http://informahealthcare.com/jmf ISSN: 1476-7058 (print), 1476-4954 (electronic) J Matern Fetal Neonatal Med, Early Online: 1–5 ! 2015 Informa UK Ltd. DOI: 10.3109/14767058.2015.1045864
ORIGINAL ARTICLE
Fully automated simultaneous umbilical arteriovenous exchange transfusion in term and late preterm infants with neonatal hyperbilirubinemia J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
Hu¨seyin Altunhan, Ali Annagu¨r, Nuriye Tarakc¸i, Murat Konak, Sabahattin Ertug˘rul, and Rahmi O¨rs Faculty of Medicine, Selcuk University, Neonatology, Konya, Turkey
Abstract
Keywords
Objectives: The purpose of this study was to compare the efficacy and safety of two different catheterization techniques of exchange transfusion (ET) used in the therapy of newborn jaundice: fully automated two-way ET technique and the classical one-way ET. Patients and methods: The study included babies at gestational age of 434 weeks. In total, 107 ETs were performed on 86 babies. Totally, the umbilical vein (UV) group included 54 babies having undergone 69 ETs and the UV/UA group included 32 babies having undergone 38 ETs. Results: The declines in bilirubin levels right after ET (p ¼ 0.018) and 8 h after ET (p ¼ 0.014) were higher in the fully automated UV/UA technique than in the classical UV technique. Furthermore, the duration of intensive phototherapy following ET was shorter in the UV/UA method than in the UV method (p ¼ 0.003). There was no difference between the two methods in terms of ETassociated complications (p ¼ 0.927). Conclusions: In neonatal hyperbilirubinemia, ET with fully automated UV/UA technique is more efficient than the classical ET technique, causing no additional side-effects. It is also more physiological than the classical technique, since it minimizes the fluctuations in the blood volume and intravascular pressure during ET.
Automated exchange transfusion, exchange transfusion, neonatal hyperbilirubinemia, umbilical artery, umbilical vein
Introduction The prevalence of exchange transfusion (ET) in the therapy of neonatal jaundice has markedly declined particularly in developed countries due to the wide use of intensive phototherapy and protective measures taken [1]. However, ET is still a frequently used therapeutic approach in countries where neonatal jaundice is prevalent [2]. The standard technique in ET is umbilical vein (UV) catheterization where double volume blood exchange is performed. In this method, each time a volume of blood of less than 5–10 mL/kg is removed and transfused. However, the simultaneous use of UV and umbilical artery (UA), that is, the automated UV/UA (two-way) technique performed with infusion pumps, is expected to be more efficient than the classical technique, which has the following disadvantages: insufficient ET experience of pediatricians in centers where ET is rarely performed, long procedure leading to loss of manpower, no phototherapy during ET, some changes in hemodynamics, infection risk due to the technique, a lower decline in the bilirubin level than expected due to the dead space in the set and injectors during each blood drawing and Address for correspondence: Dr. Ali Annagu¨r, Faculty of Medicine, Selcuk University, Neonatology, Konya, Turkey. E-mail: aliannagur @yahoo.com
History Received 29 January 2015 Revised 24 April 2015 Accepted 24 April 2015 Published online 1 June 2015
transfusing [3–9]. In this study, we compared the classical one-way ET where only the UV was used with the fully automated two-way ET technique where UV and UA were simultaneously used.
Patients and methods This retrospective investigation was carried out by studying the files of newborns with neonatal jaundice who had undergone ET with classical or fully automated UV/UA technique in the Neonatal Intensive Care Unit of Meram Medical School, Konya, Turkey, between February 2008 and October 2014. The study included babies at gestational age of 434 weeks (term and late preterm infants). Babies with gestational age under 34 weeks were excluded from the study because they may have had various problems that could complicate the clinical picture [10]. The babies’ demographic characteristics, indications for ET, bilirubin levels and ET-associated side-effects were noted. Babies undergoing ET for reasons other than newborn jaundice (such as polycythemia, anemia) were not included in the study. The possible etiologies of hyperbilirubinemia with one or more causes have been categorized as follows: Rh incompatibility (Rh-positive baby from an Rh-negative mother), ABO incompatibility (baby with blood group A or B from
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
2
H. Altunhan et al.
mother with blood group O), glucose-6-phosphate dehydrogenase (G6PD) deficiency, polycythemia, hereditary spherocytosis, infections (sepsis, urinary infection, etc.), breast milk jaundice, extravascular hemorrhage (such as cephalhematoma, diffuse ecchymosis) and idiopathic hyperbilirubinemia. Indications for ET were established according to the protocols set forth by our clinic. All catheterizations and ET procedures were carried out under aseptic conditions by the fellow or senior pediatrics assistant in the Newborn Unit. The localizations of the catheters were radiologically checked before ET. The catheters were withdrawn when ET was considered sufficient. In both methods of ET, as donor blood, Rh- and ABO-compatible erythrocyte suspension mixed with AB type plasma was used [11,12]. All babies underwent isovolumetric double volume ET. Until 2011, in our clinic, we performed classical ET on babies in whom UV catheter could be placed. After 2011, we have started to perform automated two-way ET with infusion pumps on babies in whom both UV and UA catheters could be placed. We have used the classical ET in babies in whom UV catheter can be placed, but not the UA catheter. In classical ET, each time, 5 mL/kg of blood was drawn from UV and same volume of blood was administered, completing a cycle. For each 100 mL of blood, 0.1 g of calcium gluconate infusion was administered. The duration of ET with the fully automated method was 90 min. As blood was withdrawn from UA with the infusion pump, the donor blood was simultaneously given to UV at the same rate with the infusion pump. Furthermore, at the same time, the calcium gluconate necessary (for each 100 mL of blood, 0.1 g calcium gluconate) during the whole procedure (90 min) was infused. Thus, completely closed and automated ET was completed in 90 min. Before and after ET, all babies received intensive phototherapy. Babies who had fully automated ET also received intensive phototherapy during the ET procedure. During ET and in the 24 h after ET, the oxygen saturation, pulse, respiration, skin temperature and blood pressure of each baby were closely monitored. Moreover, serum bilirubin, sodium, potassium, chlorine, calcium and albumin levels and the complete blood count were assessed before and after ET. The serum bilirubin level and complete blood count were determined every 8 h (if needed, more frequently) until bilirubin attained reasonable levels. Any abnormality formerly absent and seen within the first week after ET was accepted as ET-associated side-effect. The main ET-associated side-effects described are thrombocytopenia (platelet count, 5100 000/mL), hypoglycemia (glucose, 550 mg/dL), hypocalcemia (in asymptomatic babies, 57 mg/ dL; in symptomatic babies, 58 mg/dL), necrotizing enterocolitis (NEC) (according to modified Bell’s staging criteria), sepsis (clinical sepsis and/or blood culture positive), omphalitis (infection in the umbilicus), respiratory distress, hypotension, hypertension, convulsions and renal insufficiency. All data were analyzed using the Statistical Package for the Social Sciences version 15.0 (SPSS Inc., Chicago, IL). Statistical analysis was performed using the Mann–Whitney U test (Wilcoxon rank-sum test) for continuous variables. The Pearson 2 and the Fisher’s exact tests were used for
J Matern Fetal Neonatal Med, Early Online: 1–5
categorical variables, as appropriate. Significance was defined as a p value of 50.05.
