Preload Dependence Fails to Predict Hemodynamic Instability During a Fluid Removal Challenge in Children.

ABSTRACT

OBJECTIVES: Fluid overload increases morbidity and mortality in PICU patients. Active fluid removal improves the prognosis but may worsen organ dysfunction. Preload dependence in adults does predict hemodynamic instability induced by a fluid removal challenge (FRC). We sought to investigate the diagnostic accuracy of dynamic and static markers of preload in predicting hemodynamic instability and reduction of stroke volume during an FRC in children. We followed the Standards for Reporting of Diagnostic Accuracy statement to design conduct and report this study. DESIGN: Prospective noninterventional cohort study. SETTINGS From June 2017 to April 2019 in a pediatric cardiac ICU in a tertiary hospital. PATIENTS Patients 8 years old or younger, with symptoms of fluid overload after cardiac surgery, were studied. INTERVENTIONS: We confirmed preload dependence by echocardiography before and during a calibrated abdominal compression test. We then performed a challenge to remove 10-mL/kg fluid in less than 120 minutes with an infusion of diuretics. Hemodynamic instability was defined as a decrease of 10% of mean arterial pressure. MEASUREMENT AND MAIN RESULTS: We compared patients showing hemodynamic instability with patients remaining stable, and we built receiver operative characteristic (ROC) curves. Among 58 patients studied, 10 showed hemodynamic instability. The area under the ROC curve was 0.55 for the preload dependence test (95% CI, 0.34-0.75). Using a threshold of 10% increase in stroke volume index (SVi) during calibrated abdominal compression, the specificity was 0.30 (95% CI, 0.00-0.60) and the sensitivity was 0.77 (95% CI, 0.65-0.88). Mean arterial pressure variation and SVi variation were not correlated during fluid removal; r = 0.19; 95% CI -0.07 to 0.43; p = 0.139. CONCLUSIONS: Preload dependence is not accurate to predict hemodynamic instability during an FRC. Our data do not support a reduction in intravascular volume being mainly responsible for the reduction in arterial pressure during an FRC in children.