Introduction

Continuous renal replacement therapy (CRRT) is the preferred choice for blood purification and volume control in critically ill children [1,2]. The reported overall survival rate for children requiring CRRT is 60% [3], and mortality in infants is comparable with that of older children and adolescents [4].

It is imperative in paediatric CRRT that equipment be adaptable to accommodate large variations in size (2–100 kg). Until recently at the Hospital for Sick Children, CRRT was provided using the PRISMA™ (Gambro AB, Stockholm, Sweden) circuits designed for adult use. A ‘Hot-Line™’(Smiths-Level 1 Inc., Rockland, MA) blood warmer prevented hypothermia but increased the extracorporeal circuit volume by 25 ml, which then exceeded 10% of the blood volume for patients weighing <15 kg. The PRISMA M10™, a 50 ml paediatric circuit, reduced the weight for a blood prime to ≤8 kg, but we elected not to use it given its limited functional capabilities based on product specifications. The PRISMA™ haemofilter AN69 membrane is associated with a ‘bradykinin release phenomenon’ on contact with acidotic blood including a blood prime [5], producing transient, but potentially life-threatening cardiovascular instability [6,7]. Protocols to overcome this, incorporating administration of large doses of alkali and calcium to the patient [8], are labour-intensive and were, in our experience, largely unsuccessful.

To address these concerns, the AQUARIUS™ (Edwards Lifesciences AG, Irvine, CA, USA) was purchased in 2004. In the paediatric programme the machine's specifications include a scale accuracy of ±20 ml and the ability to support a blood-flow rate of 10 to 200 ml/min, a pre-dilution rate of 0 or 100–6000 ml/h and post-dilution rate of 0 or 100–4000 ml/h [9]. In conjunction with this we introduced the AQUAMAX™ (Edwards Lifesciences AG, Irvine, CA, USA) polyethersulphone filters (HFO3—0.3 m2 surface area, priming volume of 32 ml or HFO7—0.7 m2 surface area, priming volume of 54 ml) and Aqualine tubing (Aqualine—110 ml priming volume and paediatric Aqualine S—64 ml priming volume) [10]. We report 14 months of clinical experience of CRRT with this equipment in the first 11 children, highlighting the success, benefits and technical issues.

Methods

We conducted a Research Ethics Board-approved retrospective chart review of the first 11 children treated with the AQUARIUS™ from August 2004 to October 2005 to obtain information on demographic factors, CCRT prescription, circuit life span, haemodynamic stability on initiating therapy, feasibility and technical problems. The patients were co-managed by the intensivists and nephrologists with CRRT orders written by the attending nephrologist.

Results

Table 1 shows the characteristics of the 11 children treated with the AQUARIUS™. Their mean age was 3.93 ± 5.99 years (mean ± SD) and mean weight 18.2 ± 20.2 kg.

Table 1.

Patient characteristics

PatientAgeWeight (kg)DiseaseOrgan failureCoagulopathyBaseline urine output (ml/kg/h)
111 m7.5Post BM transplant pneumoniaRenalNo2.6
Pulmonary
24 d3.0HyperammonaemiaNo1.6
38.5 m6.3Macrophage activation syndromeMultiple organ failureYes1.0
410 m7.3Neonatal liver failureMultiple organ failureYes0.1
54 d3.1HyperammonemiaNo3.1
616 y58.6Gram positive sepsisMultiple organ failureYes0.0
Intracerebral haemorrhage
DKA
73 m7.8Gram positive sepsisPulmonaryNo0.9
Cardiac arrhythmiasRenal
815 y50Reyes syndromeRenalYes4.5
PneumoniaLiver
Intracranial hypertension
96 y36.5Post BM transplantLiverYes0.6
Adenovirus viraemiaRenal
103 y13.0Diarrhoea + HUS (CNS involvement)RenalYes0.0
116 m6.8Congenital heart diseaseMultiple organ failureNo0.1
PatientAgeWeight (kg)DiseaseOrgan failureCoagulopathyBaseline urine output (ml/kg/h)
111 m7.5Post BM transplant pneumoniaRenalNo2.6
Pulmonary
24 d3.0HyperammonaemiaNo1.6
38.5 m6.3Macrophage activation syndromeMultiple organ failureYes1.0
410 m7.3Neonatal liver failureMultiple organ failureYes0.1
54 d3.1HyperammonemiaNo3.1
616 y58.6Gram positive sepsisMultiple organ failureYes0.0
Intracerebral haemorrhage
DKA
73 m7.8Gram positive sepsisPulmonaryNo0.9
Cardiac arrhythmiasRenal
815 y50Reyes syndromeRenalYes4.5
PneumoniaLiver
Intracranial hypertension
96 y36.5Post BM transplantLiverYes0.6
Adenovirus viraemiaRenal
103 y13.0Diarrhoea + HUS (CNS involvement)RenalYes0.0
116 m6.8Congenital heart diseaseMultiple organ failureNo0.1

DKA, diabetic ketoacidosis; BM, bone marrow; HUS, haemolytic uremic syndrome; CNS, central nervous system. Age represented as months (m), years (y) and days (d).

