# Perioperative hemodynamic parameters monitored by three noninvasive technologies in children with congenital heart disease: A prospective study

## TL;DR

This study compares three noninvasive methods for monitoring heart function in children with heart defects during surgery, finding that some parameters are better at predicting complications.

## Contribution

The study identifies specific noninvasive hemodynamic parameters that accurately predict adverse events in children with congenital heart disease.

## Key findings

- VISmea, uCImea, and eCImea showed the highest accuracy in predicting adverse events.
- LVEF in TTE could not predict adverse events and did not fully reflect cardiovascular function.
- Parameters from EC and VIS correlated with longer ICU stays and mechanical ventilation duration.

## Abstract

This study aims to compare the efficiencies of three noninvasive technologies in monitoring the perioperative hemodynamics of children with congenital heart disease (CHD) including ventricular septal defects with or without atrial septal defects. Three noninvasive technologies included transthoracic echocardiography (TTE), electrical cardiometry (EC), and vasoactive inotropic score (VIS). Parameters included left ventricular ejection fraction (LVEF) and cardiac index (cardiac index monitored by ultrasound, uCI) in TTE, cardiac index (cardiac index monitored by electrical cardiometry, eCI) and systemic vascular resistance index (SVRI) in EC, and VIS. Seventy‐four children were eligible. Three types of adverse events (AEs) related to disease activity and prognosis were observed, including cardio‐pulmonary resuscitation in five cases (5/74, 6.76%), hypoxic‐ischemic brain damage in four cases (4/74, 5.41%) and hemopurification in four cases (4/74, 5.41%). Except for LVEF, eight parameters (VISmax [maximum VIS], VISmea [mean VIS], uCImea [mean uCI], uCImin [minimum uCI], eCImea [mean eCI], eCImin [minimum eCI], SVRImea [mean SVRI], and SVRImin [minimum SVRI]) showed predictive value for any AE (p < 0.05). VISmea, uCImea, and eCImea demonstrated the highest accuracy and linear associations (AUROC > 0.9, p = 0.00). Linear associations also existed between the three groups of parameters and the duration of mechanical ventilation (MV) and the length of stay (LOS) in the intensive care unit (ICU). The duration of MV and the LOS in the ICU increased as VISmea rose, or uCImea and eCImea fell (p < 0.05). LVEF in TTE could not predict any AE (p > 0.05) and not fully reflect the cardiovascular function. Therefore, most parameters obtained in TTE, EC, and VIS can reflect the perioperative hemodynamics of children with CHD, with VISmea, uCImea, and eCImea being most accurate.

ROC curves illustrates the predictive value of parameters obtained from three methods in predicting any of the AEs. AEs, adverse events; FPR, false positive rate; TPR, true positive rate; 1 (LVEFmea line overlaps with LVEFmin line); 2 (SVRImea line overlaps with SVRImin line); 3 (SVRImea line, SVRImin line, and LVEFmea line overlap); 4 (SVRImea line overlaps with LVEFmea line); 5 (eCImea line overlaps with eCImin line); 6 (uCImea line overlaps with eCImea line); 7 (uCImea line overlaps with uCImin line); 8 (eCImea line overlaps with uCImin line); 9 (uCImea line, eCImin line, and uCImin line overlap); 10 (uCImea line overlaps with eCImin line); 11 (VISmax line overlaps with VISmea line).