Assisted Fluid Management (AFM) System and Postoperative Outcome After High-risk Abdominal Surgery — PEFLA  

Outcome Clinical Trial

Official title: "Personalized Fluid Administration Using an AFM System for Goal Directed Fluid Therapy on Postoperative Outcome in High-risk Patients Undergoing High-risk Abdominal Surgery: A Multicenter Stepped-wedge, Cluster-randomized Clinical Trial

Status Recruiting

Clinical Trial Summary

Goal directed fluid therapy (GDFT) or "Personalized fluid therapy" may benefit high-risk surgical patients but these strategies are infrequently implemented. It has also been shown that without any goal or protocol for fluid resuscitation, large inter- and intra-provider variability exist that have been correlated with poor patient outcomes.

Recently, an "Assisted Fluid Management" (AFM) system has been developed to help ease some of the work associated with GDFT protocol implementation. The AFM system may help increase GDFT protocol adherence while leaving direction and guidance in the hands of the care providers. This artificial intelligence-based system can suggest administration of fluid boluses, analyse the hemodynamic effects of the bolus, and continually re-assess the patient for further fluid requirements.

To date, there are no large outcome study using this AFM system. The primary objective of this trial is thus to evaluate the impact of this AFM system to guide fluid bolus administration on a composite of major postoperative complications in high-risk patients undergoing high-risk abdominal surgery.

Clinical Trial Description

Many trials have indicated that goal-directed fluid therapy (GDFT) strategies or more recently "personalized fluid therapy" may benefit high-risk surgical patients but these strategies are infrequently implemented. It has also been shown that without any goal or protocol for fluid resuscitation, large inter- and intra-provider variability exist that have been correlated with poor patient outcomes. Even under ideal study conditions, strict adherence to GDFT protocols is hampered by the workload and concentration required for consistent implementation. Hemodynamic monitors and protocols alone do not enable optimal fluid titration to be provided consistently to all patients - there must also be appropriate and timely interpretation and intervention.

To address this problem of consistency and protocol adherence, a decision support system, "Assisted Fluid Management" (AFM), has been developed to help ease some of the work associated with GDFT protocol implementation. The AFM system (released on the European market in March 2017) may help increase GDFT protocol adherence while leaving direction and guidance in the hands of the care providers. This artificial intelligence-based system can suggest administration of fluid boluses, analyse the hemodynamic effects of the bolus, and continually re-assess the patient for further fluid requirements.

This system was recently implemented in a before-and-after study in a Belgian academic hospital, where the authors reported that the implementation of this AFM software system allowed a better adherence to the GDFT algorithm. However, as this was a pilot study with a small number of patients, the study was not powered to demonstrate a beneficial effect on the incidence of postoperative complications. More recently, a group from the Cleveland Clinic demonstrated that using AFM system resulted in more boluses being effective when compared to the administration of boluses without AFM support.

There are no randomized controlled studies to date comparing this AFM system to standard of care on patient outcome. We therefore aim to conduct a multicenter stepped-wedge, cluster-randomized trial involving patients undergoing high risk abdominal surgery to compare a GDFT strategy guided by the AFM system with usual care.

A stepped wedge, cluster-randomized trial approach was chosen in which clusters will be randomized to commence the intervention at different times following an initial control period in which outcomes will be measured for usual care.

So, each center (cluster) began in the control phase and transitioned to the intervention phase at a randomly assigned time (wedge). The order in which each center will move from control to intervention phase will be randomly allocated by a computer algorithm performed by the study statistician.

We selected cluster randomization rather than randomization of individual patients because the control group could be very different among centers (from GDFT strategy using a flow monitoring to GDFT with a written protocol to no clear strategy (use of an arterial line only without any advanced monitoring). Interestingly, this approach also decreases the Hawthorne effect which has been shown to decrease the incidence rate of the primary outcome in recent randomized trials because clinicians know that their patients are included in a research protocol. ;

Related Conditions & MeSH terms

• Decision Support System
• Outcome

• NCT number: NCT06011187
• Study type: Interventional
• Source: Assistance Publique - Hôpitaux de Paris
• Contact: ALEXANDRE JOOSTEN, MD PhD
• Phone: +33145213441
• Email: joosten-alexandre@hotmail.com
• Status: Recruiting
• Phase: N/A
• Start date: February 1, 2024
• Completion date: December 30, 2026