Restrictive versus liberal fluid therapy in geriatric patients undergoing major abdominal surgery: a randomized controlled trial

BACKGROUND: The optimal fluid therapy in elderly patients undergoing major abdominal surgery remains unclear. Although some trials have reported a restrictive fluid therapy may lead to better outcomes, there is no evidence whether it is suitable for elderly patients. METHODS: In a double-blinded pragmatic trial, 107 elderly patients undergoing major abdominal surgery were randomized to receive either a liberal (L group) or restrictive intravenous-fluid therapy (R group) . The postoperative fluid therapy was similar in the two groups. The primary outcome was vital organ injury included Myocardial injury after noncardiac surgery (MINS) and acute kidney injuryAKI, the secondary outcomes included hypotension needed intervention intraoperatively, length of stay (LOS) , death and other complications defined up to 30 days. Analysis was performed by intention-to-treat. RESULT: 50 patients in the L group had an average intravenous fluid of 1943ml, as compared to 1295.61ml in 57 patients in the R group (P<0.001). The baseline Characteristics and operative details were similar between the groups. Patients in the L group had a lower rate of AKI (10% vs 35.1%, P=0.002) and surgical-site infection (0 vs 10.5%, P=0.029) than in the R groupMINS (20% vs 20.8%, P=0.724) and the other postoperative complications showed no differences between two groups. One patient died in the R group. No significant difference was found for the length of hospital staymedian(range) L: 15(8-49) vs R: 17(8-80); P=0.27. The follow-up was 30 days. CONCLUSION: In geriatric patients undergoing major abdominal surgery, a liberal fluid regimen was associated with a lower rate of AKI and postoperative infection than restrictive fluid regimen and did not increase the risk of postoperative complications. Trial registration: Registered 2018

Some trials reported that a restrictive fluid therapy led to fewer complications and shorten LOS [1,2]. Recent perioperative guidelines and consensus statements also supported restrictive fluid therapy [3,4]. Nevertheless, the evidences for restriction of perioperative fluid therapy is inconclusive [5][6][7] and these findings were not always replicated. A restrictive intravenous-fluid regimen might increase rate of acute kidney injury (AKI) [5] and it is more likely to treat hypotension with vasopressor therapy, impairing the perfusion of organ.
Historically perioperative fluid management tended to more liberal. Since the 1950s, when the so called 'third space losses' was claimed that fluids are redistributed to a third space, perioperative fluid management should replace of such losses with a large amount of crystalloid. Besides, clinicians administer liberal volumes of fluids perioperative for concerning about the preoperative dehydration and unstable of hemodynamics intraoperatively.
Such liberal fluid infusion is also controversial by increasing the risk of cardiopulmonary complications, enteral edema, nausea and vomiting and prolonging LOS [8]. Elderly patients are vulnerable to excessive fluid management because of their decline of multisystem physiologic function. Too much intravenous-fluid might cause disfunction of vital organs and 4 led to more postoperative complications.
Since the most effective fluid regimen is unclear and the ubiquity of fluid therapy in major abdominal surgery, we conducted this trial to test the liberal (more traditional) against restrictive fluid therapy in elderly patients.

Methods
We enrolled 107 patients who had planned for major abdominal surgery in Guangdong provincial people's hospital between November 2018 and April 2019. Patients were eligible if they planned for major abdominal surgery and at least 65 years old, an expected operative duration of at least 2 hours and 3 days of hospital stay. Urgent or time-critical surgery, liver resection, or less extensive surgery (laparoscopic cholecystectomy or closure of colostomy) , or patients with end-stage kidney failure requiring dialysis were excluded.
Patients allergic to artificial colloid fluid or experienced intraoperative hemorrhagic shock were also excluded. Written informed consent was obtained from all participants before the day of surgery (Fig 1) . The trial has been approved by the Research Ethics committee of Guangdong Provincial People's Hospital (NO. GDREC2018575H) and registered in Chinese Clinical Trial Registry (ChiCTR1800019022) .
Patients were randomized to either restrictive fluid therapy group (R group) or liberal fluid therapy group (L group) in the preparing room by one assigned anesthetist, so the patients and surgeons were kept blinded. The randomization sequence was made by another research member and delivered in sealed consecutively numbered envelopes. The blinding of all research staff members were kept until the end of the trial, but the attending anesthetist had knowledge of the group assignments in order to perform different intravenous-fluid therapy during operation.

Trial treatments
Perioperative surgery care was conducted in line with enhanced recovery principles.
Mechanical bowl preparation was discouraged; all patients received prophylactic antibiotics and were allowed to drink clear liquid before 2 hours of surgery [4].
If the patient underwent open surgery, transversus abdominis plane block was performed with ropivacaine. Patients were monitored with electrocardiogram (ECG) , heart rate(HR) , non-invasive blood pressure, SpO2, end tidal CO2 and body temperature. Central venous pressure and arterial pressure monitoring was left to the discretion of the anesthetist and hypothermia was avoided by employing routine intraoperative warming strategies.
The liberal fluid therapy was defined as more traditional infusion therapy [5,9]