Results Totally, 99 newborns underwent ET in our unit between February 2008 and October 2014. The study included a total of 86 babies who had undergone ET with UV or UV/UA technique and whose hospital files contained all the necessary data. Of the babies, 54 had undergone classical ET (by placing only a UV catheter) and 32 had fully automated ET (by simultaneous installation of UV and UA catheters). Of the newborns, ET was performed only once on 68, twice on 15 and thrice on 3. The total number of ETs performed was 107; the total number of ETs performed in the UV group, which included 54 babies, was 69, and in the UV/UA group, which included 32 babies was 38. As seen in Table 1, the possibility of repeat ET was higher in the UV group than in the UV/UA group, but this difference was statistically insignificant (p ¼ 0.243). There was no statistically significant difference between the two groups in terms of gender, gestational age, birth weight, the day ET was performed and the etiology of the jaundice (p40.05) (Table 1). The most frequent cause of neonatal jaundice requiring ET was blood group incompatibilities (Rh, ABO and subgroup incompatibility), followed by idiopathic jaundice, breast milk feeding jaundice, hereditary spherocytosis, G6PD deficiency, infections and extravascular hemorrhage. Among 54 babies who had undergone classical ET, five babies were first chosen for fully automated ET placing UV and UA catheters simultaneously, but due to the difficulty in placing a UA catheter, these babies had to be treated with classical ET. Totally, four babies died within 3 days after ET. Two of these babies were in the UV group and two in the UV/ UA group. But the cause of death in all these four babies was Table 1. Demographic characteristics of the included patients. Parameter ET cases, n Frequency of ET procedures, n Gestational age, wk Mean ± SD Birth weight, g Mean ± SD Gender, n (%) Female Male Age at ET, d Mean ± SD Etiology, n (%) Blood group incompatibility Unknown Breast milk feeding Hereditary spherocytosis G6PD deficiency Infection Extravascular hemorrhage Other
UV 54y 69z
UV/UA 32y 38z
p* 0.243
37.6 ± 1.5
37.2 ± 1.9
0.392
3130 ± 625
3075 ± 582
0.516 0.985
31 (44.9) 38 (55.1)
17 (44.7) 21 (55.3)
3.9 ± 2.5
4.1 ± 2.3
0.218
20 15 6 5 3 2 1 2
12 8 4 3 2 1 1 1
0.966 0.778 0.846 0.986 0.894 0.888 0.705 0.888
(37) (27.8) (11) (9.3) (5.6) (3.7) (1.9) (3.7)
(37.5) (25) (12.5) (9.4) (6.3) (3.1) (3.1) (3.1)
UV, umbilical vein; UV/UA, umbilical vein/umbilical artery; ET, exchange transfusion; G6PD, glucose-6-phosphate dehydrogenase. *p values were by 2 test, Fisher’s exact test, or Mann–Whitney U test when appropriate. yNumber of neonates. zNumber of ETs.
Fully automated exchange
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
DOI: 10.3109/14767058.2015.1045864
not ET-associated complications. The cause of death in two neonates was severe hyperbilirubinemia and hydrops due to blood incompatibility, in one baby, it was the formerly present sepsis, and in one baby, it was the formerly present severe respiratory insufficiency. There was no significant difference in death between the two groups (p ¼ 0.537). Out of 86 babies, four developed kernicterus; two of the babies were in the UV group and two in the UV/UA group. There were findings of acute bilirubin encephalopathy in these four babies prior to ET. Of these four, the first baby had a total bilirubin level of 36.2 mg/dL and sepsis. The second baby was born in another medical center and was 19-h-old when referred to our clinic with a total bilirubin level of 29.6 mg/ dL. The third baby was a 49-h-old baby transferred to our clinic because of hyperbilirubinemia due to ABO incompatibility; on arrival, the baby’s total bilirubin level was 41.8 mg/ dL. The fourth baby was 31-h-old and had a total bilirubin level was 39.7 mg/dL; the etiology of jaundice could not be determined. Out of thrombocytopenia, no serious ET-associated complications were observed in our cases (such as cardiac arrest, sepsis, ET-procedure associated death). Thrombocytopenia requiring treatment was determined in three babies (platelet count 550 000). Of these three babies, two were in the UV group and one in the UV/UA group. In these three babies, the platelet count before ET was under 100 000. The first baby had a serious hyperbilirubinemia due to Rh incompatibility with a platelet count of 89 000 before ET. After ET, the platelet count decreased to 26 000, which required platelet transfusion. The second baby had sepsis before ET and a platelet count of 77 000. After ET, the platelet count dropped to 19 000, which required platelet transfusion. The third baby had a platelet count of 81 000 before ET. The baby had hydrops due to Rh incompatibility, and after ET, the platelet count dropped to 22 000 requiring platelet transfusion. However, there was no significant difference between the two groups in terms of thrombocytopenia and other complications (p40.05). All complications occurring within 3 days after ET disappeared without causing any sequelae. When the total serum bilirubin levels determined prior to ET were compared, there was no statistically significant difference between the two groups (p ¼ 0.537). When the total serum bilirubin levels determined after ET were compared, the decline in the UV/UA group was more than that in the UV group (p ¼ 0.018) (Table 2). As another parameter, when the bilirubin levels 8 h after ET (rebound bilirubin) (in babies undergoing more than one ET, if the next ET were performed earlier than 8 h, right before this ET), the bilirubin level was significantly lower in the UV/UA group than in the classical UV group (p ¼ 0.014). As an indicator of ET efficiency, the durations of phototherapy after ET were compared. The duration of phototherapy after ET in babies in the UV/UA group was significantly shorter than that in the classical UV group (p ¼ 0.003). In the comparison of duration of phototherapy in both the groups, the four babies who had died after ET were not included in the evaluation. There was no significant difference in hematocrit levels determined before ET between the two groups (p ¼ 0.827). The hematocrit level determined after ET in the UV/UA group was higher than that
3
in the classical UV group, but this difference was statistically insignificant (p ¼ 0.085).
