!!""#$% 
!""
 &
     


  


  '$()

 
 *  '   
   '+  
 $ 


)
, 

,- .(/0
 


(/ ' ''$

1

 
+ + 



 )
  '
 


  
 +
 
 
   ' $ 
 
+) 
''  

, 
  
  ' $2
,

,
  
 
 $

3$4
%5
  
5  
 * 
'
6
57
 86


9

%&":

$)
;+
 $ '
$ $
1898
1 2
Comparable Clearance of Valproic Acid with High-Flux
Hemodialysis (HFHD) and Continuous Veno-Venous
Hemodiafiltration (CVVHDF)
S. Walters
University of Michigan Health System, C.S. Mott Children’s Hospital,
Ann Arbor, Michigan, USA
scotwalt@umich.edu
Two female cousins (P1, 11 y/o; P2, 16 y/o) presented comatose
after recreational use of an unknown amount of valproic acid, along
with Gabapentin. They were found together, unresponsive. After IV
fluids, NG lavage, activated charcoal and intubation, they were
transferred to our PICU. Laboratory results showed elevated valproic
acid levels (P1, 536.9 ug/mL; P2, 424.5 ug/mL) and hypernatremia.
Within ~4 hours upon arrival to our PICU both patients had an acute
catheter placed for dialysis. P1 received 3 hours of HFHD, while P2
received ~12 hours of CVVHDF. The t1/2 of valproic acid was reduced
to 2.24 hrs during P1’s 3 hour HFHD run, but t1/2 at 11 hrs after
HFHD was 7.44 hrs. P2’s valproic acid t1/2 decreased to 4.42 hrs and
4.71 hrs at 6 hours and 9 hours, respectively after starting CVVHDF.
P1 was hemodynamically unstable during and after HFHD, requiring
vasopressor medications. Both patients had neurologic recovery.
Several case reports have shown hemodialysis to be effective in the
clearance of valproic acid. The pharmacokinetic properties of valproic
acid of low molecular (144 Da), small volume of distribution (0.5–
1 L/kg), along with increased percentage of unbound drug during
acute ingestions most likely facilitate its ability to be cleared by
dialysis. However, hemodynamic instability may prevent a patient
from receiving HFHD. These cases demonstrate that CVVHDF is an
additional valuable therapy in the setting of acute valproic acid
ingestion.
600
Valproic Acid Levels (ug/mL)
500
400
300
200
100
0
0 6 12 18 24
Time (hours)
CVVHDF HFHD
Pediatr Nephrol (2008) 23:1897–1906
Comparative Efficacy of Priming Options for an Infant
on Prolonged Continuous Renal Replacement Therapy (CRRT)
S. Muneeruddin, R. Hopfner, O. Mansoor, R. Fuentes, B. Gelman, T.
Kato, C. Abitbol
University of Miami/Holtz Children’s Hospital, Miami, Florida, USA
cabitbol@med.miami.edu
CRRT provides life-sustaining treatment for critically ill infants with
technical limitations of requiring an extracorporeal circuit that may
exceed 20% of the blood volume. Consequently, the circuit requires
priming with blood products and/or albumin at the initiation of each
new CRRT system exposing the infant to risks inherent in multiple
transfusions as well as the “bradykinin release syndrome”. It is
recommended that the CRRT system be electively replaced every
72 hours and whenever it clots. We report our experience with a 6 kg
infant who received a multivisceral transplant for liver and intestinal
failure due to microvillous inclusion disease. Post-operatively, she
suffered prolonged oligo-anuria requiring CRRT for 36 days which is
on-going at the time of this report. The PRISMA® M60 with the AN69
membrane was used to deliver continuous veno-venous hemodiafiltra-
tion (CVVHDF) with Prismasate® solutions for replacement and
dialysate. Heparin was used for anticoagulation at a rate of 10–25
units/kg/hour. Three types of prime were used: 1) 5% albumin (5%Alb)
with packed red blood cells (PRBC)(10 ml/kg) pushed separately; 2)
Naturalized blood (nBlood): 100 ml 5%Alb+100 ml PRBC+Heparin
200 units-recirculated for 15 minutes on CVVHD; 3)Elective circuit
exchange of blood from the old circuit to the new circuit (EE-CRRT).
Average lifetime of the circuit was 60±23 hours and was not different
between the priming options. Profound hypotension and cardiac arrest
occurred on one occasion with 5%Alb. Thereafter, only nBlood was
used when elective exchange could not be used. EE-CRRT was
preferred because no hypotension occurred during the exchange and
no blood transfusions were required.
3
Therapeutic Plasma Exchange for Hyperbilirubinemia in Two
Newborns During Extra Corporeal Membrane Oxygenation
L. Koster-Kamphuis, T. Antonius, A. van Heijst
Raboud University Nijmegen Medical Center
Nijmegen, The Netherlands
lkoster@cukz.umcn.nl
30 36 Background Extra Corporeal Membrane Oxygenation (ECMO) is an
established therapy for respiratory and circulatory failure in newborns.
Therapeutic plasma exchange (TPE) as treatment for hyperbilirubine-
mia in neonates is described. We describe two neonates with
Pediatr Nephrol (2008) 23:1897–1906
hyperbilirubinemia while on ECMO treated with TPE after insuffi-
cient result of phototherapy and exchange transfusions.
Patients The first patient was an at full term born male infant weighing
3.6 kg, diagnosed with early-onset group B streptococcal sepsis
requiring ECMO for respiratory and circulatory failure. The second
patient was a male infant born at 34 5/7 weeks, weighing 3.4 kg,
requiring ECMO for respiratory failure associated with a ruptured
omphalocele. Both patients developed severe pulmonary hypertension.
In the first days of life both infants developed an unconjugated
hyperbilirubinemia. Despite intensive phototherapy and exchange
transfusions while on ECMO the bilirubin concentration increased
