Non-invasive cardiac output monitoring in children: clinical validation

Introduction

Continuous noninvasive cardiac output (CO) provides valuable information for patient management. Non-invasive cardiac output monitoring (NICOM) measures CO based on chest bioreactance and is validated in adults [1]. Validated data in children are lacking. Our objective was to evaluate NICOM in children with pulmonary artery catheter thermodilution (PAC) as reference.

Methods

Paired CO values using NICOM and TD were recorded during cardiac catheterization in children with congenital heart disease. Children with intracardiac or extracardiac shunts were excluded. PAC was inserted through the femoral vein and CO was measured after bolus injection of 5 ml iced saline. NICOM was connected in accordance with the manufacturers' directrix.

Results

Nineteen pairs of CO measurements were collected in nine patients. Mean age was 4.6 years (range: 0 to 12 years) and mean weight 16.8 kg (range: 4.8 to 34 kg). Cardiac diagnosis was dilated cardiomyopathy or interventional procedures. Mean CO values were 2.18 l/minute (PAC) and 1.88 l/minute (NICOM). Correlation between two methods was significant (r = 0.826; P = 0.0005). Bland-Altman analysis shows a mean difference between the reference method and NICOM of +0.33 l/minute. Ninety-five percent of measurements were inside the limits of agreement (±1.96SD) but these limits were broad (-1.24 to 1.96 l/minute) (Figure 1).

Figure 1

Bland-Altman analysis.

Conclusions

CO measurements with NICOM and PAC show a significant correlation. Bland-Altman demonstrates an acceptable agreement; however, the limits of agreement are broad. Depending on the CO range, NICOM reveals a trend to systematically overestimate or underestimate CO. Additional studies in larger and more heterogeneous pediatric patient populations are warranted for further validation.