Excessive fluid loss is often reported in gastrointestinal surgical patients due to preoperative fasting and bowel preparations. Insufficient fluid infusion may cause hypovolemia and tissue hypoperfusion, which may delayed postoperative recovery and even induce postoperative acute renal failure. The aim of this study is to compare the effects of Goal-directed fluid therapy (GDFT) strategy with that of the conventional fluid management on the morbidity and mortality of postoperative complications, length of postoperative hospital stay, and medical expense, so as to provide clinical evidences for optimized intraoperative fluid management for patients undergone gastrointestinal surgery.

I. Background Excessive fluid loss is often reported in gastrointestinal surgical patients due to preoperative fasting and bowel preparations. Many factors may further exacerbate the lack of fluid volume, including intraoperative blood loss and fluid loss, anesthetic drug-induced peripheral vasodilatation, as well as systemic inflammatory response syndrome, systemic capillary leak syndrome, endothelial barrier dysfunction and transfer of intravascular fluid to the third space caused by surgical trauma.

Insufficient fluid infusion may cause hypovolemia and tissue hypoperfusion, which may delayed postoperative recovery and even induce postoperative acute renal failure. On the other hand, fluid overload can lead to edema, increased oxygen diffusion distance between lung mesenchyme and cells, then causing tissue ischemia, hypoxia, and acidosis and even affecting the recovery of gastrointestinal function and the healing of incision. It has been proved that improper fluid therapy in the perioperative period is associated with poor outcomes in patients with gastrointestinal surgery. Goal-directed fluid therapy (GDFT) regimen emphasizes the personalized fluid infusion to optimize the intraoperative hemodynamics and tissue perfusion, including rising stroke volume, increasing arterial oxygen saturation, which may result in improving postoperative outcomes.

Flo-Trac/Vigileo system is developed by Edwards Technology, and monitors the continuous cardiac output via peripheral arteries. The monitoring approaches of this system prove to be convenient, safe and minimally invasive. As an important indicator of the system, stroke volume variation (SVV) can accurately reflect the volume status of patients. The principle of SVV is the preload change, which is caused by the intrathoracic pressure changes between inspiratory and expiratory phases, will lead to arterial pulse pressure fluctuations, thus affecting stroke volume (SV). SV may vary under different volume loads. Through measuring and analyzing these differences and then assessing the circulation status, SVV provides guidance for fluid therapy. The continuous and real-time SVV monitoring can provide data of patients' circulation status, thereby optimizing clinical fluid treatment options and avoiding excessive or insufficient volume. For patients with gastrointestinal surgery, SVV guidance during the surgery can accurately regulate fluid infusion, which will contribute to the early postoperative recovery and improve patient prognosis.

In the present study, SVV-directed fluid therapy will be performed in patients with gastrointestinal surgery and then will be compared with conventional fluid treatment regimen in the mobility and mortality of postoperative complications, postoperative recovery, and length of hospital stay. The investigators are aiming to provide clinical evidences for reasonable intraoperative fluid management of such patients.

II. Objectives The aim of this study is to compare the effects of GDFT strategy with that of the conventional fluid management on the morbidity and mortality of postoperative complications, length of postoperative hospital stay, and medical expense, so as to provide clinical evidences for optimized intraoperative fluid management for patients undergone gastrointestinal surgery.

III. Research Plan 1. Design: quantitative observational comparative effectiveness research. 2. Sample Volume: 200 cases of subjects in the experimental group and control group, respectively.

3. Research phases The research is divided into three phases

1. Retrospective analysis of perioperative data from patients undergone gastrointestinal surgery according to traditional practices from January to June 2015. Data about postoperative length of hospital stay, the morbidity and mortality of complications is collected from electronic database. The fluid was administrated according to the traditional method. Total intake volume = compensatory volume of expansion (CVE) + physically required volume+ cumulative loss volume + continued loss volume + the loss volume of the third space. Both physically required volume and cumulative loss volume were replaced according to 4-2-1 rule. Crystalloid and colloid were given with the ratio of 2/1. According to the bleeding, RBC and plasma were given if necessary. HR50 /min~100/min and MAP60~100mmHg were maintained during the surgery in both groups. When SBP<90 mmHg or MAP < 60 mmHg, the patients were given the vasoactive drug of bitartrate noradrenaline.
2. Training phase: the GDFT protocol training will be implemented after receiving ethic approval estimated from July to September 2015. A 3-month training period allowed all research staff to become familiar with the protocol.
3. Implementation phase: After 3 months training, estimated from October 2015 till March in 2016, the fluid management is conducted to patients with the gastrointestinal surgery on the basis of GDFT protocol. Then perioperative data of the included patients in this phase will be collected retrospectively from electronic database.

