Only one randomised trial fulfilled the criteria for inclusion in the analysis. This study compared single and double volume exchange transfusion in jaundice due to ABO hemolytic jaundice. The study found no significant difference in bilirubin levels following exchange. This study did not look at any long term neurodevelopmental outcome (brain damage). Based on the available data, there is insufficient evidence to support or refute the use of single volume exchange transfusion as opposed to double volume exchange transfusion in jaundiced newborns.
Prior to the introduction of exchange transfusion, liveborn infants with severe rhesus hemolytic disease had a 35-40% mortality, with a 90% risk of severe neurological damage among survivors (Diamond 1948). A reduction in mortality to 20% and reduction in adverse neurological outcome to 30% was observed following the introduction of double volume exchange transfusion (Diamond 1948).
Although exchange transfusion achieved its goal of reducing mortality, it has become apparent over the years that the outcome of many of the sensitised infants who underwent exchange transfusion would have been no worse if they had not been treated (Lucey 1966). Over the past few decades, major advances have been made in prevention (e.g. rhesus immunoglobulin) and treatment (e.g. intrauterine transfusions, intensive phototherapy) of rhesus hemolytic disease, so that use of exchange transfusion, especially within twelve hours of birth, has become uncommon. Most centres reporting outcomes in the management of severe Rhesus isoimmune hemolytic anaemia have shown markedly improved survival rates during the past 25 years, and much of this is attributable to improved antenatal management. Of equal importance, long-term neurodevelopmental outcome in these infants is excellent (Maisels 1999).
The efficacy of ET with regards to long term neurological outcome is related to the etiology of hyperbilirubinemia and the population in which ET is undertaken. ABO incompatibility often results in a less severe hemolytic disease, needing fewer exchange transfusions compared to rhesus incompatibility (Kanto 1978). In developing countries, babies with severe jaundice may be referred late (with babies already showing signs of kernicterus), and therefore exchange transfusion may not be useful in preventing neurological damage. Bilirubin encephalopathy occurs at lower serum bilirubin levels in preterm infants and the threshold for exchange transfusion is usually lower. Sub-group analyses will be done to compare single and double volume exchanges in these populations.
Efficacy and complications of exchange transfusion can also depend upon the route (peripheral vs. umbilical) (Merchant 1992), size of aliquots, method of exchange (continuous vs. push pull) (Schober 1990) and use of albumin during exchange. Use of intrauterine transfusions (Grannum 1988) and metalloporphyrins can reduce the need of ET (Dennery 2001). Sub-group analyses will be done with regards to different techniques used for exchange transfusion.
Mortality directly attributable to ET (reported to be at least 1%) is often due to unexplained cardiac arrest, cardiac arrhythmia or air embolism (Boggs 1960). Severe necrotising enterocolitis and bowel perforation requiring surgery occur in about 1% (Lucey 1969). Other major complications include severe bleeding or coagulopathy requiring intervention, and severe bradycardia or apnea during or following exchange. There are several reports of portal vein thrombosis and portal hypertension occurring following exchange transfusion using umbilical route (Yadav 1993).
It is likely that the complications of exchange would increase with the duration of exchange and volume of blood exchanged. Therefore, a double volume exchange is potentially more hazardous than single volume exchange. With improved outcomes of jaundiced babies without exchange, the risk: benefit ratio of exchange transfusion may well have changed. It is important to examine if limiting the procedure to a single volume exchange minimises adverse effects without compromising long term outcome.
This review aims to examine whether there is any evidence supporting the continued use of double, as opposed to single, volume exchanges with regards to efficacy (especially in terms of long term neurodevelopmental outcome) and safety.