further. TPE was performed to treat the hyperbilirubinemia.
Methods The PRISMA machine was used with TPE 2000 set. The
system was primed with donor blood and connected to the ECMO
circuit between the child and the bladder box. TPE was performed
with fresh frozen plasma as replacement fluid. Blood flow was
100 ml/min in both children. In one child the exchange volume was
250 ml (0.7×estimated total plasma volume) and in the other 500 ml
(1.3×estimated total plasma volume).
Results The TPE had good result in lowering bilirubin values. One
session was sufficient in both children. No serious adverse events
were seen during the treatment.
Conclusion TPE is an effective treatment for unconjugated hyper-
bilirubinemia of the newborn. TPE is possible to perform in
combination with ECMO.
4 assess the impact of this resource intensive therapy on both short-term
Discrepancy Between Prescribed and Delivered Dose
of Continuous Veno-Venous Hemodiafiltration in Pediatric
and Adult Patients
N. Amin, W. A. Jimenez, M. K. Nguyen, B. Gales, G. Ramos, I. B.
Salusky, O. Yadin and J. J. Zaritsky.
Department of Pediatric Nephrology, David Geffen School of
Medicine at UCLA, Los Angeles, California, USA
namin@mednet.ucla.edu
While much emphasis had been placed on determining the optimal
prescribed dose of continuous renal replacement therapy (CRRT), it has
yet to be shown that the prescribed dose is actually delivered. In order to
determine whether the clearance prescribed during CRRT is achieved,
we performed a prospective study comparing the prescribed clearance
(corrected for time off therapy, body surface area and ultrafiltration) to
the achieved clearance in both adult and pediatric patients receiving
continuous veno-venous hemodiafiltration (CVVHDF) therapy. A total
of 128 daily creatinine clearance (CrCl) measurements were made over
an average of 4.8±2.3 and 6.5±1.6 consecutive days of CVVHDF
therapy in 16 pediatric and 8 adult patients, respectively. The delivered
clearance was only 71% and 74% of the prescribed clearance in the
pediatric and adult populations, respectively (delivered vs. prescribed
dose, P<0.001). The prescribed dose was significantly higher in the
pediatric compared to the adult patients (P<0.005). The principal cause
1899
of temporary cessation of therapy was filter clotting, accounting for 39%
and 49% of total time off CVVHDF in pediatric and adults respectively.
These results demonstrate that in both populations, despite accounting
for time off therapy, the delivered daily CrCl was significantly less than
prescribed. The reasons for this relatively poor delivered dose have yet
to be identified, but may be secondary to inaccuracies in the current
methods used to calculate the prescribed dose or suboptimal CVVHDF
filter performance. Thus, in order to achieve targeted clearance during
CVVHDF therapy, relatively higher CrCl needs to be prescribed.
5
A Retrospective Study of Outcomes in Pediatric
Hematology/oncology Patients Receiving Continuous
Venovenous Hemodialysis (CVVHD)
Y. Avent, N. Henderson, T. Collie, R. F. Tamburro, L. Elbahlawan,
R. R. Morrison, S. Rajasekaran
St. Jude Children’s Research Hospital, Memphis, Tennessee, USA
Yvonne.avent@stjude.org
Background Children undergoing therapy for cancer are prone to
develop tumor lysis, renal failure or fluid overload as a result of
necessary treatments. Often these patients require modalities of renal
replacement therapy such as CVVHD to support and preserve
electrolyte and fluid homeostasis. The purpose of this project was to
ICU survival and long-term survival in oncology patients 6 months
after an episode of CVVHD.
Methods A retrospective review of an institutional database identified all
patients receiving CVVHD in an 8-bed pediatric oncologic ICU from
January 2003 through December 2007. Abstracted data included
demographics, diagnoses, indications for CVVHD, use of vasoactive
medications, respiratory support, electrolyte values, survival to ICU
discharge and 6 month survival after treatment with CVVHD. For patients
requiring a second episode of CVVHD, 6 month survival was analyzed
only if onset of the second episode was at least 6 months after the first.
Results 41 patients underwent 48 episodes of CVVHD averaging 172
treatment days/year. The median age of this cohort was 12 years. Of
the 48 episodes of CVVHD, primary indications were acute renal
failure (n=30, 62.5%), fluid overload (n=12, 25%), chronic renal
failure (n=5, 10.4%), and tumor lysis syndrome (n=1, 2.1%). 30
hematopoietic stem cell transplant (HSCT) patients underwent
CVVHD, accounting for an average of 15.9±2.02 days of therapy.
Primary diagnoses prior to HSCT included leukemia/lymphoma (n=
18, 60%), solid tumor/neuro oncology (n=6, 20%) and others (n=6,
20%). Among HSCT patients, survival to ICU discharge occurred
after 13 of 36 (36%) episodes of CVVHD. Six-month survival after
CVVHD occurred in one HSCT patient. 11 non-HSCT patients
received an average of 9.6±2.9 days/year of CVVHD therapy. Of non-
HSCT patients, 9 had leukemia (81.8%), 1 had solid tumor (9.1%) and
1 was pre-transplant (9.1%). Among non-HSCT patients, survival to
ICU discharge occurred after 5 of 12 (42%) episodes of CVVHD.
1900
Four non-HSCT patients survived beyond 6 months (36.3%).
Mechanical ventilation, vasoactive infusions and hyperglycemia at