4. GDFT protocol

1. Anesthesia management The radial artery line is placed before anesthesia. The invasively arterial blood pressure, ECG, HR and SPO2 are monitored. The Flo-Trac/Vigileo monitoring system is connected with the radial artery line of the GDFT group, CO, SV, SVV and other hemodynamic parameters are monitored. In addition, the bispectral index (BIS) is continuously monitored to assess the depth of sedation.

   Induction: 0.01~0.03 mg/kg of midazolam, 0.2~0.4μg/kg of sufentanil, 1~2 mg/kg of propofol and 0.6~0.9 mg/kg of rocuronium are injected intravenously for anesthesia induction. After the completion of endotracheal intubation, the internal jugular puncture catheterization is undergone at the right side to monitor the central vein pressure.

   Maintenance: After endotracheal intubation, the mechanical ventilation is initiated. The tidal volume is set as 6~8 mL/kg, respiratory rate as 10 to 14 times/min and inhalation to exhalation ratio as 1:2. In the surgery, the partial pressure of carbon dioxide in end expiratory gas is kept at the level of 35 - 45 mmHg. And the remifentanil (target concentration of effect compartment is 4 - 8 ng / mL) and propofol (target concentration of plasma is 3 - 5μg / mL) of intravenous target-controlled infusion are given to maintain the anesthesia effect. BIS value is maintained at 40 - 50. Besides, 0.1~0.3μg/kg sufentanil and 0.1~0.2 mg/kg rocuronium are injected intermittently. About 30 minutes before the end of the surgery, a bolus of 0.1~0.3μg/kg sufentanil is administered.

   Recovery: The depth of anesthesia is reduced gradually at the end of the surgery. In the recovery period, the atropine and neostigmine are given as muscle relaxant antagonists. Extubation is performed according to the awareness, breathing, muscle strength and reflex recovery of the patient. All patients are delivered to ICU after surgery.

2. Fluid Treatment Compensatory volume of expansion (CVE) is replaced with 5mL/kg of Ringer's lactate solution before induction. The maintained dose of infusion of crystalloid during surgery is: 2~4 ml/kg/h for open procedure (maximum dose of 400ml/h), and 1~2 ml/kg/h for laparoscopic procedure (maximum dose of 200ml/h) respectively. Fluid replacement is directed by SV value. After incision (after pneumoperitoneum for laparoscopic cases) give a 200 mL Hydroxyethyl Starch 130/0.4 and Sodium Chloride bolus over <10 min. If SV increases by >10%, repeat bolus until SV increases by <10%. Record peak value achieved. Give a further colloid bolus when SV drops 10% from peak value and repeat cycle. According to the bleeding, RBC and plasma are given if necessary. SaO2≥95%, Hb > 70g/L, core body temperature> 37 ℃, HR50 /min~100/min and MAP60~100mmHg are maintained during the surgery in both groups. When SBP<90 mmHg or MAP < 60 mmHg, the two groups are given the vasoactive drug of bitartrate noradrenaline.

5. Monitoring variables

1. Intraoperative intake and output volume: volume of intraoperative crystalloid, volume of colloid, blood transfusion volume, urine output, blood loss and use of vasoactive agents.
2. Hemodynamic variables: MAP, HR, SaO2, CVP, CO, CI, SV, SVV, stroke volume index (SVI) and systemic vascular resistance (SVR) at the following time points: baseline, immediately after induction and intubation, at beginning of the surgery and every five minutes following that, at the end of surgery, etc.

Inclusion Criteria:

1. Elective laparotomy or laparoscopic gastrointestinal surgery.
2. Age ≥ 18 years old.
3. ASA grade II - IV.
4. General anesthesia.

Exclusion Criteria:

1. Patients with aortic regurgitation.
2. Patients with major artery stenosis disease, peripheral vascular disease and arterial catheterization contraindication.
3. Patients with cognitive dysfunction and uncooperative subjects.
4. Failure to obtain informed consent

patients undergone gastrointestinal surgery