ET was defined as any procedure in which whole blood was removed from the circulation and replaced with either fresh whole blood or diluted packed red blood cells. Phototherapy was defined as the use of artificial light to photoisomerise unconjugated bilirubin. A continuous technique for ET was defined as one in which two catheters were used and withdrawal and infusion of blood occurred simultaneously. A 'push-pull' technique for ET was defined as one in which one catheter was used and withdrawal and infusion of blood occurred sequentially. Central or peripheral vascular catheters was defined as catheters whose tip lies within the thorax or abdomen, or all other sites, respectively (Mills 2003)
The safety and efficacy of SVET and DVET were analysed in the following sub-groups: Rhesus hemolytic disease, hemolytic disease other than rhesus disease, infants without hemolysis, preterm babies, infants with established kernicterus, infants treated with metalloporphyrins and for various pre-exchange serum bilirubin levels. Further sub-group analyses was not undertaken on method of exchange transfusion (continuous vs push pull), route of exchange (peripheral vs central) and aliquot size because of insufficient data.
The standard statistical methods of the Cochrane Collaboration were used. For dichotomous and continuous variables, the relative risk (RR) and weighted mean difference (WMD) were calculated respectively. 95% confidence intervals were used and a fixed effects model was assumed. Heterogeneity in the study population was prospectively addressed by subgrouping on the basis of different etiologies of hyperbilirubinemia. Sub grouping was also done on the method of exchange transfusion to avoid heterogeneity related to intervention. Statistical heterogeneity was examined by I2 test.
Included Studies
Amato 1988 (Amato 1988)
Amato et al studied 20 full term newborn infants with hyperbilirubinemia due
to ABO incompatibility, requiring exchange transfusion. The diagnosis of ABO
isoimmunization was made on the basis of clinical data, mother infant blood
group incompatibility, positive direct or indirect Coombs test, high
reticulocyte count and hyperbilirubinemia. The study was done at two centers
over a one year period. 84 babies were admitted to the two intensive care units
with hyperbilirubinemia due to ABO incompatibility. Babies who had perinatal
asphyxia (Apgar score < 4 at one minutes and < 6 at five minutes), babies
with congenital malformations, suspected or proven bacterial infection,
respiratory distress and babies who had hyperbilirubinemia due to maternal
drugs, polycythemia, skin hematomas, cephalhematoma were excluded. No patients
were treated with phenobarbitone. Indication for exchange transfusion was
bilirubin levels according to the modified Polacek curves (Cockington
1979)
Twenty term babies who met the inclusion criteria, and in whom a decision to exchange was taken, were randomised into two groups i.e. one group receiving SVET and the other group receiving DVET. Packed cells (2/3 red cell volume and 1/3 plasma) suspended in plasma was used for exchange. No immunoglobulin or clotting factors were present. The mean hemoglobin level of the blood units was 20 g/dl (SD 3.1) and hematocrit 68% (SD 5).
Base line characteristics were similar with regards to birth weight 3260g (SD 390) vs. 3350g SD (410), gestational age 39 (SD 1) week vs. 40 (SD 0.8) weeks, 5 min Apgar scores 7.5 (SD 0.6) vs. 7.8 (SD 0.8) and age at exchange transfusion 17.2 (SD 6.3) hours vs. 18.7 (SD 6.2) hours in babies who received single volume and double volume exchange transfusion. The immediate pre exchange hemoglobin levels 16.7 (SD 2.4) g/dL vs. 16.6 (SD 2.5) g/dL, platelet levels 253 (SD 83) x 109/L vs. 298 (SD 95) x 109/L, rate of rise in bilirubin 199 (SD 33) micromol/L vs. 216 (SD 55) micromols/L and bilirubin level 199 (SD 33) micromol/L vs. 216 (SD 55) micromol/L were similar single and double volume exchange groups. Both groups were treated equally apart from the volume of blood used for exchange transfusion.
Single volume exchange transfusion transfused 80 ml/kg and double volume
exchange 180 ml/kg. Exchange transfusion was performed through umbilical vein
over one hour using a disposable exchange transfusion set in 10 ml aliquots. No
additional calcium or albumin was given. No deficit was created. Continuous
phototherapy was started immediately after ET using double blue light
phototherapy. Babies were given 10% dextrose at 120 ml/kg and once bilirubin
levels were less than 205 mmol/L, phototherapy was discontinued. None of the
babies required more than one exchange transfusion.