the onset of CVVHD did not impact mortality in this study.
Conclusions CVVHD is an effective renal replacement modality in
critically ill children with cancer. Our 6 month survival rate in non-
HSCT patients compares favorably with that of the general pediatric
ICU population receiving this therapy. While the observed short-term
benefit of CVVHD in HSCT patients is encouraging, it does not
translate into 6 month survival. Further research is required to delineate
factors which may convert the short-term benefits of CVVHD into long
term survival in the high-risk group of HSCT patients.
6 nous infusions of two or more of the following agents sodium benzoate,
Concurrent Continuous Renal Replacement Therapy (CRRT)
and Plasmapheresis in Pediatrics
D. Eding, L. Jelsma, B. Lovejoy, R. Hackbarth, T. Bunchman
Helen DeVos Children’s Hospital, Grand Rapids, Michigan, USA
dawn.eding@devoschildrens.org
Pediatric patients can occasionally require the addition of plasmaphe-
resis therapy to their CRRT treatment. We report a technique combining
CRRT and plasmapheresis using the Gambro Primsaflex and the Cobe
Spectra concurrently. Connections were made by the addition of two 3-
way stopcocks on the access line of the CRRT circuit. Access for
plasmapheresis therapy was connected to the first stopcock, and return
was connected via the second stopcock. Anticoagulation was main-
tained via the plasmapheresis circuit. No CRRT flow adjustments were
required; plasmapheresis flows were adjusted per protocol. This
technique is easily accomplished and done without complications. It
provides optimal therapy for patients requiring both CRRT and
plasmapheresis, preventing disruptions in hemofiltration therapy
7 was 50%, all in intensive care.
High Volume Haemofiltration (CVVH) in the Management
of Neonates With Hyperammonaemia
C. Westrope, G. C. Morrison
Birmingham Children’s Hospital
West Midlands, UK
Claire@westropes.co.uk
Neonatal hyperammonaemia, due to an inborn error of metabolism
(IEM), may often be severe enough to precipitate coma and result in
patient death. Existing therapies displace ammonia from the serum,
either through biochemical manipulation of nitrogen clearing process-
es, or, by utilizing renal replacement therapies. The latter are
particularly indicated when urgent reduction in the serum ammonia
level is required. Haemodiafiltration (HDF) can offer efficient
ammonia clearance and patient stability, but may be cumbersome.
Given reports that high volume CVVH may offer similar efficacy in
Pediatr Nephrol (2008) 23:1897–1906
the clearance of ammonia we reviewed our experience of CVVH in a
population of hyperammonaemic neonates.
12 neonates with severe hyperammonaemia (median 880 µmol/L,
range 329 -1887 µmol/L) were admitted to the paediatric intensive care
unit of the Birmingham Children’s Hospital between 2000 and 2007.
The median age was 4 days (range 1–10 days) while the median body
weight was 2.7 kg (range 1.7 – 3.9 kg). In all patients the underlying
diagnosis was unknown at time of admission. Subsequently, hyper-
ammonaemia was found to be secondary to an organic acidaemia in 5
patients, while 7 patients had a urea cycle defect. Prior to initiating
CVVH, 11 children required mechanical ventilation, and 5 had
commenced inotropic support. In addition all were receiving intrave-
phenylacetate or arginine All were commenced on high volume CVVH
(median ultrafiltrate flow 111 ml/kg/hr; range 78 – 250 ml/kg/hr)
utilizing the BM 25 (Baxter) or AQUARIUS (Edwards Lifesciences).
Median circuit bloodflow was 12.1 ml/kg/min (range 5.7 – 24.0 ml/kg/min).
In those patients whose CVVH was conducted using the BM25, the
FH22 polyamide filter was employed. The AQUARIUS system was
coupled with the HFO7 polyethersulphone filter.
11 patients completed the course of CVVH. One patient experi-
enced an extravasation complication following dialysis catheter
insertion. CVVH was abandoned and peritoneal dialysis therapy was
employed. Two patients experienced acute haemodynamic compro-
mise on initiation of circuit bloodflow. Both were subsequently
stabilized and CVVH continued. The mean decrease in mean
ammonia levels at 12 hours of CVVH was 61% (+/- 7.9%) while at
24 hours the mean ammonia level was decreased by 81% (+/- 3.5%)
and, 9 of 11 patients had achieved “non-toxic” ammonia levels
(<200 µmol/L). No correlation was detected (Pearson coefficient)
between the circuit blood or ultrafiltrate flow and the rapidity of
reduction in serum ammonia. This may be due to small patient
numbers and the confounding effects of co-administered medical
therapies and use of different haemofilters. The mortality in the group
High volume CVVH can produce a reduction in ammonia load in a
timespan comparable to that of HDF and may simplify the use of
continuous renal replacement therapy in hyperammonaemic conditions.
8
Continuous Renal Replacement Therapy (CRRT)
in the Treatment of Hyperammonemia Associated with Inborn
Errors of Metabolism
D. Eding, H. Marine, N. Hautala, L. Border, J. Harley,
R. Hackbarth, T. Bunchman
Helen DeVos Children’s Hospital
Grand Rapids, Michigan, USA
dawn.eding@devoschildrens.org
Children presenting with severe hyperammonemia require rapid
correction of their ammonia levels to minimize neurologic injury.
Pediatr Nephrol (2008) 23:1897–1906 1901