Table 1.

Patient characteristics

PatientAgeWeight (kg)DiseaseOrgan failureCoagulopathyBaseline urine output (ml/kg/h)
111 m7.5Post BM transplant pneumoniaRenalNo2.6
Pulmonary
24 d3.0HyperammonaemiaNo1.6
38.5 m6.3Macrophage activation syndromeMultiple organ failureYes1.0
410 m7.3Neonatal liver failureMultiple organ failureYes0.1
54 d3.1HyperammonemiaNo3.1
616 y58.6Gram positive sepsisMultiple organ failureYes0.0
Intracerebral haemorrhage
DKA
73 m7.8Gram positive sepsisPulmonaryNo0.9
Cardiac arrhythmiasRenal
815 y50Reyes syndromeRenalYes4.5
PneumoniaLiver
Intracranial hypertension
96 y36.5Post BM transplantLiverYes0.6
Adenovirus viraemiaRenal
103 y13.0Diarrhoea + HUS (CNS involvement)RenalYes0.0
116 m6.8Congenital heart diseaseMultiple organ failureNo0.1
PatientAgeWeight (kg)DiseaseOrgan failureCoagulopathyBaseline urine output (ml/kg/h)
111 m7.5Post BM transplant pneumoniaRenalNo2.6
Pulmonary
24 d3.0HyperammonaemiaNo1.6
38.5 m6.3Macrophage activation syndromeMultiple organ failureYes1.0
410 m7.3Neonatal liver failureMultiple organ failureYes0.1
54 d3.1HyperammonemiaNo3.1
616 y58.6Gram positive sepsisMultiple organ failureYes0.0
Intracerebral haemorrhage
DKA
73 m7.8Gram positive sepsisPulmonaryNo0.9
Cardiac arrhythmiasRenal
815 y50Reyes syndromeRenalYes4.5
PneumoniaLiver
Intracranial hypertension
96 y36.5Post BM transplantLiverYes0.6
Adenovirus viraemiaRenal
103 y13.0Diarrhoea + HUS (CNS involvement)RenalYes0.0
116 m6.8Congenital heart diseaseMultiple organ failureNo0.1

DKA, diabetic ketoacidosis; BM, bone marrow; HUS, haemolytic uremic syndrome; CNS, central nervous system. Age represented as months (m), years (y) and days (d).

Table 2 outlines the CRRT prescription. The continuous renal replacement therapy (CRRT) circuit was incorporated into an extracorporeal membrane oxygenation (ECMO) circuit in one patient. Initial placement of the access and return lines pre-oxygenator resulted in high access and transmembrane pressures (TMP) and was resolved by placing the access line post-oxygenator. The mean duration of therapy was 88.9 ± 106 h (range 23–371 h) and the mean circuit lifespan was 25.8 ± 21.1 h (range 1–75 h) (Table 2). Heparin was used exclusively in five patients; citrate anticoagulation was attempted in two patients both of whom developed metabolic alkalosis, and no anticoagulation was attempted in two patients which shortened the circuit life to 10.6 ± 4.4 h.

Table 2.

CRRT prescriptions, access, tubing and cardiovascular profile on initiating therapy