Excluded Studies
No other studies comparing single and double volume exchange
transfusions were found. Several case series were found on double volume
exchange transfusions and one case series on single volume exchange transfusion
(Soulie
1999). Soulie et al retrospectively reviewed 60 exchange transfusions in 48
newborns (26 babies with Rhesus hemolytic disease) who had single volume
exchange transfusion (0.72 - 1 times blood volume exchanged), with satisfactory
lowering of bilirubin levels. However, no long term neurological effects were
examined.
Amato 1988 (Amato 1988)
Full details are provided in the Table, Characteristics of Included Studies.
Method of subject allocation: Table of random numbers was used for
randomisation.
Concealment of allocation: There was no mention of allocation
concealment.
Masking of caregivers: This was not done in the study.
Completeness of outcome assessment: Outcomes were reported in all randomised
infants up to 3 months.
Masking of outcome assessment: We could not
ascertain if this was done in the study.
Secondary
Total number of exchange transfusions:
None
of the babies needed a second exchange transfusion in either group.
Hemoglobin (g/dl) or hematocrit (%) measured 3-6 weeks after exchange
transfusion:
No significant difference was seen in either group at 10
days and at three month follow up after exchange. The mean haemoglobin
immediately post exchange was higher in double volume group 20.4 g/dl (SD 2.1)
than single volume exchange transfusion group 18.3 (SD1.1) g/dL (p = 0.01)
(Outcome 01.02). This was not a secondary outcome planned in the review.
Thrombocytopenia or coagulopathy following exchange, receiving
intervention:
No intervention was needed in any of the groups in the
study. Mean platelet levels immediately post exchange was significantly lower in
double volume group. At 10 days post exchange there was no difference in
platelet levels (Outcome 01.03).
Infection attributable to exchange
transfusion:
This outcome was not reported.
Portal vein
thrombosis and/or portal hypertension following exchange:
This outcome
was not reported.
Necrotising enterocolitis within 72 hours of
exchange:
None of the infants developed necrotising enterocolitis.
When introduced in the 1940's, exchange transfusion (double volume) was shown to reduce the incidence of kernicterus in Rhesus hemolytic disease (Diamond 1948). Early exchange transfusions were often performed early in order to remove hemolysed cells, rather than to reduce serum bilirubin. Studies have shown that the double volume exchange removes 80 - 85 % of RBC and single volume exchange removes about 65% of the circulating RBCs (Lathe 1955; Sproul 1961).
Forfar et al measured the bilirubin mass removed during exchange transfusion and correlated this with the volume of exchange. The bilirubin removed was 45% higher than the fall in serum bilirubin (Forfar 1958). This is because of the movement of bilirubin from the tissues into the blood during exchange. The bilirubin removed correlated with the volume of exchange (r = +0.34). An extra 20 - 70% increase in amount of bilirubin was noticed when single volume is increased to a double volume exchange. An exchange volume of 160 - 180 ml/kg resulted in optimum removal of bilirubin. Further increase in exchange volume did remove more bilirubin, but to a much lesser extent. Authors have considered lowering of total body bilirubin, rather than serum bilirubin, more important in prevention of kernicterus and, therefore, recommended double volume exchange transfusion as opposed to single volume in Rhesus hemolytic disease.
The spectrum of hemolytic jaundice and indications for exchange transfusion in jaundice has changed markedly in the past decade in developed countries following the increased use of Rhesus anti D. Exchange transfusion for Rhesus incompatibility has become uncommon and early exchange transfusions are hardly undertaken these days. It is not clear the same criteria for the exchange volume used in cases of Rhesus isoimmunization would apply to jaundiced babies without Rhesus hemolytic disease ( e.g. ABO incompatibility). Since the hemolysis is much lesser, one could argue that single volume exchange transfusion is as effective as double volume exchange for non Rhesus hemolysis with regards to neurodevelopmental outcome and may lead to fewer side effects. However, currently there is insufficient evidence to support a change in the common practice from double volume exchange to single volume exchange for non Rhesus conditions.