Discussion Modification efforts have been made for further development of the classical ET using UV [3–5,8,9,14–16], which was first introduced by Diamond et al. [13]. One of these modifications is the automated ET where infusion pumps are used instead of the classical push–pull technique, and the other modification is the simultaneous use of UV and UA or peripheral arteries and veins. In our study, we compared the fully automated twoway ET technique performed with infusion pumps, where the UV and the UA were simultaneously used (UA/UV) and the classical one-way ET, where only the UV was used. The negative aspects of classical ET with push–pull technique using only UV have been a matter of discussion for long. One of these negative aspects, maybe the most important, is the potential hemodynamic impairment [5,6,14,17,18]. During the classical ET, 5–10 mL/kg blood is pulled-out and pushed-in in a very short time of about 30– 40 times. This pull and push procedure leads to sudden and serious fluctuations in the baby’s blood pressure, particularly to hypovolemia and decrease in cardiac output during removal of blood. These fluctuations in blood pressure may affect the cerebral blood flow and thus cause intraventricular hemorrhage particularly in premature and/or sick neonates [5,6,17,18]. Likewise, this hemodynamic impairment may Table 2. Biochemical and hematological characteristics of the patients and adverse events of ET in the two groups.
Parameter
UV (n ¼ 69*)
UV/UA (n ¼ 38*)
Pre-ET bilirubin (mg/dL) Mean ± SD 27.6 ± 6.1 29.2 ± 6.6 Post-ET bilirubin (mg/dL) Mean ± SD 13.6 ± 3.4 11.2 ± 3.1 Bilirubin levels after 8 h following the ET (mg/dL) Mean ± SD 14.3 ± 3.5 11.8 ± 3.2 Post ET duration of phototherapy (h) Mean ± SD 42 ± 16 29 ± 11 Pre-ET hematocrit (%) Mean ± SD 41.5 ± 7.2 40.6 ± 6.8 Post-ET hematocrit (%) Mean ± SD 37.6 ± 5.9 39.5 ± 6.1 Adverse events of ET, n (%) Total events 46 (5.1) 25 (5.0) Thrombocytopenia 18 (26) 11 (29) Death 2 (2.9) 2 (5.3) Hypocalcemia 6 (8.7) 4 (10.5) Hypoglycemia 1 (1.4) 0 (0.0) Hypercalcemia 2 (2.9) 1 (2.6) Apnea 3 (4.3) 1 (2.6) Bradycardia 3 (4.3) 1 (2.6) Hyperkalemia 1 (1.4) 0 (0.0) Infection 2 (2.9) 0 (0.0) Cyanosis 4 (5.8) 2 (5.3) Hypotension 2 (2.9) 0 (0.0) Hyponatremia 1 (1.4) 1 (2.6) Omphalitis 1 (1.4) 2 (5.3)
py 0.537 0.018 0.014 0.003 0.827 0.085 0.927 0.75 0.537 0.756 0.999 0.999 0.654 0.654 0.999 0.538 0.999 0.538 0.999 0.287
UV, umbilical vein; UV/UA, umbilical vein/umbilical artery; ET, exchange transfusion. *Number of ETs. yp values were by 2 test, Fisher’s exact test, or Mann–Whitney U test when appropriate.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
4
H. Altunhan et al.
also affect the gastrointestinal system [19]. However, in the fully automatic technique, the pull and push procedure continues with a fixed velocity and so, at least theoretically, it does not significantly change the blood volume and intravascular pressure [5]. There are many studies where peripheral vessels instead of umbilical vessels or umbilical vessels with peripheral vessel are used for ET [4,5,8,14,15]. Although these studies report that catheterization of peripheral vessels is easier than catheterization of umbilical vessels, in our routine work, we have found umbilical catheterization to be faster and easier. We perform peripheral catheterization only in cases where umbilical catheterization is not possible. Another important parameter in studies comparing the ET techniques is the prevalence of ET-associated unwanted effects. Weng and Chiu [4] have reported that the side-effects of the classical ET technique are less than those of the ET techniques using UV/UA or peripheral veins. Many studies comparing the side-effects of ET methods have reported different results, but similar rates of life-threatening sideeffects [4,8,14,15]. In our study, the unwanted effects in both the groups were found to be similar. There was no significant difference between the two groups in terms of the rates of serious side-effects, namely, infant death and kernicterus. In each group, two babies developed kernicterus and two babies died; the deaths were not associated with ET. Weng and Chiu [4] have reported that there is no significant difference between the UV, UV/UA and femoral vein groups in terms of hypocalcemia, death, apnea and infection, but the unwanted effect of thrombocytopenia and total unwanted effects are observed more often in the UV/UA group. Chen et al. [14] compared ET using the UV technique with the ET using peripheral vein/peripheral artery, and reported that the total side-effect rate was 4.7% in the UV group, but the difference in the side-effect rate between the two groups was insignificant. Jackson [10] reported that the thrombocytopenia rate in classical ET patients was 16% and the mortality rate in formerly sick neonates was 18%. In our study, there was no significant difference between the two groups in terms of the total unwanted effects as well as thrombocytopenia, hypocalcemia, apnea and hypoglycemia. In the classical ET using only UV, the dead space in the injector and the catheter is inevitable, and blood in the dead space never takes part in ET. On the other hand, in the doubleway UV/UA method, the whole of the donor blood enters the baby’s circulation, leading to complete exchange [3,5,8,9,15]. The elimination of the dead space is particularly more important in neonates for providing an efficient ET. With the UV/UA technique performed with infusion pumps, the aimed duration and volume in ET is more readily reached. Accordingly, we showed that the fully automated UV/UA technique was more efficient than the classical UV technique. Another parameter compared was the number of ETs required. We found that with the fully automated UV/UA technique, the number of repeat ETs required was fewer than that of the classical UV technique, but the difference was statistically insignificant (p ¼ 0.243). When the falls in total serum bilirubin were compared, we found that the falls in bilirubin levels right after ET and 8 h after ET (rebound) were
J Matern Fetal Neonatal Med, Early Online: 1–5