PatientAccessAquamax filter/tubingCRRT modeBlood flow (ml/min)Dialysate flow (ml/1.73 m2/h)Pre/post dilution flow (ml/1.73 m2/h)Systemic anti-coagulationCircuit life (h)Duration of therapy (h)InotropesPre-CRRT BPBP post-initiating CRRTBlood prime
18F-12cm HospalHFO3/aqualine SCVVH50–60Pre: 2160Heparin54201Dopamine113/55111/50Yes
CVVH50–60Pre: 870+ Post: 1300Heparin72–78
26.5F GamcathHFO3/aqualine SCVVHD50–601650–3700Heparin1–2023Dopamine56/4489/55Yes
36.5F GamcathHFO3/aqualine SCVVH30–50Pre: 1900–3300Heparin16–38371Milrinone, dopamine, epinephrine110/78105/75Yes
CVVHD501400–2800Citrate6–52
4Med comp 10 F SplitcathHFO3/aqualine SCVVHD50–801050–4200Heparin173481/49116/73Yes
56.5F GamcathHFO3/aqualine SCVVHD30–402900–3850Heparin444464/3968/48Yes
611.5F-12 cm MahurkarHFO7/aqualineCVVH100–160Pre: 2700Heparin20–2578Milrinone, dopamine105/64112/64No
CVVH130Post: 2700None2
CVVH100–160Pre: 2700None11
CVVHD1302700Citrate1
76.5F GamcathHFO3/aqualine SCVVH50Pre: 2160–2600Heparin8–2053Milrinone, dopamine, epinephrine105/5291/48Yes
811.5F-12 cm QuintonHFO7/aqualineCVVHD90–1001000–2100Heparin16–2541Dopamine, epinephrine, norepinephrine141/65128/86No
911.5F QuintonHFO7/aqualineCVVHD70–1501900Heparin3–2730114/68116/78No
108F-12 cm HospalHFO3/aqualine SCVVHD501900None4–123198/5799/49Yes
CVVHDF501900Post: 1900None15
11R carotid artery 12F, L.int jug 14FHFO3/aqualineCVVHD +ECMO*502000Heparin7272Milrinone, dopamine46/4346/44Yes
PatientAccessAquamax filter/tubingCRRT modeBlood flow (ml/min)Dialysate flow (ml/1.73 m2/h)Pre/post dilution flow (ml/1.73 m2/h)Systemic anti-coagulationCircuit life (h)Duration of therapy (h)InotropesPre-CRRT BPBP post-initiating CRRTBlood prime
18F-12cm HospalHFO3/aqualine SCVVH50–60Pre: 2160Heparin54201Dopamine113/55111/50Yes
CVVH50–60Pre: 870+ Post: 1300Heparin72–78
26.5F GamcathHFO3/aqualine SCVVHD50–601650–3700Heparin1–2023Dopamine56/4489/55Yes
36.5F GamcathHFO3/aqualine SCVVH30–50Pre: 1900–3300Heparin16–38371Milrinone, dopamine, epinephrine110/78105/75Yes
CVVHD501400–2800Citrate6–52
4Med comp 10 F SplitcathHFO3/aqualine SCVVHD50–801050–4200Heparin173481/49116/73Yes
56.5F GamcathHFO3/aqualine SCVVHD30–402900–3850Heparin444464/3968/48Yes
611.5F-12 cm MahurkarHFO7/aqualineCVVH100–160Pre: 2700Heparin20–2578Milrinone, dopamine105/64112/64No
CVVH130Post: 2700None2
CVVH100–160Pre: 2700None11
CVVHD1302700Citrate1
76.5F GamcathHFO3/aqualine SCVVH50Pre: 2160–2600Heparin8–2053Milrinone, dopamine, epinephrine105/5291/48Yes
811.5F-12 cm QuintonHFO7/aqualineCVVHD90–1001000–2100Heparin16–2541Dopamine, epinephrine, norepinephrine141/65128/86No
911.5F QuintonHFO7/aqualineCVVHD70–1501900Heparin3–2730114/68116/78No
108F-12 cm HospalHFO3/aqualine SCVVHD501900None4–123198/5799/49Yes
CVVHDF501900Post: 1900None15
11R carotid artery 12F, L.int jug 14FHFO3/aqualineCVVHD +ECMO*502000Heparin7272Milrinone, dopamine46/4346/44Yes

*‘Medtronic 1/4’ Carmeda coated circuit; oxygenator — Medos Hilite 2400 LT; priming volume — 400 ml.

Table 2.

CRRT prescriptions, access, tubing and cardiovascular profile on initiating therapy

