Use of preload responsiveness tests in pediatric intensive care units: A nationwide prospective clinical practice...

To avoid unnecessary volume expansions (VE), several preload-responsiveness tests exist in children but most have strict validity conditions, or are time-consuming. Although underuse or misuse of these tests could lead to inappropriate VE and fluid overload, their actual use in pediatric intensive care units (PICU) has never been evaluated. This study aimed to assess the use of preload-responsiveness tests before VE in PICU, identify clinical factors associated with their use, and document misuse when validity conditions, as defined in the literature, were not met.

We conducted a prospective nationwide observational study in 34 French PICU. VE prescriptions to children ≤15 years were collected during a 6-week period. Exclusion criteria were preterm neonates, mechanical circulatory support, and single-ventricle physiology. For each VE, physicians completed a questionnaire at bedside, focusing on preload-responsiveness test use before VE. Relevant data to interpret tests validity were also collected, including spontaneous breathing and ventilation parameters. The use of preload-responsiveness test, their misuse and the characteristics associated with their use were analyzed. Misuse of a test was defined as its application when validity criteria were not met, regardless of the method used to perform the test which was not recorded.

A total of 471 VE were analyzed, prescribed to children with a median age of 7.9 months, on catecholamine support in 202/471 (42.9%) cases, and intubated in 292/471 (62.0%) cases. The median volume was 10 ml.kg^-1. A preload-responsiveness test was performed before 296/471 (62.8%) VE (95%CI=[58.5%-67.2%]). Inferior vena cava visual analysis was the most common test, performed before 165 VE, followed by the respiratory variability of the peak aortic velocity (ΔVPeak), a dynamic index, measured before 92 VE. Among 294 dynamic indices (i.e., based on respiratory variation in hemodynamic signals) or preload challenges performed, misuse was observed in 188/294 (63.9%) cases. Performing a preload-responsiveness test was independently associated with mechanical ventilation (OR = 3.73, 95%CI=[1.95-7.15], p < 0.001) and catecholamine support (OR=4.08, 95%CI=[2.00-8.31], p < 0.001).

In this large national prospective study, preload-responsiveness tests were routinely used before VE in PICU, especially in severe patients. However, the selected tests were often either poorly reliable or misused.