| Study | Methods | Participants | Interventions | Outcomes | Notes | Allocation concealment |
| Amato 1988 | Prospective Randomised Control trial Blinding of intervention: No |
20 full term infants with jaundice due to ABO hemolytic disease | Single volume (80 ml/kg) vs double volume exchange transfusions(180 ml/kg) | Bilirubin, platelets and hamoglobin level after exchange transfusion, duration of phototherapy, need of further exchange transfusion and adverse events during exchange transfusion | B |
| Study | Reason for exclusion |
| Soulie 1999 | Retrospective case series on single volume exchange transfusion. No comparison with double volume exchange is given |
* Amato M, Blumberg A, Hermann U Jr, Zurbrugg R. Effectiveness of single versus double volume exchange transfusion in newborn infants with AB0 hemolytic disease. Helvetica Paediatrica Acta 1988;43:177-86.
Soulie JC, Larsen M, Andreu G, Berry M, Gabai A et al. Retrospective study of exchange transfusion for newborn infants with reconstituted blood. Review of 60 exchanges. Transfusion Clinique et Biologique 1999;6:166-73.
* indicates the primary reference for the study
Bell MJ, Ternberg JL, Feigin RD, Keating JP, Marshall R, Barton L, Brotherton T. Neonatal necrotizing enterocolitis. Therapeutic decisions based upon clinical staging. Annals of Surgery 1978;187:1-7.
Boggs TR, Westphal MC. Mortality of exchange transfusion. Pediatrics 1960;26:745-55.
Bollmann R, Schilling H, Ihle W, Mockel A, Zienert A. Intrauterine exchange transfusion in Rh immunization. Zentralblatt fur Gynakologie 1988;110:54-9.
Bowman, JM. RhD hemolytic disease of the newborn. New England Journal of Medicine 1998;339:1775-7.
Ceccon ME, Diniz EM, Ramos JL, Vaz FA. Exchange transfusion in newborn infants with perinatal hemolytic disease. Efficacy of the procedure. Revista Paulista de Medicina 1993;111:348-53.
Cockington R. A guide to the use of phototherpay in the management of neonatal hyperbilirubinemia. Journal of Pediatrics 1979;95:281-5.
Comley A, Wood B. Albumin administration in exchange transfusion for hyperbilirubinaemia. Archives of Diseases in Childhood 1968;43:151-4.
Dennery PA, Seidman DS, Stevenson DK. Neonatal hyperbilirubinemia. New England Journal of Medicine 2001;344:581-90.
Diamond LK. Replacement transfusion as a treatment for erythroblastosis fetalis. Pediatrics 1948;2:520-4.
Flores J, Yanez P, Ramirez R. Portal vein thrombosis in children. Presentation, treatment and outcome. Journal of Pediatric Gastroenterology & Nutrition 1997;25:481.
Fok TF, So LY, Leung KW, Wong W, Feng CS, Tsang SS. Use of peripheral vessels for exchange transfusion. Archives of Diseases in Childhood 1990;65:676-8.
Forfar JO, Keay AJ, Elliot WD, Cumming RA. Exchange transfusion in neonatal hyperbilirubinemia. Lancet 1958;2:1131-7.
Funato M, Shimada S, Tamai H, Taki H, Yoshioka Y. Automated exchange transfusion and exchange rate. Acta Paediatrica Japonica 1989;31:572-7.
Grannum PA, Copel JA, Moya FR, Scioscia AL, Robert JA, Winn HN et al. The reversal of hydrops fetalis by intravascular intrauterine transfusion in severe isoimmune fetal anemia. American Journal of Obstetrics and Gynaecology 1988;158:914-9.
Hoontrakoon S, Suputtamongkol Y. Exchange transfusion as an adjunct to the treatment of severe falciparum malaria. Tropical Medicine & International Health 1998;3:156-61.
Kanto WP Jr, Marino B, Godwin AS, Bunyapen C. ABO hemolytic disease: a comparative study of clinical severity and delayed anemia. Pediatrics 1978;62:365-9.