PatientAccessAquamax filter/tubingCRRT modeBlood flow (ml/min)Dialysate flow (ml/1.73 m2/h)Pre/post dilution flow (ml/1.73 m2/h)Systemic anti-coagulationCircuit life (h)Duration of therapy (h)InotropesPre-CRRT BPBP post-initiating CRRTBlood prime
18F-12cm HospalHFO3/aqualine SCVVH50–60Pre: 2160Heparin54201Dopamine113/55111/50Yes
CVVH50–60Pre: 870+ Post: 1300Heparin72–78
26.5F GamcathHFO3/aqualine SCVVHD50–601650–3700Heparin1–2023Dopamine56/4489/55Yes
36.5F GamcathHFO3/aqualine SCVVH30–50Pre: 1900–3300Heparin16–38371Milrinone, dopamine, epinephrine110/78105/75Yes
CVVHD501400–2800Citrate6–52
4Med comp 10 F SplitcathHFO3/aqualine SCVVHD50–801050–4200Heparin173481/49116/73Yes
56.5F GamcathHFO3/aqualine SCVVHD30–402900–3850Heparin444464/3968/48Yes
611.5F-12 cm MahurkarHFO7/aqualineCVVH100–160Pre: 2700Heparin20–2578Milrinone, dopamine105/64112/64No
CVVH130Post: 2700None2
CVVH100–160Pre: 2700None11
CVVHD1302700Citrate1
76.5F GamcathHFO3/aqualine SCVVH50Pre: 2160–2600Heparin8–2053Milrinone, dopamine, epinephrine105/5291/48Yes
811.5F-12 cm QuintonHFO7/aqualineCVVHD90–1001000–2100Heparin16–2541Dopamine, epinephrine, norepinephrine141/65128/86No
911.5F QuintonHFO7/aqualineCVVHD70–1501900Heparin3–2730114/68116/78No
108F-12 cm HospalHFO3/aqualine SCVVHD501900None4–123198/5799/49Yes
CVVHDF501900Post: 1900None15
11R carotid artery 12F, L.int jug 14FHFO3/aqualineCVVHD +ECMO*502000Heparin7272Milrinone, dopamine46/4346/44Yes
PatientAccessAquamax filter/tubingCRRT modeBlood flow (ml/min)Dialysate flow (ml/1.73 m2/h)Pre/post dilution flow (ml/1.73 m2/h)Systemic anti-coagulationCircuit life (h)Duration of therapy (h)InotropesPre-CRRT BPBP post-initiating CRRTBlood prime
18F-12cm HospalHFO3/aqualine SCVVH50–60Pre: 2160Heparin54201Dopamine113/55111/50Yes
CVVH50–60Pre: 870+ Post: 1300Heparin72–78
26.5F GamcathHFO3/aqualine SCVVHD50–601650–3700Heparin1–2023Dopamine56/4489/55Yes
36.5F GamcathHFO3/aqualine SCVVH30–50Pre: 1900–3300Heparin16–38371Milrinone, dopamine, epinephrine110/78105/75Yes
CVVHD501400–2800Citrate6–52
4Med comp 10 F SplitcathHFO3/aqualine SCVVHD50–801050–4200Heparin173481/49116/73Yes
56.5F GamcathHFO3/aqualine SCVVHD30–402900–3850Heparin444464/3968/48Yes
611.5F-12 cm MahurkarHFO7/aqualineCVVH100–160Pre: 2700Heparin20–2578Milrinone, dopamine105/64112/64No
CVVH130Post: 2700None2
CVVH100–160Pre: 2700None11
CVVHD1302700Citrate1
76.5F GamcathHFO3/aqualine SCVVH50Pre: 2160–2600Heparin8–2053Milrinone, dopamine, epinephrine105/5291/48Yes
811.5F-12 cm QuintonHFO7/aqualineCVVHD90–1001000–2100Heparin16–2541Dopamine, epinephrine, norepinephrine141/65128/86No
911.5F QuintonHFO7/aqualineCVVHD70–1501900Heparin3–2730114/68116/78No
108F-12 cm HospalHFO3/aqualine SCVVHD501900None4–123198/5799/49Yes
CVVHDF501900Post: 1900None15
11R carotid artery 12F, L.int jug 14FHFO3/aqualineCVVHD +ECMO*502000Heparin7272Milrinone, dopamine46/4346/44Yes

*‘Medtronic 1/4’ Carmeda coated circuit; oxygenator — Medos Hilite 2400 LT; priming volume — 400 ml.

The AQUARIUS™ fluid warmer was able to maintain normothermia in the three older children, but not in six children below 1 year of age. One child was febrile and required no warming, and the remaining patient was on an ECMO circuit. The Hotline™ restored normothermia but increased the extracorporeal volume by 25 ml.

All patients were haemodynamically stable on initiating CRRT (Table 2). The difference in the systolic and diastolic blood pressure (BP) before and after starting was −4.36 mmHg [95% confidence interval (CI) −15.1 to 6.4] and −5.09 mmHg (95% CI −12.29 to 2.11), respectively.

Among the 11 children receiving CRRT, 7 survived, achieving an overall survival rate of 64%.

For full specifications of the AQUARIUS™ and AQUAMAX™, we refer the reader to the company literature [9,10]. The benefits and technical issues are outlined in Table 3 and in the discussion.

Table 3.