higher in the UV/UA group than those in the classical UV group, and these falls in total serum bilirubin levels were statistically significant (p ¼ 0.018 and p ¼ 0.014, respectively). Furthermore, the duration of intensive phototherapy after FT was significantly shorter in the UV/UA group than in the classical UV group (p ¼ 0.003). The shorter intensive phototherapy will surely shorten the baby’s stay in the hospital. Another parameter compared was the hematocrit level after ET. In the fully automated UV/UA group, the hematocrit levels were higher than those in the classical UV group, but the difference was statistically insignificant (p ¼ 0.085). Among relevant studies in the literature, Chen et al. [14] and Weng and Chiu [4] found no difference in efficiency between the classical pull–push method and fully automated ET, but Goldman and Tu [5], Schober [15], Funato et al. [8] and some other authors [3,9] reported that the fully automated ET was more effective in decreasing the serum bilirubin levels. The advantages of ET using the fully automated UV/UA technique can be summarized as follows: (1) Since it is an automated and closed system, it is quite sterile and minimizes air escapes; (2) It provides an ideal possibility of reaching the aimed duration and volume during ET; (3) Since it is automated, it reduces the work load of the staff, it saves on personnel and time, and is less tiring; (4) It enables phototherapy of the baby during ET; (5) There is no need to immobilize the baby during ET; since there is no pull–push technique involved, the discomfort of the baby, such as frequent crying, is markedly reduced; (6) The elimination of dead space increases the efficacy of ET; it shortens the baby’s stay in the hospital; (7) It provides a simultaneous and isovolumic ET, which reduces the marked fluctuations in cardiovascular balance and thus, the impairments in cerebral blood flow that consequently lead to decrease in ETassociated side-effects particularly in premature and/or very sick babies. Although automated ET has been known for long years, it has not been widely used [5,8,9,15,16]. However, today, with the development of advanced and sensitive infusion pumps, the automated method can be used more efficiently. This study has some limitations. The most important limitation is that it was not a randomized controlled study. The second limitation is that premature babies 534-week gestational age were not included in the study. The third limitation is that the cases studied were of limited number. The fourth limitation is that there was no comparison with ET using peripheral vessels. In view of these limitations, we can say that future randomized controlled studies including younger premature babies and a larger patient population for comparing the techniques mentioned above, will guide us further on the subject. In conclusion, the fully automated ET performed with infusion pumps using umbilical vein and artery simultaneously, when compared with the classical ET using UV only, is more efficient in reducing the serum bilirubin level and does not cause an increase in ET-associated side-effects. The automated ET reduces the risk of infection, since it is a completely closed and automated system. The system reduces the workload of the staff. Lastly, the automated method can
DOI: 10.3109/14767058.2015.1045864
provide a more physiological ET by minimizing the changes in blood volume and intravascular pressure during ET.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
References 1. Steiner LA, Bizzarro MJ, Ehrenkranz RA, Gallagher PG. A decline in the frequency of neonatal exchange transfusions and its effect on exchange-related morbidity and mortality. Pediatrics 2007;120: 27–32. 2. Sanpavat S. Exchange transfusion and its morbidity in ten-year period at King Chulalongkorn Hospital. J Med Assoc Thai 2005;88: 588–92. 3. Ata M, Holman CA. Simultaneous umbilical arteriovenous exchange transfusion. Br Med J 1966;2:743–5. 4. Weng YH, Chiu YW. Comparison of efficacy and safety of exchange transfusion through different catheterizations: femoral vein versus umbilical vein versus umbilical artery/vein. Pediatr Crit Care Med 2011;12:61–4. 5. Goldman SL, Tu HC. Automated method for exchange transfusion: a new modification. J Pediatr 1983;102:119–21. 6. Bada HS, Chua C, Salmon JH, Hajjar W. Changes in intracranial pressure during exchange transfusion. J Pediatr 1979;94:129–32. 7. Lou HC, Lassen NA, Friis-Hansen B. Impaired autoregulation of cerebral blood flow in the distressed newborn infant. J Pediatr 1979;94:118–21. 8. Funato M, Shimada S, Tamai H, et al. Automated exchange transfusion and exchange rate. Acta Paediatr Jpn 1989;31:572–7.
Fully automated exchange
5
9. Philpott MG, Banerjee A. Automated method for exchange transfusion. Arch Dis Child 1972;47:815–18. 10. Jackson JC. Adverse events associated with exchange transfusion in healthy and ill newborns. Pediatrics 1997;99:E7. 11. Sharma DC, Rai S, Mehra A, et al. Study of 25 cases of exchange transfusion by reconstituted blood in hemolytic disease of newborn. Asian J Transfus Sci 2007;1:56–8. 12. Yigit S, Gursoy T, Kanra T, et al. Whole blood versus red cells and plasma for exchange transfusion in ABO haemolytic disease. Transfus Med 2005;15:313–18. 13. Diamond LK, Allen Jr FH, Thomas Jr WO. Erythroblastosis fetalis. VII. Treatment with exchange transfusion. N Engl J Med 1951;244: 39–49. 14. Chen HN, Lee ML, Tsao LY. Exchange transfusion using peripheral vessels is safe and effective in newborn infants. Pediatrics 2008; 122:e905–10. 15. Schober PH. Automated exchange transfusion in premature and newborn infants with hyperbilirubinemia using a peripheral arteriovenous vascular access device. Wien Klin Wochenschr 1990;102:471–5. 16. Chen HW, Huang WM. Automated peripheral arteriovenous exchange transfusion for treatment of severe hyperbilirubinemia in neonates. Nan Fang Yi Ke Da Xue Xue Bao (J South Med Univ) 2010;30:2396–8. 17. Murakami Y, Yamashita Y, Nishimi T, et al. Changes of cerebral hemodynamics and oxygenation in unstable septic newborns during exchange transfusion. Kurume Med J 1998;45:321–5. 18. van de Bor M, Benders MJ, Dorrepaal CA, et al. Cerebral blood volume changes during exchange transfusions in infants born at or near term. J Pediatr 1994;125:617–21. 19. Touloukian RJ, Kadar A, Spencer RP. The gastrointestinal complications of neonatal umbilical venous exchange transfusion: a clinical and experimental study. Pediatrics 1973;51:36–43.