Two newborns presenting with hyperammonemic encephalopathy 10

were treated with hemodialysis (HD) then transitioned to continuous
veno-venous hemodiafiltration. CRRT was started using the Gambro Pediatric CRRT Nurse Model: The Transition to an ICU Based
Prismaflex, a blood flow of 25 mL/kg/minute and citrate anti- Model
coagulation. A M60 filter was utilized on an infant with methylma-
lonic acidemia, with initial flow rates of filter replacement fluid (FRF) T. Mottes, J. Vamos, W. Wieneke, J. Juno
at 1.5 L/hour and dialysate at 2 L/hour. Flow rates were weaned over a University of Michigan Hospitals, Ann Arbor, Michigan, USA
16 hour period and CRRT was discontinued. A HF 1000 filter was tmottes@umich.edu
utilized on an infant with a urea cycle defect with both FRF and
dialysate at 4 L/hour. Flow rates were decreased over a 52 hour period
Over the past 16 years our patient activity has increased by 65%, from
and CRRT was discontinued. Both infants had ammonia levels of over
14 to 40 patients in the past year. Our current nursing care delivery
1000 mcmol/L prior to therapy, which rapidly decreased on HD and
model consists of the pediatric dialysis nursing staff providing the set
continued to decline on CRRT to levels less than 100 mcmol/L for the
up, prime and initiation, while the PICU nursing staff provides the
duration of the therapy. Sequential HD/CRRT is effective in the acute
bedside hourly care. The increased activity along with projected
management of hyperammonemia associated with metabolic disease.
continual growth would stretch our current care delivery model past
its current adaptive potential. Thus, the need to explore other models
9
became imperative to maintaining our existing level of patient care. Our
new model allows for continued growth, is fiscally responsible,
Inaccuracy of Concomitant CVVH in ECMO Patients
compliments the excellent nursing bedside care and doesn’t overburden
the nursing staff with its implementation. With those goals in mind, a
M. I. Paden, C. Reid, S. F. Wagoner, P. Sucosky, L. P. Dasi, A. P.
model was developed with the care being shifted from the pediatric
Yoganathan, J. D. Fortenberry
dialysis nursing staff to the pediatric ICU nursing staff.
Emory University/Children’s Healthcare of Atlanta/Georgia Institute
Transition to the new care delivery model necessitated the implemen-
of Technology, Atlanta, Georgia, USA
tation of a training program, along with the creation of a nursing
mpaden@emory.edu
leadership position, the program coordinator, designed to train the ICU
staff to assume the set up and initiation responsibilities. The Initiator
Introduction: Renal failure complicates care of critically ill children
Education program consists of 3 Steps; Hands-on demonstration of the
on ECMO. On ECMO, CVVH can be delivered either by standard
machine set up and the different initiation procedures, 5 assisted patient
systems (Braun Diapact) or by a simplified system driven by ECMO
initiations, and ongoing education, including CRRT drills. This training
pump flow using IV pumps (Weber, 1998). Accuracy concerns exist
program is in addition to the current education program for CRRT.
for concomitant CVVH and ECMO, due to circuit pressure effects. We
During the transition to it is important that we track the effects on the
evaluated accuracy of the inline and Diapact systems using an in vitro
nursing staff. To measure these effects a survey tool was developed that
ECMO circuit model.
has the staff rank their perceived comfort and knowledge of CRRT.
Methods: Two identical saline primed ECMO circuits were used. One
During the transition period, the survey will be implemented every
circuit added an inline hemofilter system, using IV pumps (Model 8100,
4 months, with first survey being just prior to the start of the transition
Alaris Medical Systems) to deliver replacement fluid (RF) and create
process. In brief summary, the baseline survey results with 72% of all
ultrafiltrate (UF), that is measured with a urometer (Criticore, Bard) The
CRRT trained nurses responding indicate an overall comfort level
other circuit used a Diapact for CVVH. Both methods diverted saline
average score of 4.2. During this transition time, evaluating and
post ECMO pump but pre-membrane, sent it through a PAN 6 hemofilter,
trending the data will allow us to adapt to the nursing education needs.
and returned to the bladder. CVVH was prescribed with zero balance and
UF rates from 0.5–2 L/hour. RF and UF bags were weighed hourly.
Results: Forty eight hourly measurements were analyzed (26 hrs Alaris, 11
22 hrs Braun). Adjusting for varying UF rates, the Alaris pump delivered
a median 4.3% (range+3% to -25%) less RF per hour than set and created AN69 Surface-Treated (AN69ST) Membrane and Hemodynamic
a median of 0.8% (+7% to -12%) less UF/hr than prescribed. The Braun Profile of Critically Ill Children During Initiation of CRRT
Diapact delivered a median of 1% (+10% to -7%) more RF/hour and
created a median of 1% (+6% to -8%) per hour more UF than prescribed. W. Kechaou, P. Jouvet, V. Phan, C. Litalien.
Conclusions: In this in vitro CVVH/ECMO model, both IV pump and Centre Hospitalier Universitaire Sainte-Justine, Montreal, Canada
Diapact systems had clinically significant error rates, with potential for Wassim.kechaou@hotmail.fr
unexpected fluid removal or excess delivery. Careful clinical assess-
ment of volume status is essential with concomitant ECMO/CVVH. Background: Hypotension secondary to bradykinin release via
Further work is needed to develop more accurate fluid delivery for contact activation of clotting factor XII by AN69 membrane is a
CVVH on ECMO. known complication usually occurring within 15 minutes after the
1902 Pediatr Nephrol (2008) 23:1897–1906

