The goal of this randomized clinical trial is to evaluate whether balanced gelatin solution is more effective and safe than balanced crystalloid solution for perioperative fluid management in adults with sepsis undergoing emergency abdominal surgery. Sepsis often causes severe fluid loss from the bloodstream into tissues, leading to low blood pressure, impaired organ function, and the need for urgent fluid resuscitation. Balanced gelatin, a colloid solution, may help maintain intravascular volume more effectively than crystalloid alone.

In this study, participants are randomly assigned in a 1:1 ratio to receive either balanced gelatin or Ringer's acetate during surgery and in the first 24 hours afterward. All patients receive standardized anesthesia care, goal-directed fluid therapy, and protocolized use of vasoactive drugs. The main questions the study aims to answer are:

• Does balanced gelatin reduce positive fluid balance within 24 hours after surgery?
• Does it improve hemodynamic stability during the early postoperative period?
• What effects does balanced gelatin have on kidney function, microcirculation, postoperative recovery, and other clinical outcomes?

Participants will be followed throughout hospitalization and contacted again on postoperative day 28 and day 90 to assess survival, complications, and health-related quality of life. The trial is double-blind, meaning that patients, clinicians, and outcome assessors do not know which fluid is being used. An independent Data and Safety Monitoring Board will oversee patient safety during the study.

The findings of this trial are expected to provide important evidence to guide perioperative fluid resuscitation strategies for septic patients undergoing emergency surgery.

Study Background Sepsis is a life-threatening syndrome caused by a dysregulated host response to severe infection, which can rapidly progress to septic shock, multiple organ dysfunction, and death.Despite advances in recognition and treatment in recent years, sepsis remains a major global health challenge, with overall mortality rates still ranging from 18% to 30%.Pathophysiologically, sepsis is characterized by increased capillary permeability and vasodilation, resulting in substantial fluid extravasation into the interstitial space, tissue edema, and intravascular volume depletion. These changes lead to inadequate tissue perfusion, impaired oxygen delivery, and ultimately organ dysfunction. Therefore, timely and appropriate fluid resuscitation remains one of the most fundamental and essential components of sepsis management. Individualized, goal-directed fluid therapy aims to improve systemic perfusion while minimizing the risks associated with fluid overload.

In 2001, Rivers et al. introduced the concept of early goal-directed therapy (EGDT), demonstrating significant mortality reduction (30.5% vs 45%) in patients with severe sepsis and septic shock.This finding prompted widespread adoption of early aggressive fluid resuscitation. However, three large multicenter trials published between 2014 and 2015 (ARISE, ProCESS, and ProMISe) later found no significant mortality differences between EGDT and usual care (approximately 18-25%).These results suggested that with improvements in routine clinical practice, contemporary usual care may already include adequate hemodynamic optimization.

Based on this evidence, the 2016 Surviving Sepsis Campaign (SSC) guidelines recommended administration of at least 30 mL/kg of crystalloid fluid within the first 3 hours for patients with sepsis-induced hypoperfusion, followed by hemodynamic-guided adjustment.Because many participants in these large trials had already received similar volumes before randomization, this dosage became associated with favorable outcomes. The 2021 SSC guideline retained this recommendation, though downgraded from a "strong" to a "weak" recommendation in the absence of new high-quality evidence.This highlights persistent uncertainty and the need for further research on optimal fluid type and strategy in sepsis resuscitation.

Increasing attention has been paid to microcirculation in recent years. While macrocirculatory parameters such as blood pressure may normalize after resuscitation, this does not necessarily indicate restoration of tissue-level perfusion.Persistent microcirculatory dysfunction may contribute to inadequate oxygen delivery and delayed organ recovery. Understanding how different fluid therapies influence both macro- and microcirculation is therefore crucial for optimizing outcomes.

Currently used resuscitation fluids include crystalloids and colloids. Crystalloids rapidly redistribute into the interstitial space and may exacerbate tissue edema-particularly in sepsis where capillary permeability is increased-potentially impairing microcirculatory flow.Colloids contain larger molecules that remain intravascular for longer periods, generating oncotic pressure and maintaining circulating volume more effectively, which may theoretically support microcirculatory perfusion and oxygen delivery. Clinically used colloids include albumin, hydroxyethyl starch (HES), and gelatin. Albumin is effective but costly and limited in availability. HES has fallen out of favor due to its association with kidney injury and increased mortality in sepsis.Gelatin, derived from bovine collagen hydrolysates, is now the only artificial colloid still recommended by guidelines for hypovolemia in sepsis patients.Nevertheless, high-quality randomized controlled evidence comparing gelatin with crystalloids in sepsis remains insufficient.