Technical problems encountered with the AQUARIUS™, and possible solutions

Problem (consequences)Explanation, actual and theoretical solutions
(1) Maintaining normothermiaAddition of the Hotline™ to CRRT circuit
(2) Frequent TMP alarms, with minute-to-minute variation between high and low pressures (frequent interruptions to treatment and increased clotting tendency)Pumps speed up and down to give the desired fluid delivery and loss
Revision of the software led to a partial resolution but also increased the minimum replacement and dialysis fluid rates in the paediatric mode from 100 ml/h to ≥600 ml/h. By maintaining these higher rates the issue has largely resolved
(3) The minimum recommended dialysate and replacement fluid rate in the paediatric program is 600 ml/h (excessive clearance of drugs and solutes)
(4) Balance alarmsMalfunction of the effluent pump resolved by revising the software. Alarms decreased by keeping the dialysate and replacement fluid rate ≥600 ml/h
• frequent balance alarms (stopping the blood pump if deviation >20 ml in the paediatric program)Persistence of balance alarms—possibly from partial obstruction of the substitution fluid flow exiting from the bags (cone port connection difficult & the broken seal occludes the water channel; spiking the bag may cause kinking of tubing).
• sudden spinning of effluent pump [excessive ultrafiltration (UF)]PRISMA™ effluent bags not used on AQUARIUS™ (PRISMA™ fill top-down, AQUARIUS™ from bottom-up)
• dialysate leak with PRISMA™ effluent bags (inaccurate balance records)
(5) Negative ‘pressure drop’ i.e. return pressure > access pressureMore likely with low blood-flow rate, small filter, addition of the Hotline™ to the circuit or post-dilutional CVVHDF
Lower limit: −30 mmHg
Increase the pre-dilution or blood-flow rate
(6) Recurrent clots in the circuit, especially in the bubble chamberMaintain ACT 170–220 if low blood flow rate, or history of recurrent clotting
Limit the filtration fraction (<20%)
Prophylactic circuit change 48 h
Consider pre-filtered normal saline infusion or dilute the heparin and run from a separate infusion pump
(7) Single weigh scale for dialysate and replacement fluid (only one solution can be used in the CVVHDF mode; high risk of developing metabolic alkalosis on citrate anticoagulation with commercially available solutions)Citrate anticoagulation with a pharmacy-prepared low-bicarbonate solution
CVVHD using an external infusion to give normal saline (declined for safety issues)
Consider heparin-free or tight heparin protocols in conjunction with frequent saline flushes pre-filter
(8) Only post-dilutional CVVHDF is possible (increased risk of clotting; negative ‘pressure drop’)In the CVVHDF mode the pre-filter pump drives the dialysate and the post-filter pump drives the replacement fluid
(9) Leak from the back of the return-pressure sensor (once—during therapy; multiple—on dismantling)If sudden drop in return-pressure check for a blood leak
Policy of wearing gloves on dismantling
(10) Master keyboard failure during priming (frequent attempts of priming resulting in excessive waste)Resolved by pressing the blood pump so that the screen and pump communicate
The software is currently undergoing revision
Problem (consequences)Explanation, actual and theoretical solutions
(1) Maintaining normothermiaAddition of the Hotline™ to CRRT circuit
(2) Frequent TMP alarms, with minute-to-minute variation between high and low pressures (frequent interruptions to treatment and increased clotting tendency)Pumps speed up and down to give the desired fluid delivery and loss
Revision of the software led to a partial resolution but also increased the minimum replacement and dialysis fluid rates in the paediatric mode from 100 ml/h to ≥600 ml/h. By maintaining these higher rates the issue has largely resolved
(3) The minimum recommended dialysate and replacement fluid rate in the paediatric program is 600 ml/h (excessive clearance of drugs and solutes)
(4) Balance alarmsMalfunction of the effluent pump resolved by revising the software. Alarms decreased by keeping the dialysate and replacement fluid rate ≥600 ml/h
• frequent balance alarms (stopping the blood pump if deviation >20 ml in the paediatric program)Persistence of balance alarms—possibly from partial obstruction of the substitution fluid flow exiting from the bags (cone port connection difficult & the broken seal occludes the water channel; spiking the bag may cause kinking of tubing).
• sudden spinning of effluent pump [excessive ultrafiltration (UF)]PRISMA™ effluent bags not used on AQUARIUS™ (PRISMA™ fill top-down, AQUARIUS™ from bottom-up)
• dialysate leak with PRISMA™ effluent bags (inaccurate balance records)
(5) Negative ‘pressure drop’ i.e. return pressure > access pressureMore likely with low blood-flow rate, small filter, addition of the Hotline™ to the circuit or post-dilutional CVVHDF
Lower limit: −30 mmHg
Increase the pre-dilution or blood-flow rate
(6) Recurrent clots in the circuit, especially in the bubble chamberMaintain ACT 170–220 if low blood flow rate, or history of recurrent clotting
Limit the filtration fraction (<20%)
Prophylactic circuit change 48 h
Consider pre-filtered normal saline infusion or dilute the heparin and run from a separate infusion pump
(7) Single weigh scale for dialysate and replacement fluid (only one solution can be used in the CVVHDF mode; high risk of developing metabolic alkalosis on citrate anticoagulation with commercially available solutions)Citrate anticoagulation with a pharmacy-prepared low-bicarbonate solution
CVVHD using an external infusion to give normal saline (declined for safety issues)
Consider heparin-free or tight heparin protocols in conjunction with frequent saline flushes pre-filter
(8) Only post-dilutional CVVHDF is possible (increased risk of clotting; negative ‘pressure drop’)In the CVVHDF mode the pre-filter pump drives the dialysate and the post-filter pump drives the replacement fluid
(9) Leak from the back of the return-pressure sensor (once—during therapy; multiple—on dismantling)If sudden drop in return-pressure check for a blood leak
Policy of wearing gloves on dismantling
(10) Master keyboard failure during priming (frequent attempts of priming resulting in excessive waste)Resolved by pressing the blood pump so that the screen and pump communicate
The software is currently undergoing revision

Hotline™ (Smiths-Level 1 Inc., Rockland, MA, USA); PRISMA™ (Gambro AB, Stockholm, Sweden); AQUARIUS™ (Edwards Lifesciences AG, Irvine, CA, USA).

Table 3.

