Finding covert fluid: methods for detecting volume overload in children on dialysis

Abstract

Background: Lung ultrasound is a novel technique for detecting generalized fluid overload in children and adults with end-stage renal disease (ESRD). Echocardiography and bioimpedance spectroscopy are established methods, albeit variably adopted in clinical practice. We compared the practicality and accuracy of lung ultrasound with current objective techniques for detecting fluid overload in children with ESRD.

Methods: A prospective observational study was performed to compare lung ultrasound B-lines, echocardiographic measurement of inferior vena cava parameters and bioimpedance spectroscopy in the assessment of fluid overload in children with ESRD on dialysis. The utility of each technique in predicting fluid overload, based on short-term weight gain, was assessed. Multiple linear regression models to predict fluid overload by weight were explored.

Results: A total of 22 fluid assessments were performed in 13 children (8 on peritoneal dialysis, 5 on haemodialysis) with a median age of 4.0 (range 0.8-14.0) years. A significant linear correlation was observed between the number of B-lines detected by lung ultrasound and fluid overload by weight (r = 0.57, p = 0.005). A non-significant positive linear correlation was observed between fluid overload by weight and bioimpedance spectroscopy (r = 0.43, p = 0.2), systolic blood pressure (r = 0.19, p = 0.4) and physical examination measurements (r = 0.19, p = 0.4), while a non-significant negative linear relationship was found between the inferior vena cava collapsibility index and fluid overload by weight (r = -0.24, p = 0.3). In multiple linear regression models, a combination of three fluid parameters, namely lung ultrasound B-lines, clinical examination and systolic blood pressure, best predicted fluid overload (R ² = 0.46, p = 0.05).

Conclusions: Lung ultrasound may be superior to echocardiographic methods and bioimpedance spectroscopy in detecting volume overload in children with ESRD. Given the practicality and sensitivity of this new technique, it can be adopted alongside clinical examination and blood pressure in the routine assessment of fluid status in children with ESRD.

Keywords: Bioimpedance; Dialysis; Echocardiography; Fluid overload; Paediatrics; Ultrasonography.

Fig. 1

Lung ultrasound B-lines. a, b A-lines (parallel to the pleural line), c, d Z-lines (comet tails that do not obliterate A-lines and do not penetrate to the bottom of the ultrasound window; arrows), e, f single B-line (arising from a single point at the pleural line; arrows), g, h three B-lines (arising from three distinct points at the pleural line; arrows)

Fig. 2

Correlation of fluid parameters with proportional increase in weight from the prescribed target weight. SBP Systolic blood pressure, IVC inferior vena cava, OH/ECW overhydration/extracellular water

Fig. 3

Correlation of B-lines on lung ultrasound with measured weight overload in patients receiving peritoneal dialysis and haemodialysis

Fig. 4

Bland–Altman analysis comparing B-lines on the lung ultrasound image to the gold standard of weight increase following natural logarithmic transformation to account for differences in scale between parameters with limits of agreement represented by the shaded area.

Fig. 5

Dissociation between SBP elevation and measured parameters of fluid overload. Vertical lines Cutoff measurements for volume overload, horizontal lines 50th centile SBP. Dissociated readings are in the upper-left and lower-right quadrants of each plot, SBP Systolic blood pressure, IVC inferior vena cava, BIS Bioimpedance spectroscopy