start of CRRT. AN69ST membrane was designed to prevent this pH-
dependent reaction. To date, there are few data regarding the use of
this membrane in children. Objectives: To evaluate the hemodynamic
profile of critically ill children at the start of CRRT when AN69ST
membrane is used. Methods: A retrospective study of critically ill
children who underwent CRRT with AN69ST membrane. Hemody-
namic changes during the first hour after CRRT initiation were
assessed by recording from the charts heart rate and systolic blood
pressure at 10 min intervals, changes in vasoactive score1 and need for
volume and HCO3- administration. Severity of illness, priming
techniques and blood pH were also evaluated. Hemodynamic
instability was defined as hypotension, increase in vasoactive score
or need to administer≥10 ml/kg of volume. Results: Eleven patients
with a median age of 7.5 yrs (5 months–15 yrs) were included. At
CRRT initiation, median PELOD score, blood pH and vasoactive
score were 30 (10–43), 7.33 (7.21–7.46) and 24 (0–97). Priming
techniques were normal saline (8/11), albumine 5% (1/11) and bypass
system described by Brophy (2/11). During the first hour of CRRT, 9
patients remained hemodynamically stable while 2 had an increase in
their vasoactive score from 25 to 27 and from 35 to 50. Both had an
initial PELOD score>40. No volume administration ≥ 10 ml/kg was
required and one patient received HCO3-. Conclusion: AN69ST
membrane was associated with hemodynamic stability at CRRT
initiation in most critically ill children. Further studies are needed to
better evaluate whether or not the use of AN69ST membrane prevents
the occurrence of hypotension at the start of CRRT, especially in
infants less than 10 kg when the bypass system is not used.
Results
Filter Fluid Cmax Sc MRT T1/2
(mg.ml) (min) (min)
During CVVH, drug clearance is dependant on many factors. Dose
adjustments assume reduced drug clearance by the renal system and
CVVH. Practice variation(pre-dilution, high volume haemofiltration)
alters drug removal. Dosing during CVVH is complex; under- or
overdosing may occur. We studied the PK of ceftriaxone during
CVVH in an in vitro model Methods: Renoflow filters were used to
model 6 & 20 kg patient. After priming, each circuit and reservoir was
prepared with a known vol. of Hartmann’s, 4.5% HAS or blood using
the Infomed 400. Blood pump speed & exchange rate for each circuit
was protocolised. HBO used as replacement fluid, 70% predilution.
Following paired sampling from circuit and ultrafiltrate fluid,
Ceftriaxone (80 mg/kg) was injected into the post-filter port (time
0). Paired samples were taken at 10 time points (0–720 mins).
Ceftraxione concentrations determined using HPLC.
Conclusion: Estimates of high Sc and short circuit half-life from this
in-vitro model suggest Ceftriaxone is rapid cleared during CVVH
(cleared by 240 min), with important implications for dosing during
in-vivo CVVH. The albumin circuit had the lowest Sc and longest
terminal half-life, reflecting protein binding of drug, Ceftriaxone
clearance may increase in hypoalbuminaemic patients.
13
Pediatric Continuous Renal Replacement Therapy Program:
Maintaining Nursing Skills in a High Risk Low Volume Area
G. Bonin, H. Cooper, C. Press, S. Kowalski, J. Plouffe, K. Pederson,
M. Kesselman, T. Blydt-Hansen
Children’s Hospital Health Sciences Centre
Winnipeg, Manitoba, Canada
gbonin@exchange.hsc.mb.ca