ORIGINAL ARTICLE
Fully automated simultaneous umbilical arteriovenous exchange transfusion in term and late preterm infants with neonatal hyperbilirubinemia J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
Hu¨seyin Altunhan, Ali Annagu¨r, Nuriye Tarakc¸i, Murat Konak, Sabahattin Ertug˘rul, and Rahmi O¨rs Faculty of Medicine, Selcuk University, Neonatology, Konya, Turkey
Abstract
Keywords
Objectives: The purpose of this study was to compare the efficacy and safety of two different catheterization techniques of exchange transfusion (ET) used in the therapy of newborn jaundice: fully automated two-way ET technique and the classical one-way ET. Patients and methods: The study included babies at gestational age of 434 weeks. In total, 107 ETs were performed on 86 babies. Totally, the umbilical vein (UV) group included 54 babies having undergone 69 ETs and the UV/UA group included 32 babies having undergone 38 ETs. Results: The declines in bilirubin levels right after ET (p ¼ 0.018) and 8 h after ET (p ¼ 0.014) were higher in the fully automated UV/UA technique than in the classical UV technique. Furthermore, the duration of intensive phototherapy following ET was shorter in the UV/UA method than in the UV method (p ¼ 0.003). There was no difference between the two methods in terms of ETassociated complications (p ¼ 0.927). Conclusions: In neonatal hyperbilirubinemia, ET with fully automated UV/UA technique is more efficient than the classical ET technique, causing no additional side-effects. It is also more physiological than the classical technique, since it minimizes the fluctuations in the blood volume and intravascular pressure during ET.
Automated exchange transfusion, exchange transfusion, neonatal hyperbilirubinemia, umbilical artery, umbilical vein
Introduction The prevalence of exchange transfusion (ET) in the therapy of neonatal jaundice has markedly declined particularly in developed countries due to the wide use of intensive phototherapy and protective measures taken [1]. However, ET is still a frequently used therapeutic approach in countries where neonatal jaundice is prevalent [2]. The standard technique in ET is umbilical vein (UV) catheterization where double volume blood exchange is performed. In this method, each time a volume of blood of less than 5–10 mL/kg is removed and transfused. However, the simultaneous use of UV and umbilical artery (UA), that is, the automated UV/UA (two-way) technique performed with infusion pumps, is expected to be more efficient than the classical technique, which has the following disadvantages: insufficient ET experience of pediatricians in centers where ET is rarely performed, long procedure leading to loss of manpower, no phototherapy during ET, some changes in hemodynamics, infection risk due to the technique, a lower decline in the bilirubin level than expected due to the dead space in the set and injectors during each blood drawing and Address for correspondence: Dr. Ali Annagu¨r, Faculty of Medicine, Selcuk University, Neonatology, Konya, Turkey. E-mail: aliannagur @yahoo.com
History Received 29 January 2015 Revised 24 April 2015 Accepted 24 April 2015 Published online 1 June 2015
transfusing [3–9]. In this study, we compared the classical one-way ET where only the UV was used with the fully automated two-way ET technique where UV and UA were simultaneously used.
Patients and methods This retrospective investigation was carried out by studying the files of newborns with neonatal jaundice who had undergone ET with classical or fully automated UV/UA technique in the Neonatal Intensive Care Unit of Meram Medical School, Konya, Turkey, between February 2008 and October 2014. The study included babies at gestational age of 434 weeks (term and late preterm infants). Babies with gestational age under 34 weeks were excluded from the study because they may have had various problems that could complicate the clinical picture [10]. The babies’ demographic characteristics, indications for ET, bilirubin levels and ET-associated side-effects were noted. Babies undergoing ET for reasons other than newborn jaundice (such as polycythemia, anemia) were not included in the study. The possible etiologies of hyperbilirubinemia with one or more causes have been categorized as follows: Rh incompatibility (Rh-positive baby from an Rh-negative mother), ABO incompatibility (baby with blood group A or B from
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
2
H. Altunhan et al.
mother with blood group O), glucose-6-phosphate dehydrogenase (G6PD) deficiency, polycythemia, hereditary spherocytosis, infections (sepsis, urinary infection, etc.), breast milk jaundice, extravascular hemorrhage (such as cephalhematoma, diffuse ecchymosis) and idiopathic hyperbilirubinemia. Indications for ET were established according to the protocols set forth by our clinic. All catheterizations and ET procedures were carried out under aseptic conditions by the fellow or senior pediatrics assistant in the Newborn Unit. The localizations of the catheters were radiologically checked before ET. The catheters were withdrawn when ET was considered sufficient. In both methods of ET, as donor blood, Rh- and ABO-compatible erythrocyte suspension mixed with AB type plasma was used [11,12]. All babies underwent isovolumetric double volume ET. Until 2011, in our clinic, we performed classical ET on babies in whom UV catheter could be placed. After 2011, we have started to perform automated two-way ET with infusion pumps on babies in whom both UV and UA catheters could be placed. We have used the classical ET in babies in whom UV catheter can be placed, but not the UA catheter. In classical ET, each time, 5 mL/kg of blood was drawn from UV and same volume of blood was administered, completing a cycle. For each 100 mL of blood, 0.1 g of calcium gluconate infusion was administered. The duration of ET with the fully automated method was 90 min. As blood was withdrawn from UA with the infusion pump, the donor blood was simultaneously given to UV at the same rate with the infusion pump. Furthermore, at the same time, the calcium gluconate necessary (for each 100 mL of blood, 0.1 g calcium gluconate) during the whole procedure (90 min) was infused. Thus, completely closed and automated ET was completed in 90 min. Before and after ET, all babies received intensive phototherapy. Babies who had fully automated ET also received intensive phototherapy during the ET procedure. During ET and in the 24 h after ET, the oxygen saturation, pulse, respiration, skin temperature and blood pressure of each baby were closely monitored. Moreover, serum bilirubin, sodium, potassium, chlorine, calcium and albumin levels and the complete blood count were assessed before and after ET. The serum bilirubin level and complete blood count were determined every 8 h (if needed, more frequently) until bilirubin attained reasonable levels. Any abnormality formerly absent and seen within the first week after ET was accepted as ET-associated side-effect. The main ET-associated side-effects described are thrombocytopenia (platelet count, 5100 000/mL), hypoglycemia (glucose, 550 mg/dL), hypocalcemia (in asymptomatic babies, 57 mg/ dL; in symptomatic babies, 58 mg/dL), necrotizing enterocolitis (NEC) (according to modified Bell’s staging criteria), sepsis (clinical sepsis and/or blood culture positive), omphalitis (infection in the umbilicus), respiratory distress, hypotension, hypertension, convulsions and renal insufficiency. All data were analyzed using the Statistical Package for the Social Sciences version 15.0 (SPSS Inc., Chicago, IL). Statistical analysis was performed using the Mann–Whitney U test (Wilcoxon rank-sum test) for continuous variables. The Pearson 2 and the Fisher’s exact tests were used for
J Matern Fetal Neonatal Med, Early Online: 1–5
categorical variables, as appropriate. Significance was defined as a p value of 50.05.
