Multivariable haemodynamic approach to predict the fluid challenge response: A multicentre cohort study

Area of Science:

• Critical Care Medicine
• Cardiovascular Physiology
• Hemodynamic Monitoring

Background:

• Beat-to-beat stroke volume (SV) is determined by left ventricular function and arterial load.
• Fluid challenges trigger dynamic changes in cardiac performance and vascular characteristics.

Purpose of the Study:

• To determine if early hemodynamic response determinants during fluid challenge can predict final fluid responsiveness at 10 and 30 minutes.
• To develop a predictive model for fluid responsiveness in ICU patients.

Main Methods:

• An observational multicentric cohort study involving 85 ICU patients with acute circulatory failure.
• A fluid challenge of 500ml Ringer's solution over 10 minutes was administered.
• Hemodynamic parameters including SV, pulse pressure variation (PPV), arterial elastance, systolic-dicrotic pressure difference (SAP-Pdic), and cardiac cycle efficiency (CCE) were measured.

Main Results:

• A multivariable model using baseline PPV and early changes in arterial elastance, CCE, and SAP-Pdic accurately classified responders and non-responders at 10 minutes (80.5% and 90.7%).
• A separate model incorporating changes in PPV, CCE, SAP-Pdic, and arterial elastance predicted the 30-minute response with high accuracy (93.3% responders, 91.4% non-responders).

Conclusions:

• A statistical model analyzing changes in PPV, CCE, arterial elastance, and SAP-Pdic reliably predicts both early and late responses to fluid challenges in mixed ICU populations.
• This predictive model aids in optimizing fluid management strategies for critically ill patients.