Technical problems encountered with the AQUARIUS™, and possible solutions

Problem (consequences)Explanation, actual and theoretical solutions
(1) Maintaining normothermiaAddition of the Hotline™ to CRRT circuit
(2) Frequent TMP alarms, with minute-to-minute variation between high and low pressures (frequent interruptions to treatment and increased clotting tendency)Pumps speed up and down to give the desired fluid delivery and loss
Revision of the software led to a partial resolution but also increased the minimum replacement and dialysis fluid rates in the paediatric mode from 100 ml/h to ≥600 ml/h. By maintaining these higher rates the issue has largely resolved
(3) The minimum recommended dialysate and replacement fluid rate in the paediatric program is 600 ml/h (excessive clearance of drugs and solutes)
(4) Balance alarmsMalfunction of the effluent pump resolved by revising the software. Alarms decreased by keeping the dialysate and replacement fluid rate ≥600 ml/h
• frequent balance alarms (stopping the blood pump if deviation >20 ml in the paediatric program)Persistence of balance alarms—possibly from partial obstruction of the substitution fluid flow exiting from the bags (cone port connection difficult & the broken seal occludes the water channel; spiking the bag may cause kinking of tubing).
• sudden spinning of effluent pump [excessive ultrafiltration (UF)]PRISMA™ effluent bags not used on AQUARIUS™ (PRISMA™ fill top-down, AQUARIUS™ from bottom-up)
• dialysate leak with PRISMA™ effluent bags (inaccurate balance records)
(5) Negative ‘pressure drop’ i.e. return pressure > access pressureMore likely with low blood-flow rate, small filter, addition of the Hotline™ to the circuit or post-dilutional CVVHDF
Lower limit: −30 mmHg
Increase the pre-dilution or blood-flow rate
(6) Recurrent clots in the circuit, especially in the bubble chamberMaintain ACT 170–220 if low blood flow rate, or history of recurrent clotting
Limit the filtration fraction (<20%)
Prophylactic circuit change 48 h
Consider pre-filtered normal saline infusion or dilute the heparin and run from a separate infusion pump
(7) Single weigh scale for dialysate and replacement fluid (only one solution can be used in the CVVHDF mode; high risk of developing metabolic alkalosis on citrate anticoagulation with commercially available solutions)Citrate anticoagulation with a pharmacy-prepared low-bicarbonate solution
CVVHD using an external infusion to give normal saline (declined for safety issues)
Consider heparin-free or tight heparin protocols in conjunction with frequent saline flushes pre-filter
(8) Only post-dilutional CVVHDF is possible (increased risk of clotting; negative ‘pressure drop’)In the CVVHDF mode the pre-filter pump drives the dialysate and the post-filter pump drives the replacement fluid
(9) Leak from the back of the return-pressure sensor (once—during therapy; multiple—on dismantling)If sudden drop in return-pressure check for a blood leak
Policy of wearing gloves on dismantling
(10) Master keyboard failure during priming (frequent attempts of priming resulting in excessive waste)Resolved by pressing the blood pump so that the screen and pump communicate
The software is currently undergoing revision
Problem (consequences)Explanation, actual and theoretical solutions
(1) Maintaining normothermiaAddition of the Hotline™ to CRRT circuit
(2) Frequent TMP alarms, with minute-to-minute variation between high and low pressures (frequent interruptions to treatment and increased clotting tendency)Pumps speed up and down to give the desired fluid delivery and loss
Revision of the software led to a partial resolution but also increased the minimum replacement and dialysis fluid rates in the paediatric mode from 100 ml/h to ≥600 ml/h. By maintaining these higher rates the issue has largely resolved
(3) The minimum recommended dialysate and replacement fluid rate in the paediatric program is 600 ml/h (excessive clearance of drugs and solutes)
(4) Balance alarmsMalfunction of the effluent pump resolved by revising the software. Alarms decreased by keeping the dialysate and replacement fluid rate ≥600 ml/h
• frequent balance alarms (stopping the blood pump if deviation >20 ml in the paediatric program)Persistence of balance alarms—possibly from partial obstruction of the substitution fluid flow exiting from the bags (cone port connection difficult & the broken seal occludes the water channel; spiking the bag may cause kinking of tubing).
• sudden spinning of effluent pump [excessive ultrafiltration (UF)]PRISMA™ effluent bags not used on AQUARIUS™ (PRISMA™ fill top-down, AQUARIUS™ from bottom-up)
• dialysate leak with PRISMA™ effluent bags (inaccurate balance records)
(5) Negative ‘pressure drop’ i.e. return pressure > access pressureMore likely with low blood-flow rate, small filter, addition of the Hotline™ to the circuit or post-dilutional CVVHDF
Lower limit: −30 mmHg
Increase the pre-dilution or blood-flow rate
(6) Recurrent clots in the circuit, especially in the bubble chamberMaintain ACT 170–220 if low blood flow rate, or history of recurrent clotting
Limit the filtration fraction (<20%)
Prophylactic circuit change 48 h
Consider pre-filtered normal saline infusion or dilute the heparin and run from a separate infusion pump
(7) Single weigh scale for dialysate and replacement fluid (only one solution can be used in the CVVHDF mode; high risk of developing metabolic alkalosis on citrate anticoagulation with commercially available solutions)Citrate anticoagulation with a pharmacy-prepared low-bicarbonate solution
CVVHD using an external infusion to give normal saline (declined for safety issues)
Consider heparin-free or tight heparin protocols in conjunction with frequent saline flushes pre-filter
(8) Only post-dilutional CVVHDF is possible (increased risk of clotting; negative ‘pressure drop’)In the CVVHDF mode the pre-filter pump drives the dialysate and the post-filter pump drives the replacement fluid
(9) Leak from the back of the return-pressure sensor (once—during therapy; multiple—on dismantling)If sudden drop in return-pressure check for a blood leak
Policy of wearing gloves on dismantling
(10) Master keyboard failure during priming (frequent attempts of priming resulting in excessive waste)Resolved by pressing the blood pump so that the screen and pump communicate
The software is currently undergoing revision

