An in vitro comparative study of fluid balance accuracies between pediatric-specific and inlinehemofiltration devices for continuous renal replacement therapy during extracorporeal
membrane oxygenation
Arvind Santhanakrishnan, T. Nestle, B. L. Moore, A. Johannes, A. P. Yoganathan and M. L. Paden
Wallace H. Coulter Department of Biomedical Engineering, Georgia Institute of Technology and
Emory University
Division of Pediatric Critical Care, Children’s Healthcare of Atlanta at Egleston
Department of Pediatrics, Emory University School of Medicine
Failure of the cardiac or respiratory system is a common problem seen in pediatric and neonatal
intensive care units. When conventional management fails to improve a child’s condition,
extracorporeal life support such as extracorporeal membrane oxygenation (ECMO) can provide lifesaving temporary heart and lung support. Children requiring ECMO for respiratory failure often have
significant fluid overload and renal insufficiency. Recent studies have pointed to the benefits of early
intervention with hemofiltration to improve survival in neonatal and pediatric ECMO. However,
currently available CRRT devices are not Food and Drug Administration approved for use in ECMO
patients. Current CRRT devices do not integrate ideally into the pressure and flow conditions seen with
ECMO, and the procedure for clinical use in these patients involves unsafe practices such as
suppressing inbuilt alarms. We have designed a pediatric-specific CRRT device that can be easily
integrated within an ECMO system and is capable of providing accurate fluid balance across neonatal
to pediatric flow rates. Further, the pediatric CRRT device design has different alarms appropriate for
neonatal, pediatric and adult profiles. Though the fluid management system is designed to provide
perfect fluid balance, flexibility in the clinical prescription of negative or positive fluid balance has
been incorporated using an additional offset pump for fine control in the volume of replacement fluid
delivered to the patient. For the purpose of validation of the improved fluid balance accuracy achieved
during CRRT under ECMO, we conducted in vitro tests comparing 8-hour long continuous operation
of our device and a simplified inline hemofilter device that has been implemented at the Pediatric
Intensive Care Unit of Children’s Healthcare of Atlanta at Egleston (CHOA). In the latter device, the
ultrafiltrate (UF) is removed and replacement fluid (RF) is infused using two identical intravenous (IV)
pumps. An identical pulsatile pump was used in both devices to drive the hemofiltration process.
Multiple sets of conditions were considered, with ECMO flows ranging from 0.5-2 L/min and UF-RF
flows in the range of 300-900 ml/hr. A common reservoir was used for collecting the fluid drained by
the UF pump and for also providing the sourcing volume for the RF pump. The change in the fluid
volume within the reservoir, during CRRT in conjunction with ECMO, was measured using a weighing
scale every 15 minutes throughout the entire duration of the test. The results of the study showed that
the fluid balance error obtained using our pediatric-specific device ranged from 0.1% to 0.75%, which
is over an order of magnitude improvement as compared to the current clinical implementation in
CHOA. Across most of the conditions investigated, the inline hemofiltration device necessitated
additional infusion of fluid volume in the bladder of the ECMO circuit during the course of the 8-hour
period. Ongoing studies are examining the accuracies obtained from using commercial adult CRRT
devices for comparison to our novel pediatric-specific design.
315 Ferst Drive NW, Parker H. Petit Biotechnology Building
Department of Biomedical Engineering, Georgia Institute of Technology
Atlanta, GA 30332-0363
Fax: (404) 894-4243
E-mail: arvind7@gatech.edu