Haemodialysing babies weighing <8 kg with the Newcastle infant dialysis and ultrafiltration system (Nidus): comparison with peritoneal and conventional haemodialysis

Abstract

Background: To compare the efficacy of the Newcastle infant dialysis and ultrafiltration system (Nidus) with peritoneal dialysis (PD) and conventional haemodialysis (HD) in infants weighing <8 kg.

Methods: We compared the urea, creatinine and phosphate clearances, the ultrafiltration precision, and the safety of the Nidus machine with PD in 7 piglets weighing 1-8 kg, in a planned randomised cross-over trial in babies, and in babies for whom no other therapy existed, some of whom later graduated to conventional HD.

Results: Two babies entered the randomised trial; 1 recovered rapidly on PD, the other remained on the Nidus as PD failed. Additionally, 9 babies were treated on the Nidus on humanitarian grounds: 3 because of failed PD, and 3 with permanent kidney failure later converted to conventional HD. We haemodialysed 10 babies weighing between 1.8 and 5.9 kg for 2,475 h during 354 Nidus sessions without any clinically important incidents, and without detectable haemolysis. Single-lumen vascular access was used with no blood priming of circuits. The urea, creatinine and phosphate clearances using the Nidus were around 1.5 to 2.0 ml/min in piglets and babies, and were consistently higher than PD clearances, which ranged from about 0.2 to 0.8 ml/min (p ≤ 0.0002 for each chemical). Ultrafiltration was achieved to microlitre precision by the Nidus, but varied widely with PD. Fluid removal using conventional HD was imprecise and resulted in some hypovolaemic episodes requiring correction.

Conclusion: The Nidus can provide HD in the Pediatric Intensive Care Unit (PICU) and outpatient intermittent HD without blood priming for babies weighing <8 kg, It generates higher dialysis clearances than PD, and delivers more precise ultrafiltration control than either PD or conventional HD.

Fig. 1

The Newcastle infant dialysis and ultrafiltration system (Nidus machine) in use. The numbered parts are described in the text. The parents gave their informed consent to use the photograph

Fig. 2

Chemical clearances in 7 piglets treated using the Newcastle infant dialysis and ultrafiltration system (Nidus, solid symbols and lines) and by peritoneal dialysis (PD, open symbols and broken lines), expressed a as absolute values, ml/min/piglet, and b relative to body surface area, ml/min/1.73 m². In this case, the ultrafiltration was set at 40 ml/h for Nidus, and 3.86 % glucose dialysate was used for PD. Squares urea, triangles creatinine, circles phosphate

Fig. 3

Absolute clearances of urea, creatinine and phosphate a in infants, and b in piglets, comparing those generated by the Nidus machine (dark grey columns) and by Peritoneal dialysis (light grey). For the babies, the values were the first ones measured in that child. The number of cases is shown on each column.*Comparing babies with unpaired t tests; ^ϕcomparing each piglet using paired t tests

Fig. 4

Differences between the absolute clearances of urea, creatinine and phosphate delivered by the Nidus machine and Peritoneal dialysis, ml/min, for the 4 babies (solid symbols) and 7 piglets (open symbols) that received both treatments. Means and individual values are shown. p values are for independent t tests comparing the baby and piglet clearances

Fig. 5

All of the absolute urea, creatinine and phosphate clearances measured in the babies, identified by their case number in Table 1. Treatments with the Nidus machine are shown with solid symbols, and Peritoneal dialysis treatments with open symbols. The same logarithmic time scale was used for all to improve the visual separation of individual babies’ values

Fig. 6

The estimated fluid balance errors in 2 babies who weighed about 6 kg, during routine outpatient dialysis and ultrafiltration sessions of 3 to 5 h, using the Nidus machine first, and then after changing them to a conventional paediatric haemodialysis machine (Gambro AK200). Fluid overload appears above the 0-line, dehydration below. The grey line indicates the expected error range of ±20 g from weighing the babies twice to the nearest 10 g

Fig. 7

The ultrafiltration rates achieved in 7 piglets on peritoneal dialysis, using solutions with glucose concentrations of 1.36, 2.27 and 3.86 %. Positive values indicate fluid removal; negative ones indicate fluid gain