Based on this background, the present study focuses on balanced gelatin solution-a compound solution containing 4% succinylated gelatin in a balanced crystalloid carrier. We hypothesize that, compared with balanced crystalloid alone, balanced gelatin may better support hemodynamic stability through its colloid osmotic effect, more effectively correct fluid imbalance, improve microcirculatory perfusion, promote organ function recovery, and ultimately improve clinical outcomes. This randomized controlled trial is designed to rigorously test this hypothesis.

Patient and Public Involvement (PPIE)

During the protocol development stage, this study incorporated Patient and Public Involvement (PPIE). Three public contributors without medical backgrounds were invited to review the full protocol and informed consent form prior to study initiation. They provided feedback from the perspective of typical patients and family members. The main areas of concern included:

• The need for more accessible and patient-friendly language when describing the study background and objectives.
• Difficulty understanding the double-blind design, with recommendations to use analogies or simplified explanations.
• High concern regarding safety issues, including allergic reactions, kidney injury, and fluid-related adverse events, and a desire for clear descriptions of emergency procedures and contact information.
• Interest in treatment stability, postoperative supportive care needs, long-term complications, and follow-up arrangements.
• Questions regarding treatment-related costs, the exploratory nature of the intervention, and whether future protocol optimization is planned.
• Expectation that overall study results will be shared with participants after study completion.

Based on this feedback, the research team has revised and optimized the protocol background, the explanation of "double-blind" in the informed consent form, the safety management procedures, the follow-up plan, and the approach for disseminating study results. These adjustments aim to enhance the clarity, acceptability, and overall engagement of patient participants.

Study Objectives Primary Objective To evaluate the effectiveness of balanced gelatin solution in perioperative fluid management for septic patients undergoing emergency non-cardiac surgery, with emphasis on:reducing perioperative positive fluid balance；promoting hemodynamic stability；improving microcirculatory perfusion；and supporting organ function recovery and clinical outcomes.

Secondary Objective To assess the safety of balanced gelatin solution in this population, focusing on its effects on kidney function, coagulation status, and common postoperative complications.

Study Design This is a prospective, multicenter, randomized, double-blind, controlled clinical trial with an adaptive design incorporating sample size re-estimation. An initial sample size of 318 patients (159 per group) will be recruited, with an interim analysis after enrollment of 50% of participants to reassess the required sample size. Patients are randomized in a 1:1 ratio to receive either balanced gelatin or crystalloid solution (Ringer's acetate). Randomization is stratified by baseline blood lactate level (≤ 4 mmol/L vs > 4 mmol/L) using a central dynamic allocation system to ensure balance between groups.

Population Eligible patients are adults (≥18 years) with sepsis (per Sepsis-3 criteria) due to abdominal infection requiring emergency surgery. Inclusion requires a SOFA score ≥2 and lactate >2 mmol/L. Key exclusions include prior colloid use within 24 hours, expected death within 48 hours, advanced heart failure, severe ARDS, pre-existing renal replacement therapy, severe coagulopathy, liver failure, or allergy to gelatin.

Interventions All participants receive standardized anesthesia care. Intraoperative fluid therapy follows a goal-directed fluid therapy (GDFT) protocol guided by stroke volume (SV) monitoring. To ensure double-blinding, the study fluid (either balanced gelatin or Ringer's acetate) is prepared by independent, unblinded personnel, remaining masked to clinicians, investigators, and patients. During anesthesia, all patients receive a baseline infusion of Ringer's acetate at 3 mL/kg/h (ideal body weight, IBW). For volume expansion, fluid boluses (3 mL/kg IBW of the assigned study fluid over 5 minutes) are administered; a bolus is repeated if SV increases by >10%. Postoperatively, the assigned study fluid remains the primary resuscitation fluid for up to 24 hours. The cumulative dose of the study fluid is capped at 30 mL/kg (IBW) within this 24-hour intervention period; if the limit is reached, subsequent resuscitation reverts to Ringer's acetate. Vasoactive or inotropic support is standardized based on predefined hemodynamic triggers after optimizing SV.