Results Totally, 99 newborns underwent ET in our unit between February 2008 and October 2014. The study included a total of 86 babies who had undergone ET with UV or UV/UA technique and whose hospital files contained all the necessary data. Of the babies, 54 had undergone classical ET (by placing only a UV catheter) and 32 had fully automated ET (by simultaneous installation of UV and UA catheters). Of the newborns, ET was performed only once on 68, twice on 15 and thrice on 3. The total number of ETs performed was 107; the total number of ETs performed in the UV group, which included 54 babies, was 69, and in the UV/UA group, which included 32 babies was 38. As seen in Table 1, the possibility of repeat ET was higher in the UV group than in the UV/UA group, but this difference was statistically insignificant (p ¼ 0.243). There was no statistically significant difference between the two groups in terms of gender, gestational age, birth weight, the day ET was performed and the etiology of the jaundice (p40.05) (Table 1). The most frequent cause of neonatal jaundice requiring ET was blood group incompatibilities (Rh, ABO and subgroup incompatibility), followed by idiopathic jaundice, breast milk feeding jaundice, hereditary spherocytosis, G6PD deficiency, infections and extravascular hemorrhage. Among 54 babies who had undergone classical ET, five babies were first chosen for fully automated ET placing UV and UA catheters simultaneously, but due to the difficulty in placing a UA catheter, these babies had to be treated with classical ET. Totally, four babies died within 3 days after ET. Two of these babies were in the UV group and two in the UV/ UA group. But the cause of death in all these four babies was Table 1. Demographic characteristics of the included patients. Parameter ET cases, n Frequency of ET procedures, n Gestational age, wk Mean ± SD Birth weight, g Mean ± SD Gender, n (%) Female Male Age at ET, d Mean ± SD Etiology, n (%) Blood group incompatibility Unknown Breast milk feeding Hereditary spherocytosis G6PD deficiency Infection Extravascular hemorrhage Other
UV 54y 69z
UV/UA 32y 38z
p* 0.243
37.6 ± 1.5
37.2 ± 1.9
0.392
3130 ± 625
3075 ± 582
0.516 0.985
31 (44.9) 38 (55.1)
17 (44.7) 21 (55.3)
3.9 ± 2.5
4.1 ± 2.3
0.218
20 15 6 5 3 2 1 2
12 8 4 3 2 1 1 1
0.966 0.778 0.846 0.986 0.894 0.888 0.705 0.888
(37) (27.8) (11) (9.3) (5.6) (3.7) (1.9) (3.7)
(37.5) (25) (12.5) (9.4) (6.3) (3.1) (3.1) (3.1)
UV, umbilical vein; UV/UA, umbilical vein/umbilical artery; ET, exchange transfusion; G6PD, glucose-6-phosphate dehydrogenase. *p values were by 2 test, Fisher’s exact test, or Mann–Whitney U test when appropriate. yNumber of neonates. zNumber of ETs.
Fully automated exchange
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
DOI: 10.3109/14767058.2015.1045864
not ET-associated complications. The cause of death in two neonates was severe hyperbilirubinemia and hydrops due to blood incompatibility, in one baby, it was the formerly present sepsis, and in one baby, it was the formerly present severe respiratory insufficiency. There was no significant difference in death between the two groups (p ¼ 0.537). Out of 86 babies, four developed kernicterus; two of the babies were in the UV group and two in the UV/UA group. There were findings of acute bilirubin encephalopathy in these four babies prior to ET. Of these four, the first baby had a total bilirubin level of 36.2 mg/dL and sepsis. The second baby was born in another medical center and was 19-h-old when referred to our clinic with a total bilirubin level of 29.6 mg/ dL. The third baby was a 49-h-old baby transferred to our clinic because of hyperbilirubinemia due to ABO incompatibility; on arrival, the baby’s total bilirubin level was 41.8 mg/ dL. The fourth baby was 31-h-old and had a total bilirubin level was 39.7 mg/dL; the etiology of jaundice could not be determined. Out of thrombocytopenia, no serious ET-associated complications were observed in our cases (such as cardiac arrest, sepsis, ET-procedure associated death). Thrombocytopenia requiring treatment was determined in three babies (platelet count 550 000). Of these three babies, two were in the UV group and one in the UV/UA group. In these three babies, the platelet count before ET was under 100 000. The first baby had a serious hyperbilirubinemia due to Rh incompatibility with a platelet count of 89 000 before ET. After ET, the platelet count decreased to 26 000, which required platelet transfusion. The second baby had sepsis before ET and a platelet count of 77 000. After ET, the platelet count dropped to 19 000, which required platelet transfusion. The third baby had a platelet count of 81 000 before ET. The baby had hydrops due to Rh incompatibility, and after ET, the platelet count dropped to 22 000 requiring platelet transfusion. However, there was no significant difference between the two groups in terms of thrombocytopenia and other complications (p40.05). All complications occurring within 3 days after ET disappeared without causing any sequelae. When the total serum bilirubin levels determined prior to ET were compared, there was no statistically significant difference between the two groups (p ¼ 0.537). When the total serum bilirubin levels determined after ET were compared, the decline in the UV/UA group was more than that in the UV group (p ¼ 0.018) (Table 2). As another parameter, when the bilirubin levels 8 h after ET (rebound bilirubin) (in babies undergoing more than one ET, if the next ET were performed earlier than 8 h, right before this ET), the bilirubin level was significantly lower in the UV/UA group than in the classical UV group (p ¼ 0.014). As an indicator of ET efficiency, the durations of phototherapy after ET were compared. The duration of phototherapy after ET in babies in the UV/UA group was significantly shorter than that in the classical UV group (p ¼ 0.003). In the comparison of duration of phototherapy in both the groups, the four babies who had died after ET were not included in the evaluation. There was no significant difference in hematocrit levels determined before ET between the two groups (p ¼ 0.827). The hematocrit level determined after ET in the UV/UA group was higher than that
3
in the classical UV group, but this difference was statistically insignificant (p ¼ 0.085).
