Population: Adult patients (≥19 years) admitted to the ICU with septic shock and AKI requiring CRRT between March 2021 and August 2025.
• Time Zero: Defined as the initiation of CRRT.
• Treatment Strategies:

Strategy A (oXiris): CRRT initiated with the oXiris® filter. Strategy B (Control): CRRT initiated with a standard filter, sourced from the Korea Sepsis Alliance registry.

• Data Sources: Data will be collected from five major medical centers in South Korea: Seoul National University Bundang Hospital, Hallym University Sacred Heart Hospital, Chung-Ang University Hospital, Samsung Medical Center, and Asan Medical Center.
• Statistical Analysis Plan:

To adjust for baseline confounding and confounding by indication, propensity score-based Inverse Probability of Treatment Weighting (IPTW) will be applied.

Primary Outcome: In-hospital mortality will be analyzed using an IPTW-weighted Cox proportional hazards model.

Secondary Outcomes:

• Fluid Balance: 3-day cumulative fluid balance using landmark analysis and IPTW-weighted linear regression.
• Vasopressor Weaning: Analyzed using a weighted Fine-Gray risk model, considering death as a competing risk.
• Organ Function & Inflammation: Repeated measures of SOFA scores and inflammatory markers (CRP, Procalcitonin, lactate) using IPTW-weighted linear mixed-effects models.
• Renal Recovery: Evaluated via IPTW-weighted logistic regression.

Inclusion Criteria:

• Adults aged 19 years or older.
• Admitted to the ICU with septic shock (Sepsis-3 criteria) between March 2021 and August 2025.
• Required Continuous Renal Replacement Therapy (CRRT) due to acute kidney injury (AKI).

Exclusion Criteria:

• Death or withdrawal of life-sustaining therapy within 48 hours of ICU admission.
• Patients with End-Stage Renal Disease (ESRD).
• Patients with missing critical covariates.

The study population consists of adult patients (aged 19 or older) diagnosed with septic shock and acute kidney injury (AKI) who required continuous renal replacement therapy (CRRT). The treatment group includes approximately 120 patients who received the oXiris® hemofilter at five tertiary academic hospitals in South Korea between March 2021 and August 2025. The control group comprises approximately 240 patients who received standard CRRT filters, identified from the Korea Sepsis Alliance (KSA) registry, a nationwide multicenter sepsis cohort. All subjects are selected based on the availability of electronic medical records (EMR) and registry data that include critical clinical covariates for Target Trial Emulation.

This is a retrospective study using electronic medical records and registry data. According to the institutional policy and data use agreements, individual participant data cannot be shared publicly to protect patient privacy.