0.7 m2 Albumin 3.5 0.23 236 164

0.7 m2 Blood 4.5 0.31 89 62 The Winnipeg Pediatric Continuous Renal Replacement Therapy
0.4 m2 Hartmans 1.65 0.52 178 124 (CRRT) program averaged two-three patients per year up to 2006. The
0.7 m2 Albumin 2.65 0.64 132 91 model of nursing care included training the majority of Pediatric
0.7 m2 Blood 6.5 0.51 29 20 Intensive Care nurses in CRRT. Nursing education at this time
0.7 m2 Hartmans 2.93 0.49 108 75
included a one-day classroom orientation and an annual four-hour
recertification class. Due to the low volume of usage, this large group
12 of nurses struggled to maintain a basic level of skill and familiarity.
In 2006 the CRRT committee re-evaluated the program. Basic
Ceftriaxone Pharmacokinetics (PK) During Continuous Pediatric setup times (>2 hours) and troubleshooting were identified as specific
Veno-veno Haemofiltration (CVVH); in vitro model challenges. The objective was to develop and apply a training and
skills maintenance model to improve the expertise and exposure of
nurses providing CRRT. This model included the following tasks:
B. Harvey, D. Yeomanson, T. Johnson, H. Mulla, & A. Mayer
Dept. of PICU Sheffield Children’s Hospital, Sheffield, UK 1. Select a limited number of nurses who are committed to
Ben.Harvey@nhs.net developing an enhanced skill-set as “specialists”.
2. Restrict the delivery of CRRT skills to nurses from this select pool.
3. Train nurses to provide basic, then more advanced care:
1
Dose of dopamine + dobutamine + (epinephrine X 100) + A. Advanced Users (AU): To act as educators and to perform
(norepinephrine X100) + (phenylephrine X 100) + (milrinone x 10). advanced troubleshooting.
Pediatr Nephrol (2008) 23:1897–1906
B. Super Users (SU): To set up the CRRT circuit including
Albumin and Blood Primes, and to provide typical
troubleshooting.
C. Associates (AS): To perform the daily operation of the
CRRT machine and provide basic trouble shooting
4. The enhanced training program included the following components:
A. Increase initial CRRT education to a day and a half for all
CRRT trained nurses.
B. “Dry lab” every three months to practice set up and trouble
shooting skills.
C. AUs received full day education for new Prismaflex. “Dry
lab” practice once a week to once a month, assistance with
Prismaflex education and development and review of
policies and procedures.
D. Present a Problem of the Month.
E. Monthly CRRT Case reviews.
5. Apply a mix of expertise to each nursing shift: At least one AU,
SU and one AS on each shift. This team provides comprehensive
patient care, and builds capacity for the AS.
6. Invite Clinical Specialists in Pediatric CRRT.
In the two years following the program changes, there has been a
marked increase in the use of CRRT (10 patients/year). This is
attributed to increase nursing expertise and physician confidence.
Specific changes include faster set-up times (<1 hour), more confident
troubleshooting and subjective improvements in nursing independence
and skill at the beside. Ongoing challenges include: a large staff
turnover rate, high acuity preventing nurses from entering the dry lab,
a low volume of trained nurses causing difficulty in scheduling and
CRRT case reviews.
CRRT is being viewed as a safe and available treatment option
where nurses feel they are valuable members of the CRRT team.
14
Children Requiring Continuous Renal Replacement Therapy
(CRRT)
L. Jelsma, D. Eding, C. Metz, A. Neumann, V.Steen, M. Oleniczak,
R. Hackbarth, T. Bunchman
Helen Devos Children’s Hospital, Grand Rapids, Michigan, USA
lindsey.jelsma@devoschildrens.org
Two pediatric patients with severe abdominal trauma required CRRT
secondary to renal insufficiency, hypercatabolism and rhabdmyolysis.
Continuous Veno-Venous Hemodiafiltration was initiated using the
1903
Gambro Prismaflex, HF1000 filter, citrate anticoagulation and an
initial blood flow of 150 mL/minute. Patient 1 sustained an abdominal
aortic dissection. Filter replacement fluid (FRF) and dialysate were
started both at 4 L/hour. After a 57 day course paient 1 transitioned to
hemodialysis. Patient 2 sustained blunt abdominal trauma with
abdominal compartment syndrome requiring a left nephrectomy and
decompressive silo. Initial flow rates were FRF at 4 L/hour and dialysate
at 3 L/hour. After a 15 day course patient 2 recovered renal function and
CRRT was discontinued. CRRT is effective in the management of acute
renal insufficiency in patients with abdominal trauma.
15
Characteristics and Mortality of Paediatric Patients Undergoing
Haemofiltration
L.Byrne, J. White, A. Durward
Pediatric Intensive Care Unit, Evelina Children's Hospital, Guys & St.
Thomas NHS Foundation Trust, London, England, UK
Louise.byrne@gstt.nhs.uk
Aim: The aim of this study was to determine the demographic
characteristics of patients receiving haemofiltration (CVVH) in a lead
paediatric intensive care unit (PICU) in United Kingdom with specific
reference to mortality.
Methods: A retrospective review of all patients who received CVVH
(n=28) in our PICU over a two year period commencing January 2006
were identified from our intensive care database and analysed. Patients
were haemofiltered using the Aquarius Platinum machine (Edwards) with
the HF03 and HF07 polysulphone filters at blood flow rates of 100 and
150 ml/min for patients below and above 15 kg respectively. Ultrafiltra-
tion rates of 60 ml/kg/hr were used for patients < 15 kg and 30 ml/kg/hr
above this weight. Demographic characteristics of patient’s receiving
CVVH were compared to those who did not receive this therapy (n=
2475). Peritoneal dialysis patients were not included in analysis (n=52).
Chi squared test was used for categorical data and Mann Whitney Test for
continuous data with a p value<0.05 considered significant. Data are
expressed as median and inter-quartile range.
Results: Twenty eight (1.1%) episodes of CVVH were identified from
2337 admissions. Case-mix of haemofiltered patients were sepsis (n=14
including 3 meningococcal), haemolytic uraemic syndrome (n=6), acute
renal failure (n=4), other (n=4 including 1 neonate with urea cycle
defect). The commonest reasons for haemofiltration included anuria
(35.7%), electrolyte abnormality (25%) and fluid overload (17.9%).
CVVH patients received inotropes more frequently (71% vs. 32%, p<
0.0001), were significantly younger, sicker and had higher mortality
1904 Pediatr Nephrol (2008) 23:1897–1906

(28.5% vs. 4%, p<0.0001) than patient’s not dialysed (Table 1). Non
survivors of CVVH had significantly higher mortality risk (PIM2 score
30 (17 to 48) vs. 5 (1 to 10), p<0.0001), shorter length of stay (0.8 days
(0.2 to 4.7) vs. 6.2 (1.8 to 17), p<0.0001) despite similar age and weight
to survivors of haemofiltration (12 kg (8 to 15) vs. 16 (9.8 to 33)
respectively.
resumed. This exchange is done over several minutes dependent on
total circuit volume and catheter flow capabilities, limiting the time off
CRRT and avoiding repeated blood exposure to the patient. Method
has been utilized successfully with the Gambro Prismaflex M60 and
HF 1000 circuits.