Hotline™ (Smiths-Level 1 Inc., Rockland, MA, USA); PRISMA™ (Gambro AB, Stockholm, Sweden); AQUARIUS™ (Edwards Lifesciences AG, Irvine, CA, USA).

Discussion

Benefits

With the AQUARIUS™ it was feasible to perform CRRT on children of all ages. The circuit was easily incorporated into an ECMO circuit. Universal demonstration of haemodynamic stability on initiating therapy supports the absence of bradykinin-mediated membrane reactions.

Synthetic dialysers such as the AN69 and polysulphone membranes are ideal for convective therapies as they are highly permeable, thick-walled membranes and can therefore tolerate high transmembrane pressures while achieving good middle-molecule clearance. In vitro data suggest that AN69 membranes are more biocompatible than polysulphone membranes resulting in less complement [11], B-cell and monocyte activation [12] and have superior protein and cytokine adsorption abilities [13]. In practical terms, these benefits became insignificant when faced with the possibility of severe life-threatening membrane reactions. New surface-treated AN69ST membranes hold the promise of complete suppression of potential blood–membrane contact reactions, independent of pH [14], but paediatric safety and efficacy data are currently not available.

Application of the AQUARIUS™ circuit did expose several technical issues, which have been summarized in point format in Table 3:

(1) Hypothermia

The fluid warmer was insufficient to prevent hypothermia in all infants despite additional warming strategies such as overhead radiant heaters. The Hotline™ blood warmer remedied the hypothermia but at the expense of adding 25 ml to the extracorporeal circuit volume.

(2) and (3) TMP alarms and minimum substitution fluid rate

Minute-to-minute variations in the TMP interrupted the treatment and became more prevalent if a zero hourly fluid balance was prescribed. We believe this occurred because the pumps were changing speed to maintain the 20 ml error in fluid balance throughout the therapy. New software (v4.01.01) has resolved the issue but at the cost of pushing the minimum recommended dialysate or replacement fluid rates in the paediatric setting to 600 ml/h. This places infants at risk of excessive drug and solute clearance and disequilibrium. Dialysate has a smaller effect on TMP; therefore, on continuous veno-venous haemodialysis (CVVHD) reduced dialysate flow, and thus reduced clearance, may be possible without triggering TMP alarms. We have not yet formally tested this theory.

(4) Balance alarms

The aforementioned software issue also resulted in sudden spinning of the effluent pump that has resolved with the software upgrade; however, balance alarms persist. One cause was the use of PRISMA™ bags on the AQUARIUS™ resulting in dialysate leak. The design of the connection between the substitution fluid bags and the circuit also appears to cause intermittent obstruction to flow. Finally, during periods of high pump-generated filtrate pressures, balance alarms are triggered. This can be remedied by allowing a higher ultrafiltrate rate or switching to convective therapies and increasing the pre-dilution rate.

(5) Negative ‘pressure drop’

The negative pressure drop refers to higher hydraulic pressure on the venous side compared with the arterial side of the haemofilter. This can be overcome by increasing the blood flow (limited by access in infants) or switching to continuous veno-venous haemofiltration (CVVH) and increasing pre-dilution-flow rates to make the TMP trend positive.

(6) Recurrent clots

In the absence of a coagulopathy, if the circuit lifespan was <24 h, the target activated clotting time was increased to 170–220 s for subsequent circuits. Other solutions are outlined in Table 3.

(7) Single substitution fluid weighing scale

Citrate chelation of calcium produces regional anticoagulation but can result in alkalosis when used in conjunction with commercially available solutions [15–16]. This can be corrected by converting to CVVHDF, infusing normal saline as the replacement fluid [17].

The AQUARIUS™ has two weigh scales, one for the effluent fluid and the other for dialysate or replacement fluid. During continuous veno-venous haemodialfiltration (CVVHDF), different dialysate and replacement fluid rates can be employed but only one substitution fluid can be used. Normocarb™ (Dialysis Solutions Inc., Whitby, Canada), our solution of choice, contains 35 mmol/l bicarbonate making it unsuitable for alkalotic patients. Furthermore, it is not licensed for infusion. Therefore citrate anticoagulation cannot be applied without a custom-made solution. A theoretical alternative includes CVVHD with a lower bicarbonate, calcium-free dialysate and higher dialysate flow rates.