Endpoints

Primary endpoints include:

Cumulative fluid balance within 24 hours after surgery. Proportion of patients achieving hemodynamic stability within 24 hours. Secondary endpoints include kidney function, SOFA score dynamics, lactate clearance, need for vasopressors or renal replacement therapy, postoperative complications, ICU and hospital length of stay, and all-cause mortality at 28 and 90 days. Safety endpoints include pulmonary edema, arrhythmias, and acute kidney injury.

Follow-up Patients will be followed during hospitalization and by structured telephone interviews at day 28 and day 90. Follow-up assessments include survival, complications, and health-related quality of life using the EQ-5D-5L questionnaire.

Blinding and Oversight The study is double-blind: patients, treating clinicians, outcome assessors, and statisticians remain unaware of group allocation. Randomization and drug packaging are handled by independent, unblinded coordinators under the supervision of independent witnesses to ensure masking integrity. Strict role isolation is maintained, ensuring that any personnel with access to allocation information are prohibited from participating in subject recruitment, clinical intervention, follow-up, or data analysis. Emergency unblinding is permitted only for patient safety. An independent Data and Safety Monitoring Board (DSMB) will oversee trial conduct, perform interim analyses for the adaptive design, and review adverse events.

Statistical Note:

A fixed-sequence hierarchical testing procedure will be applied to the primary endpoints to maintain the overall family-wise type I error rate at 0.05. Primary endpoint 1 (24-hour cumulative fluid balance) will be tested first; primary endpoint 2 (hemodynamic stability) will only be formally tested for statistical significance if the first endpoint reaches p < 0.05.

Significance By directly comparing balanced gelatin with crystalloids in septic patients undergoing emergency abdominal surgery, this trial will provide critical evidence regarding the efficacy and safety of gelatin-based fluid resuscitation. Results are expected to inform perioperative fluid management strategies and contribute to guideline development in the management of sepsis.

Inclusion Criteria:

• Age ≥ 18 years.
• Diagnosis of sepsis according to the Sepsis-3 definition, caused by intra-abdominal infection, and requiring emergency abdominal surgery for source control.
• Sequential Organ Failure Assessment (SOFA-2) score ≥ 2 (determined by the most recent available clinical data) and blood lactate > 2 mmol/L (measured within 6 hours prior to enrollment).
• Subject or legal representative can understand the study purpose and provide written informed consent.

Exclusion Criteria:

• Received any dose of artificial colloids within 24 hours prior to randomization.
• Predicted mortality within 48 hours (ASA physical status class ≥ V).
• Atrial fibrillation or congestive heart failure.
• Severe Acute Respiratory Distress Syndrome (ARDS).
• Preoperative coagulation dysfunction or receiving anticoagulant therapy.
• Preoperative requirement for renal replacement therapy (long-term or intermittent, including hemodialysis or peritoneal dialysis).
• Acute burns exceeding 10% of the total body surface area.
• Severe hepatic impairment.
• Severe electrolyte disturbance.
• Pregnancy or breastfeeding.
• Known allergy to gelatin.
• Current participation in other interventional clinical trials.
• Other conditions that the investigator considers inappropriate for inclusion.

At this time, the study team has not finalized an IPD sharing plan. This is a multicenter trial in critically ill patients; data governance, site ownership, and de-identification procedures are still under review. If sharing is pursued, it would be limited to a de-identified dataset sufficient to reproduce key findings-not the complete raw source data (e.g., free-text notes or imaging). Any decision will be guided by participant privacy, consent language, local/national regulations, and approvals by the sponsor/IRB/DSMB. Aggregate results will be reported in publications and on ClinicalTrials.gov.

This trial uses a double-blind design. Participants, treating clinicians, investigators, outcome assessors, and statisticians are all blinded to treatment allocation. To ensure masking, study fluids are packaged in identical opaque containers, prepared and labeled by independent unblinded coordinators who are not involved in subject recruitment, clinical intervention, follow-up, or data analysis. A double-check system involving an independent witness is implemented during the drug packaging and labeling process. Strict role isolation is maintained throughout the study to prevent any bias in treatment execution or outcome assessment. Emergency unblinding is strictly limited to patient safety emergencies and must be documented.