Lathe GH. Exchange transfusion as a means of removing bilirubin in haemolytic disease of newborn. British Medical Journal 1955;22:192-6.
Luban NL. Hemolytic disease of the newborn: progenitor cells and late effects. New England Journal of Medicine 1998;338:830-1.
Lucas A, Abbott R. Neonatal necrotising enterocolitis. BPSU Annual Report 1996/7 1998;11.
Lucey JF. Diagnosis and treatment of the fetus with erythroblastosis. Pediatric Clinics of North America 1966;13:1117-30.
Lucey JF. Colonic perforation after exchange transfusion. New England Journal of Medicine 1969;280:724.
Maisels MJ. Is exchange transfusion for hyperbilirubinemia in danger of becoming extinct? Pediatric Research 1999;45:210.
Maisels MJ, Newman TB. Bilirubin and neurological dysfunction—do we need to change what we are doing? Pediatric Research 2001;50:677-8.
Merchant RH, Sakhalkar VS, Rajadhyaksha SB. Exchange transfusions via peripheral vessels. Indian Pediatrics 1992;29:457-60.
Mills JF, Woodgate PG. Exchange transfusion for neonatal jaundice (Cochrane Protocol). In: The Cochrane Database of Systematic Reviews, Issue 3, 2003.
Schober PH. Automated exchange transfusion in premature and newborn infants with hyperbilirubinemia using a peripheral arteriovenous vascular access device. Wiener Klinische Wochenschrift 1990;102:471-5.
Shapiro SM. Definition of the clinical spectrum of kernicterus and bilirubin-induced neurologic dysfunction (BIND). Journal of Perinatology 2005;25:54-9.
Soorani-Lunsing I, Woltil HA, Hadders-Algra M. Are moderate degrees of hyperbilirubinemia in healthy term neonates really safe for the brain? Pediatric Research 2001;50:701-5.
Sproul A, Smith L. Bilirubin equilibration during exchange transfusion in hemolytic disease of newborn. Journal of Pediatrics 1961;65:12-26.
Stockman JA 3rd. Overview of the state of the art of Rh disease: history, current clinical management, and recent progress. Journal of Pediatric Hematology and Oncology 2001;23:554-62.
Valaes T. Bilirubin distribution and dynamics of bilirubin removed by exchange transfusion. Acta Paediatrica Scandinavia 1963;49:87-92.
Wennberg AP, Depp R, Heinrichs WL. Indications of early exchange transfusion in infants with erythroblastosis fetalis. Journal of Pediatrics 1978;92:789-92.
Yadav S, Dutta AK, Sarin SK. Do umbilical vein catheterization and sepsis lead to portal vein thrombosis? A prospective, clinical, and sonographic evaluation. Journal of Pediatric Gastroenterology and Nutrition 1993;17:392-6.
| Comparison or outcome | Studies | Participants | Statistical method | Effect size |
|---|---|---|---|---|
| 01 Single volume vs double volume exchange transfusion for ABO hemolytic disease | ||||
| 01 Immediate post exchange serum bilirubin level (micromol/L) | 1 | 20 | WMD (fixed), 95% CI | -13.00 [-45.71, 19.71] |
| 02 Hemoglobin 10 days post exchange (g/dl) | 1 | 20 | WMD (fixed), 95% CI | 0.20 [-0.81, 1.21] |
| 03 Platelet count 10 days post exchange (n X 1000,000,000/L) | 1 | 20 | WMD (fixed), 95% CI | 24.00 [-52.66, 100.66] |
01.01 Immediate post exchange serum bilirubin level (micromol/L)
01.02 Hemoglobin 10 days post exchange (g/dl)
01.03 Platelet count 10 days post exchange (n X 1000,000,000/L)
| This review is published as a Cochrane
review in The Cochrane Library, Issue 4, 2006 (see http://www.thecochranelibrary.com
for information). Cochrane reviews are regularly updated as new
evidence emerges and in response to feedback. The Cochrane Library
should be consulted for the most recent version of the
review. |