(8) Post-dilutional CVVHDF

This increases the risk of filter clotting, and in circuits with the Hotline™ warmer, will push return pressures higher and may, compared with pre-dilutional CVVHDF, push clearance rates higher [18] in the already vulnerable infants.

(9) Return-pressure-sensor leak

We have alerted the company and have increased vigilance for sudden drops in the access and return pressures.

(10) Master keyboard failure

The keyboard failure during priming is essentially a software issue that is undergoing revision. In the interim, pressing the blood pump prompts the screen and pump to communicate.

Conclusion

In our unit the leading cause of morbidity during initiation of CRRT is haemodynamic instability particularly using blood primes. By using the AQUARIUS™ and AQUAMAX™ filters, we saw an improvement in patient well-being, a reduction in acute adverse effects on initiating therapy and the safer and broader applicability of CRRT in children of all ages. The trade-off, however, was a number of technical issues. Some have been resolved, but of those remaining, the absence of safe and simple protocols for citrate anticoagulation and the obligatory 600 ml/h minimum replacement and dialysate fluid rates have the greatest potential impact on patient care.

Conflict of interest statement. None declared.

References

1

Warady B, Bunchman T. Dialysis therapy for children with acute renal failure: survey results.

Pediatr Nephrol
2000
;
15
:
11
–13

2

Burchardi H. Renal replacement therapy (RRT) in the ICU: criteria for initiating RRT. In: Ronco C, Bellomo R, La Greca G., eds.

Blood Purification in Intensive Care (Contributions to Nephrology V 132- Berlyne GM and Ronco C)
. Karger New York:
2001
;
171
–180

3

Goldstein S, Currier H, Graf J, Cosio CC, Brewer ED, Sachdeva R. Outcome in children receiving continuous venovenous hemofiltartion.

Pediatrics
2001
;
107
:
1309
–1312

4

Symons JM, Brody PD, Gregory MJ, et al. Continuous renal replacement therapy in children up to 10 kg.

Am J Kidney Dis
2003
;
41
:
984
–989

5

Lacour F, Maheut H. AN69 membrane and conversion enzyme inhibitors: prevention of anaphylactic shock by alkaline rinsing?.

Nephrologie
1992
;
13
:
135
–136

6

Tournier O, Delaunay M.

The AN-69 dialysis membrane
. Hospital Product Pamphlet
1995
1
–36

7

Renaux JL, Thomas M.

Hypersensitivity reactions and dialysis membranes
. Hospital Product Pamphlet
1996
1
–19

8

Bunchman TE, Maxvold NJ, Barnett J, Hutchings A, Benfield MR. Pediatric hemofiltration: Normocarb® dialysate solution with citrate anticoagulation.

Pediatr Nephrol
2002
;
17
:
150
–154

9

Edwards Lifesciences SA. Edwards Aquarius – Setting new standards for acute blood purification.

Product Pamphlet
2002
; Aquarius Specifications

10

Edwards Lifesciences SA. Aquamax filters.

Product Specification Pamphlet
2002

11

Schaefer RM, Kulzer P, Gilge U, Schaefer L, Heidland A. Clinical evaluation of the new, steam-sterilized polysulfone high-flux dialyzer.

Clin Nephrol
1994
;
42 [Suppl 1]
:
S21
–S25

12

Descamps-Latscha B, Herbelin A, Nguyen AT, et al. Soluble CD23 as an effector of immune dysregulation in chronic uremia and dialysis.

Kidney Int
1993
;
43
:
878
–884

13

Lonnemann G, Schindler R, Dinarello CA, Koch KM. Removal of circulating cytokines by hemodialysis membranes in vitro. In; Faist E, Meakins J, Schildberg FW, eds.

Host Defense Dysfunction in Trauma, Shock and Sepsis
,Springer Verlag, Berlin:
1993
;
613
–623

14

Hospal AG. The Nephral ST series.

Product Specification Pamphlet
2005

15

Mehta RL, McDonald BR, Aguilar MM, Ward DM. Regional citrate anticoagulation for continuous arteriovenous hemodialysis in critically ill patients.

Kidney Int
1990
;
38
:
976
–981

16

Tolwani A, Campbell R, Schenk M, Allon M, Warnock DG. Simplified citrate anticoagulation for continuous renal replacement therapy.

Kidney Int
2001
;
60
:
370
–374

17

Chadha V, Garg U, Warady BA, Alon US. Citrate clearance in children receiving continuous venovenous renal replacement therapy.

Pediatr Nephrol
2002
;
17
:
819
–824

18

Parakininkas D, Greenbaum LA. Comparison of solute clearance in three modes of continuous renal replacement therapy.

Pediatr Crit Care Med
2004
;
5
:
269
–274

Author notes

1Division of Nephrology and 2Pediatric Critical Care Unit, Hospital for Sick Children, Toronto, Canada

Comments

0 Comments
Submit a comment
You have entered an invalid code
Thank you for submitting a comment on this article. Your comment will be reviewed and published at the journal's discretion. Please check for further notifications by email.