17
Table 1 Demographic characteristics between patients receiving
CVVH and those not receiving this therapy.
The Use of CVVHD as an Adjunct for the Treatment of Cerebral
CVVH no CVVH p - value Edema Associated with Diabetic Ketoacidosis(DKA)
(n=28) (n=2475)
R. J. Cunningham III, R. C. Gensure, L. A. Kashimawo
Age (months) 9.3 (1.3 to 48) 49 (15 to 126) <0.0001
Ochsner Children’s Health Center
Weight (kg) 15 (8 to 28) 7 (3 to 15) <0.0001
New Orleans, Louisiana, USA
Mechanical 23/28 (82%) 1786/2475 (72%) 0.19
ventilation rcunningham@ochsner.org
Length of PICU 5.2 (0.7 to 13) 2.1 (1.1 to 4.1) 0.11
stay (days) Cerebral edema is a rare complication of diabaetic ketoacidosis (0.9%
PIM2 score 9.8 (2.7 to 17.8) 2.9 (1.1 to 6.0) 0.001 of patients presenting with DKA) with a mortality rate of 20–50% and
Mortality 8/28 (28.5%) 94/2309 (4.1%) <0.0001
of the survivors, 40–50% have significant neurologic deficits. The
PIM = Paediatric Index of Mortality score incidence and outcome of this condition has not changed appreciably
over the past 20 years. We present a case of severe cerebral edema that
Conclusion: Mortality of patients undergoing haemofiltration is high was treated with CVVHD in addition to the standard measures usually
and strongly influenced by disease severity on admission to PICU. employed in the treatment of this condition. Recovery was complete
and no neurologic deficits were seen on follow up. An 11 y/o known
16 diabetic (Wt.=56 kg) presented 5 weeks after hurricane Katrina in
ketoacidosis with a pH of 6.8, a bicarbonate of 5 meq/l, & a creatinine
Continuous Renal Replacement Therapy (CRRT) Circuit of 1.4 mg/dl. She was given 10 m/kg of 0.9%NS and was begun on IV
to Circuit Exchange insulin (0.1 unit/kg/hr) along with IV fluids (225 cc/hr of 0.45% NS).
She was lethargic, a bit confused but responsive. Fourteen hours later
D. Eding, R. Hackbarth, T. Bunchman after continued administration of fluid and insulin, it was noted that
Helen DeVos Children’s Hospital, Grand Rapids, Michigan, USA she had decerebrate posturing. A CT scan showed cerebral edema with
Dawn.eding@devoschildrens.org effacement of the quadrigeminal cistern. She was intubated, CVVHD
was instituted with Qb =120 cc/min, ultrafiltration rate of 150 cc/hr
Children less than 15 kg requiring CRRT present a challenge due to and dialysate flow of 100 cc/hr (25 meq/l Bicarbonate bath) & given
patient size to total circuit volume ratio necessitating a circuit blood IV mannitol (1 gram/kg). Over the next 7 hours, her net balance
prime. This can cause hemodynamic, acid base, and electrolyte decreased by 3000 ml, her pulse rate was up to 160/min (from 110/min),
disequilibrium. We developed a technique that can be used with and urine output had dropped by 50%. Her fluid balance was
elective circuit changes, or with transition from hemodialysis (HD) to maintained but dialysate flow was increased (bicarbonate was still
CRRT. The method allows exchange of patients own blood from old 9 meq/l) to 200 cc/hr. & the patient was responsive to stimuli. Sixteen
circuit to new circuit. A second CRRT machine is saline primed. At hours after initiation, CVVH was discontinued and CT scan showed
circuit transition the patient’s blood is returned via the catheter return improvement. The patient continued to improve, was awake 48 hours
line. Simultaneously the new machine is connected to the dialysis after initiation of CVVH & was discharged on the 5th hospital day.
catheter access line and the patient primes this circuit. Machines are The aim in using CVVHD was to use osmotic, convective and
started at the same time and blood flow rates are identical therefore the hydrostatic forces to encourage movement of fluid from brain to blood
exchange is volume neutral. When exchange is complete therapy is while simultaneously establishing better biochemical control.
Pediatr Nephrol (2008) 23:1897–1906
18
Combined Treatment with ECMO and CRRT in Children
After Heart Surgery
M.Yarustovsky, K. Shatalov, M. Abramian, E. Nazarova,
O. Stupchenko
Bakoulev Center for Cardiovascular Surgery, Moscow, Russia
mbyer@yandex.ru
Rather often children with low cardiac output after radical correction
of complex congenital heart disease require ECMO for the support of
hemodynamics, as well as CRRT for the correction of water-
electrolyte and metabolic balance. During 2007 we have followed 7
children aged from 19 days to 10 years. Severity APACHE-II
score was 32,7±5,4. The indications for the combined procedure
were: pronounced heart failure (EF LV =15–25%, PLA=22–30 mm Hg,
CVP=16–20 mm Hg), oligoanuria (diuresis<0,5 ml/kg/hour), K+ >
5,5 mmol/l, Cr>200 µmol/l, Ur > 15 mmol/l. The duration of therapy
was from 3 to 10 days. Qp —1,4 – 2,4 l/min., Qb – 2800–3600 turns
\min., FiO2 – 40%, Qо2 – 800 ml/min. Ultrafiltration was carried out
with AV 400 HF (Fresenius). Because of high Qp, ultrafiltration
velocity was regulated with a clamp on the supplying («arterial») line.
UF volume depended on preload and infusion volume indices.
Anticoagulation: heparin (АСТ – not more 250 sec). When it was
necessary to correct metabolic and nitrogen balances, as well as K+ >
5,5 mmol/l, dialysis was added (up to 100 ml/min). After ECMO
disconnection 2 children were transferred to peritoneal dialysis.
Mortality in the group was 71.4%. The combination of ECMO and
CRRT is safe and can be effectively used in clinical settings.
19
Animal Model to Study Feasibility of Venovenous Extracorporeal
Membrane Oxygenation Using Renal Replacement Therapy
Platform
A. Divekar, G. Bonin, H. Cooper, A. Gutsol, T. Koga
University of Iowa Children’s Hospital, Iowa City, Iowa, USA
All research done at Children’s Hospital Health Sciences Center,
University of Manitoba, Winnipeg, Canada
abhay-divekar@uiowa.edu
Objectives: We have previously reported the first successful adapta-
tion of a renal replacement therapy (RRT) platform to provide
extracorporeal venovenous membrane oxygenation (VV ECMO). This
study was intended to develop an animal model to study feasibility of