Discussion Modification efforts have been made for further development of the classical ET using UV [3–5,8,9,14–16], which was first introduced by Diamond et al. [13]. One of these modifications is the automated ET where infusion pumps are used instead of the classical push–pull technique, and the other modification is the simultaneous use of UV and UA or peripheral arteries and veins. In our study, we compared the fully automated twoway ET technique performed with infusion pumps, where the UV and the UA were simultaneously used (UA/UV) and the classical one-way ET, where only the UV was used. The negative aspects of classical ET with push–pull technique using only UV have been a matter of discussion for long. One of these negative aspects, maybe the most important, is the potential hemodynamic impairment [5,6,14,17,18]. During the classical ET, 5–10 mL/kg blood is pulled-out and pushed-in in a very short time of about 30– 40 times. This pull and push procedure leads to sudden and serious fluctuations in the baby’s blood pressure, particularly to hypovolemia and decrease in cardiac output during removal of blood. These fluctuations in blood pressure may affect the cerebral blood flow and thus cause intraventricular hemorrhage particularly in premature and/or sick neonates [5,6,17,18]. Likewise, this hemodynamic impairment may Table 2. Biochemical and hematological characteristics of the patients and adverse events of ET in the two groups.
Parameter
UV (n ¼ 69*)
UV/UA (n ¼ 38*)
Pre-ET bilirubin (mg/dL) Mean ± SD 27.6 ± 6.1 29.2 ± 6.6 Post-ET bilirubin (mg/dL) Mean ± SD 13.6 ± 3.4 11.2 ± 3.1 Bilirubin levels after 8 h following the ET (mg/dL) Mean ± SD 14.3 ± 3.5 11.8 ± 3.2 Post ET duration of phototherapy (h) Mean ± SD 42 ± 16 29 ± 11 Pre-ET hematocrit (%) Mean ± SD 41.5 ± 7.2 40.6 ± 6.8 Post-ET hematocrit (%) Mean ± SD 37.6 ± 5.9 39.5 ± 6.1 Adverse events of ET, n (%) Total events 46 (5.1) 25 (5.0) Thrombocytopenia 18 (26) 11 (29) Death 2 (2.9) 2 (5.3) Hypocalcemia 6 (8.7) 4 (10.5) Hypoglycemia 1 (1.4) 0 (0.0) Hypercalcemia 2 (2.9) 1 (2.6) Apnea 3 (4.3) 1 (2.6) Bradycardia 3 (4.3) 1 (2.6) Hyperkalemia 1 (1.4) 0 (0.0) Infection 2 (2.9) 0 (0.0) Cyanosis 4 (5.8) 2 (5.3) Hypotension 2 (2.9) 0 (0.0) Hyponatremia 1 (1.4) 1 (2.6) Omphalitis 1 (1.4) 2 (5.3)
py 0.537 0.018 0.014 0.003 0.827 0.085 0.927 0.75 0.537 0.756 0.999 0.999 0.654 0.654 0.999 0.538 0.999 0.538 0.999 0.287
UV, umbilical vein; UV/UA, umbilical vein/umbilical artery; ET, exchange transfusion. *Number of ETs. yp values were by 2 test, Fisher’s exact test, or Mann–Whitney U test when appropriate.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
4
H. Altunhan et al.
also affect the gastrointestinal system [19]. However, in the fully automatic technique, the pull and push procedure continues with a fixed velocity and so, at least theoretically, it does not significantly change the blood volume and intravascular pressure [5]. There are many studies where peripheral vessels instead of umbilical vessels or umbilical vessels with peripheral vessel are used for ET [4,5,8,14,15]. Although these studies report that catheterization of peripheral vessels is easier than catheterization of umbilical vessels, in our routine work, we have found umbilical catheterization to be faster and easier. We perform peripheral catheterization only in cases where umbilical catheterization is not possible. Another important parameter in studies comparing the ET techniques is the prevalence of ET-associated unwanted effects. Weng and Chiu [4] have reported that the side-effects of the classical ET technique are less than those of the ET techniques using UV/UA or peripheral veins. Many studies comparing the side-effects of ET methods have reported different results, but similar rates of life-threatening sideeffects [4,8,14,15]. In our study, the unwanted effects in both the groups were found to be similar. There was no significant difference between the two groups in terms of the rates of serious side-effects, namely, infant death and kernicterus. In each group, two babies developed kernicterus and two babies died; the deaths were not associated with ET. Weng and Chiu [4] have reported that there is no significant difference between the UV, UV/UA and femoral vein groups in terms of hypocalcemia, death, apnea and infection, but the unwanted effect of thrombocytopenia and total unwanted effects are observed more often in the UV/UA group. Chen et al. [14] compared ET using the UV technique with the ET using peripheral vein/peripheral artery, and reported that the total side-effect rate was 4.7% in the UV group, but the difference in the side-effect rate between the two groups was insignificant. Jackson [10] reported that the thrombocytopenia rate in classical ET patients was 16% and the mortality rate in formerly sick neonates was 18%. In our study, there was no significant difference between the two groups in terms of the total unwanted effects as well as thrombocytopenia, hypocalcemia, apnea and hypoglycemia. In the classical ET using only UV, the dead space in the injector and the catheter is inevitable, and blood in the dead space never takes part in ET. On the other hand, in the doubleway UV/UA method, the whole of the donor blood enters the baby’s circulation, leading to complete exchange [3,5,8,9,15]. The elimination of the dead space is particularly more important in neonates for providing an efficient ET. With the UV/UA technique performed with infusion pumps, the aimed duration and volume in ET is more readily reached. Accordingly, we showed that the fully automated UV/UA technique was more efficient than the classical UV technique. Another parameter compared was the number of ETs required. We found that with the fully automated UV/UA technique, the number of repeat ETs required was fewer than that of the classical UV technique, but the difference was statistically insignificant (p ¼ 0.243). When the falls in total serum bilirubin were compared, we found that the falls in bilirubin levels right after ET and 8 h after ET (rebound) were
J Matern Fetal Neonatal Med, Early Online: 1–5