Postoperative care
Standardized postoperative care was provided on surgery ward in both groups, administration to the Intensive Care Unit was at the discretion of the surgeon and anesthetist. Oral intake and early mobilization was encouraged when the patient was tolerated. Pain after surgery was treated with flurbiprofen, non-steroidal anti-inflammatory drug, or dezocine as needed.
If more fluid were required, balanced saline solution was administered on physician preference and modified fluid gelatin was permitted to restore volume if hypovolemia was occurred.
Primary outcomes was vital organ injury included Myocardial injury after noncardiac surgery (MINS) and acute kidney injury AKI , the secondary outcomes included hypotension needed 7 intervention intraoperatively, length of stay (LOS) , death and other complications defined up to 30 days. Diagnosis criteria was according to kidney disease: improving global outcomes (KDIGO) [12] or needed to renal replacement therapy (RRT). Secondary outcomes were hypotension needed intervention, other complications up to 30 days and LOS. Patients had blood collected for High Sensitivity Troponin T (Hs-TnT) assay preoperatively and on days 1, 2 after surgery for detecting MINS [13]. A complication was recorded if clinical treatment was necessary (table 2)  optimization of flow-related variables may be the best approach for fluid administration, but some studies [14,15] showed a totally different conclusion and did not encouraged to apply GDT in every patient.
Traditional fluid therapy can deliver up to 7 liters on the day of surgery in order to make up for the loss of preoperative fasting, anesthesia-induced vasodilation and accumulation of fluid in extravascular spaces [16,17]. Previous some trials [18][19][20] showed that such liberal fluid therapy could lead to tissue edema and increased interstitial lung water, leading to further complications. ERAS (enhanced recovery after surgery) pathway is a clinical care guideline proposed in last 90th [21,22], aiming at attenuating surgery stress, promoting early recovery among patients undergoing major abdominal surgery and decreasing the Length Of Stay (LOS) .
Avoiding too much fluid or a more restrictive fluid regimen is recommended in recent ERAS guidelines [23,24]. But these evidence seem not inconclusive in major abdominal surgery [18,25,26]. Especially, too restriction of fluid regimen may increase the rate of AKI and hypotension perioperatively [27]. AKI in the perioperative period is a common complication in geriatric patients associated with increased morbidity and mortality. Recent update on perioperative AKI review [28] emphasized adequate perioperative fluid to ensure hemodynamic stability and the perfusion of kidney. Oliguria could be the response for either hypovolemia or normal activity of antidiuretic hormone. In our trial, no difference was found (2% vs 3.5%, P=1) in oliguria during surgery between groups. Some studies reported that it shouldn't be use as an indication for renal injury [29], nor is there any evidence that shows diuretics protect against AKI [30]. But it is associated with adverse outcomes [28] and too restriction of fluid therapy might increase the need for vasopressor and worsen the perfusion of organs. The diagnosis of AKI postoperatively in our trial, therefore, had combined both urine output and serum creatine. Providing sufficient perfusion of kidney will be the key point in avoiding AKI perioperatively and that might be explained why liberal group did have a low risk of AKI in our trial(5% vs 20%, P=0.002), which was similar to the previous trial [5]. We also identified a higher rate of surgical-site infection in the R group, possibly caused by the hypoperfusion of anastomotic or the wound.
Myocardial injury after noncardiac surgery (MINS) is defined as myocardial injury that occurs during or within 30 days after surgery, an independently risk factor for long-term outcome [13], and it is more common in elderly patients perioperatively. Recent perioperative guidelines [31,32] have recommended patients >65 years measure troponin.
It is rather common in elderly patients. Previous studies have found that an elevated Hs-TnT is associated with long-term mortality even without ischemic symptom [13]. In our trial, MINS was diagnosed by an elevated Hs-TnT change of 5 ng/L or higher. Our first hypothesis was that liberal fluid therapy could increase the injury of cardia in elderly patients and MINS in the L group might be more obvious than R group. However, no significant difference was found (20% vs 20.8%, P=0.724 ) in our trial, nor difference were found in other complications up to 30days. But it should not to be the evidence to support the traditional fluid regimen, it should be explained that relatively liberal fluid therapy is safer than restriction ones and may have more strength in elderly patients, not increase the risk of MINS and other complications.

Limitations of the study
Our trial has several limitations. First, our anesthetist took part in this trial could not be blinded to the procedure, and this might introduce bias in documentation and some outcome monitoring. Second, we didn't separate the laparoscopic surgery from open surgery. One study [33] showed no difference outcome in different fluid strategies and there were no evidences existed that patients undergoing laparoscopic surgery should be treated differently from open surgery. Third, we did not analysis the amount of fluid included 24hours postoperatively when most intravenous fluid is given and that might cause some bias in outcomes and many patients could not be weighed on days 1 to 3. Finally, these results are limited to shorty-term outcomes, however, some authors have demonstrated the impact of early postoperative complications on long-term outcomes [28,34].

Conclusions:
In Conclusion, in elderly patients undergoing major abdominal surgery, a liberal fluid therapy was associated with a lower rate of AKI than restrictive fluid therapy and did not increase the risk of postoperative complications.  Outcome data are presented as value(%) or mean and difference. Indicates significant results with P values<0.05. ※ MINS diagnosis was followed as a elevation of Hs-TnT more than 5ng/l, or an absolute Hs-TnT was large than 14ng/ml. † AKI(acute kidney injury) defined as KDIGO guideline: either an increase of serum creatinine or a reduced urine output: In an increase of serum creatinine ≥0.3 mg/dL (26.5 µmol/L) within 48 h or an increase of serum creatinine by ≥1.5-fold above baseline, known or assumed to have occurred within 7 days, 21 or urine volume <0.5 mL/kg/h for 6 h. * data was collected among 99 patients CONSORT Flow Diagram.doc