Gelaspan vs Crystalloid Therapy in Sepsis

Primary Outcome 1: Net fluid balance within 24 hours after surgery

\[Hierarchical Testing Sequence: 1st\] This is the first primary endpoint in the hierarchical testing sequence. Net fluid balance is defined as the difference between total infused volume and total output volume during surgery and the first 24 postoperative hours. Input volume includes all study fluids (balanced gelatin solution or acetate Ringer's), albumin, blood products, and maintenance crystalloid infusion. Excluded are solvent volumes \<50 mL and non-therapeutic fluids such as irrigation or enteral/oral intake. Albumin is recorded in mL of solution administered. Blood product volumes are standardized: packed red blood cells 1 unit = 200 mL; plasma = actual volume; apheresis platelets 1 therapeutic dose = 250 mL; cryoprecipitate 1 unit = 25 mL (with center-specific adjustment allowed). Output volume includes intraoperative blood loss, urine output, and measurable drainage (thoracic, abdominal, nasogastric, etc.), excluding insensible or unmeasurable losses.

Primary Outcome 2: Proportion of patients achieving hemodynamic stability within 24 hours after surgery

\[Hierarchical Testing Sequence: 2nd\] This is the second primary endpoint in the hierarchical testing sequence. Hemodynamic stability (HDS) is defined as meeting all of the following three criteria at postoperative 24 hours: 1. Mean arterial pressure (MAP) ≥65 mmHg without vasopressor support, sustained ≥1 hour after discontinuation of vasopressors, with all subsequent MAP measurements ≥65 mmHg. 2. Blood lactate ≤2 mmol/L, based on the most recent venous or arterial sample. 3. Urine output ≥1 mL/kg/h, based on the average over the preceding 6 hours. The outcome measure is the percentage of patients in each group who fulfill all three criteria.

Secondary Outcome 1.1: Intensity of study drug use within 24 hours after surgery

Cumulative volume of study drug administered from randomization to 24 hours postoperatively, standardized by ideal body weight (mL/kg).

Secondary Outcome 1.2: Blood product utilization rate within 24 hours after surgery

Proportion of patients receiving any blood product during the intervention period, regardless of type. Each patient is counted once even if multiple blood products are used.

Secondary Outcome 1.3: Vasopressor load within 24 hours after surgery

Vasopressor dose converted to norepinephrine-equivalent according to predefined equivalence table (norepinephrine:epinephrine:dopamine:phenylephrine:vasopressin = 1:1:0.01:0.01:0.02), normalized by ideal body weight and infusion time (μg/kg/min).

Secondary Outcome 1.4: Proportion of patients receiving inotropic drugs within 24 hours after surgery

Proportion of patients treated with any inotrope (dobutamine, milrinone, levosimendan, etc.) within 24 hours postoperatively.

Safety Outcome 1.1: Incidence of postoperative respiratory complications (ARDS)

ARDS defined as PaO₂/FiO₂ ratio \<200 at any arterial blood gas measurement within 7 days postoperatively. Patients without postoperative blood gas data are excluded. Deaths without ARDS evidence are counted as negative.

Safety Outcome 1.2: Incidence of postoperative cardiovascular complications (arrhythmias)

New-onset arrhythmias within 28 days postoperatively, including supraventricular tachycardia, ventricular tachycardia, or atrial fibrillation/flutter. Multiple events in same patient count once.

Safety Outcome 1.3: Incidence of postoperative renal complications

Defined as new renal dysfunction (serum creatinine ≥2× baseline) or need for renal replacement therapy (RRT) within 28 days after surgery. One patient counts once regardless of multiple events.

Safety Outcome 2.1: Health-related quality of life at postoperative day 90

Health-related quality of life will be assessed at postoperative day 90 using the EuroQol 5-Dimension 5-Level questionnaire (EQ-5D-5L). The descriptive system covers five dimensions (mobility, self-care, usual activities, pain/discomfort, anxiety/depression), each rated on 5 levels (1 = no problems, 5 = extreme problems). Responses are converted to a health state code (e.g., 21321), which is then mapped to a utility index score using the Chinese EQ-5D-5L value set (range: -0.391 to 1.000, where 1.000 = full health and values \<0 represent states worse than death). In addition, participants will provide a self-rated health score on the EQ-VAS, a vertical visual analog scale ranging from 0 (worst imaginable health) to 100 (best imaginable health).