higher in the UV/UA group than those in the classical UV group, and these falls in total serum bilirubin levels were statistically significant (p ¼ 0.018 and p ¼ 0.014, respectively). Furthermore, the duration of intensive phototherapy after FT was significantly shorter in the UV/UA group than in the classical UV group (p ¼ 0.003). The shorter intensive phototherapy will surely shorten the baby’s stay in the hospital. Another parameter compared was the hematocrit level after ET. In the fully automated UV/UA group, the hematocrit levels were higher than those in the classical UV group, but the difference was statistically insignificant (p ¼ 0.085). Among relevant studies in the literature, Chen et al. [14] and Weng and Chiu [4] found no difference in efficiency between the classical pull–push method and fully automated ET, but Goldman and Tu [5], Schober [15], Funato et al. [8] and some other authors [3,9] reported that the fully automated ET was more effective in decreasing the serum bilirubin levels. The advantages of ET using the fully automated UV/UA technique can be summarized as follows: (1) Since it is an automated and closed system, it is quite sterile and minimizes air escapes; (2) It provides an ideal possibility of reaching the aimed duration and volume during ET; (3) Since it is automated, it reduces the work load of the staff, it saves on personnel and time, and is less tiring; (4) It enables phototherapy of the baby during ET; (5) There is no need to immobilize the baby during ET; since there is no pull–push technique involved, the discomfort of the baby, such as frequent crying, is markedly reduced; (6) The elimination of dead space increases the efficacy of ET; it shortens the baby’s stay in the hospital; (7) It provides a simultaneous and isovolumic ET, which reduces the marked fluctuations in cardiovascular balance and thus, the impairments in cerebral blood flow that consequently lead to decrease in ETassociated side-effects particularly in premature and/or very sick babies. Although automated ET has been known for long years, it has not been widely used [5,8,9,15,16]. However, today, with the development of advanced and sensitive infusion pumps, the automated method can be used more efficiently. This study has some limitations. The most important limitation is that it was not a randomized controlled study. The second limitation is that premature babies 534-week gestational age were not included in the study. The third limitation is that the cases studied were of limited number. The fourth limitation is that there was no comparison with ET using peripheral vessels. In view of these limitations, we can say that future randomized controlled studies including younger premature babies and a larger patient population for comparing the techniques mentioned above, will guide us further on the subject. In conclusion, the fully automated ET performed with infusion pumps using umbilical vein and artery simultaneously, when compared with the classical ET using UV only, is more efficient in reducing the serum bilirubin level and does not cause an increase in ET-associated side-effects. The automated ET reduces the risk of infection, since it is a completely closed and automated system. The system reduces the workload of the staff. Lastly, the automated method can
DOI: 10.3109/14767058.2015.1045864
provide a more physiological ET by minimizing the changes in blood volume and intravascular pressure during ET.
Declaration of interest The authors report no conflicts of interest. The authors alone are responsible for the content and writing of this article.
J Matern Fetal Neonatal Med Downloaded from informahealthcare.com by Nyu Medical Center on 06/05/15 For personal use only.
References 1. Steiner LA, Bizzarro MJ, Ehrenkranz RA, Gallagher PG. A decline in the frequency of neonatal exchange transfusions and its effect on exchange-related morbidity and mortality. Pediatrics 2007;120: 27–32. 2. Sanpavat S. Exchange transfusion and its morbidity in ten-year period at King Chulalongkorn Hospital. J Med Assoc Thai 2005;88: 588–92. 3. Ata M, Holman CA. Simultaneous umbilical arteriovenous exchange transfusion. Br Med J 1966;2:743–5. 4. Weng YH, Chiu YW. Comparison of efficacy and safety of exchange transfusion through different catheterizations: femoral vein versus umbilical vein versus umbilical artery/vein. Pediatr Crit Care Med 2011;12:61–4. 5. Goldman SL, Tu HC. Automated method for exchange transfusion: a new modification. J Pediatr 1983;102:119–21. 6. Bada HS, Chua C, Salmon JH, Hajjar W. Changes in intracranial pressure during exchange transfusion. J Pediatr 1979;94:129–32. 7. Lou HC, Lassen NA, Friis-Hansen B. Impaired autoregulation of cerebral blood flow in the distressed newborn infant. J Pediatr 1979;94:118–21. 8. Funato M, Shimada S, Tamai H, et al. Automated exchange transfusion and exchange rate. Acta Paediatr Jpn 1989;31:572–7.
Fully automated exchange
5
9. Philpott MG, Banerjee A. Automated method for exchange transfusion. Arch Dis Child 1972;47:815–18. 10. Jackson JC. Adverse events associated with exchange transfusion in healthy and ill newborns. Pediatrics 1997;99:E7. 11. Sharma DC, Rai S, Mehra A, et al. Study of 25 cases of exchange transfusion by reconstituted blood in hemolytic disease of newborn. Asian J Transfus Sci 2007;1:56–8. 12. Yigit S, Gursoy T, Kanra T, et al. Whole blood versus red cells and plasma for exchange transfusion in ABO haemolytic disease. Transfus Med 2005;15:313–18. 13. Diamond LK, Allen Jr FH, Thomas Jr WO. Erythroblastosis fetalis. VII. Treatment with exchange transfusion. N Engl J Med 1951;244: 39–49. 14. Chen HN, Lee ML, Tsao LY. Exchange transfusion using peripheral vessels is safe and effective in newborn infants. Pediatrics 2008; 122:e905–10. 15. Schober PH. Automated exchange transfusion in premature and newborn infants with hyperbilirubinemia using a peripheral arteriovenous vascular access device. Wien Klin Wochenschr 1990;102:471–5. 16. Chen HW, Huang WM. Automated peripheral arteriovenous exchange transfusion for treatment of severe hyperbilirubinemia in neonates. Nan Fang Yi Ke Da Xue Xue Bao (J South Med Univ) 2010;30:2396–8. 17. Murakami Y, Yamashita Y, Nishimi T, et al. Changes of cerebral hemodynamics and oxygenation in unstable septic newborns during exchange transfusion. Kurume Med J 1998;45:321–5. 18. van de Bor M, Benders MJ, Dorrepaal CA, et al. Cerebral blood volume changes during exchange transfusions in infants born at or near term. J Pediatr 1994;125:617–21. 19. Touloukian RJ, Kadar A, Spencer RP. The gastrointestinal complications of neonatal umbilical venous exchange transfusion: a clinical and experimental study. Pediatrics 1973;51:36–43.