delivering VVECMO using RRT platform. Methods: Animals were
handled according to Animal Care Committee guidelines at the
University of Manitoba and Canadian Council on Animal Care. Six
domestic swine (2.4–9 kg) were anesthetized by mask inhalation with
3–5% isoflurane with oxygen and surgical depth of anesthesia was
1905
maintained with 1–3% isoflurane. All animals were mechanically
ventilated. Continuous electrocardiogram, pulse oximetry, invasive
arterial pressure and core temperature was monitored. The femoral
vessels and external jugular vein were isolated by surgical cutdown.
The largest (7–12 French) double lumen cannula was placed in the
femoral and external jugular veins. A Prismaflex continuous RRT
platform with an ST-100 filter was used in CVVHDF mode. A Lilliput
902-D hollow fiber oxygenator was spliced in the return of the circuit
pre-air detection filter in conjunction with a Bio-Medicus heat
exchanger. The circuit was primed with Prismasol-4 and then blood
primed and was normalized against dialysate. Access was through the
external jugular vein and the blood was returned via the femoral vein.
The pre-blood pump rate was 400/min. All animals received 300 U/kg
of heparin sulphate for anticoagulation and activated clotting time was
maintained between 180–220 seconds. Sweep flow through the
oxygenator was 1 lit/min at 100% FiO2. Hypoxemic respiratory
failure was induced by reduction in minute ventilation and decreasing
FiO2 to 10% and documented by arterial blood gas. Flow was initiated
at 50 mL/min increased in 50 mL/min increments to maximum flow of
450 ml/min or until a limiting access/filter/return line pressures were
reached. Blood gas analysis was performed from the access line, post
oxygenator and from animal arterial line. Results: The first three
animals were successfully placed on support and helped identify and
correct several technical problems/limitations. Data from the next 3
animals is reported as mean + /-SD; the partial pressure of oxygen and
carbon-dioxide in mmHg. They weighed 3.57(1.69) kg. The pO2 was
28(3.79) and pCO2 56(18.5) prior to being placed on support. At a
flow of 67(19.37) ml/kg/min there was increase in pO2 to 40(3.1),
oxygen saturation to 78(5.13)% and drop in pCO2 to 31(8.02); at a
flow of 105(42.24) ml/kg/min there was increase in pO2 to 55(11.14),
oxygen saturation to 93(1.53)% and drop in pCO2 to 22(3.0). There
was minimal recirculation. Conclusions: RRT platform can be adapted
to provide VV-ECMO. As expected CO2 removal is very efficient,
improvement in oxygenation is clinically relevant. This animal model
will allow further research to adapt and further refine RRT platforms
and circuits for this modification.
20
Use of Tego Connectors to Prevent Hemodialysis Catheter
Infections in Children
N. G. McAfee, S. Seidel, S. L. Watkins, J. Flynn
Children’s Hospital and Regional Medical Center
Seattle, Washington, USA
Nancy.mcafee@seattlechildrens.org
PURPOSE: Catheter infections are a significant problem in pediatric
dialysis. To reduce infection rates in our program, we implemented the
use of TEGO™ connectors, which create a closed system, no open
catheter hubs, reducing manipulation and allowing unobstructed flow
rates >600 mL/min while remaining in place.
1906
METHODS: We retrospectively analyzed surveillance data collected
in 15 patients (mean age 12.62 years) following 4 quarters of use of
the TEGO™ connectors and compared this to preceding five quarters
of data. The data were then analyzed using Poisson generalized
estimating equations. RESULTS: Infection rates, in number of
infections per 1,000 patient-days, were as follows: 7.8 infections
per 1000 patient-days during the pre-TEGO period vs. 3.65 infections
per 1000 patient-days after the switch to the TEGO™ connector The
BSI Incidence Rate Ratio was 0.47 (95% CI: 0.23 – 0.96), which
indicates that the BSI rate with TEGO was less than half the BSI rate
in the preceding 5 quarters. This is a statistically significantly
reduction (p=0.04).
CONCLUSIONS: Although the small patient numbers in this study
limit the precision of our estimates of the reduction in bloodstream
infections, the current data are promising. We are now trialing these
connectors on our pediatric CRRT patients with central catheters to
see if the same reduction in infections is found.
21
Management of Acute Renal Failure and Increased
Intracranial Pressure with CRRT
R. Gonzales, J. Willoughby, D.M. Ford, G.M. Lum,
M.A. Cadnapaphornchai
The Children’s Hospital, 13123 E. 16th Ave. Box B328, Aurora, CO 80045
gonzales.richard@tchden.org
Multiple trauma is frequently associated with acute brain injury and
acute renal failure (ARF). Intermittent hemodialysis (IHD) can
compromise cerebral perfusion pressure through osmotic shifts and/or
Pediatr Nephrol (2008) 23:1897–1906
decreased intravascular volume. CRRT has been proposed to provide
greater intracranial stability in this setting. We report the successful use
of CRRT to control increased intracranial pressure (ICP) and hyper-
osmolality in a child with ARF.
Following handlebar injury to the abdomen, an 11-year-old male
complained of severe abdominal pain. He had a grade IV liver
laceration with abdominal compartment syndrome and required
extensive resuscitation including 17 u PRBC and 8 u fresh frozen
plasma. Upon transfer, he was anuric with acute tubular necrosis. He
underwent IHD without complications on hospital days 2 and 3. He
later developed an acutely enlarged and unreactive right pupil. CT
scan showed diffuse cerebral edema. He was treated with mannitol
and hyperventilation. Initial ICP was 23 mm Hg with serum
osmolality (SOsm) 320 mOsm/kg. With concern that IHD might
contribute to cerebral edema and tenuous fluid status, CVVHD was
initiated with dialysate osmolality 292 mOsm/kg. Over the next
5 days, adequate fluid and solute clearance was achieved with
normalization of SOsm and ICP and good neurologic outcome.
320 30
20
S Osm
ICP
300
10
280 0
3 4 5 6 7
Hospital Day
S Osm ICP