|Year : 2019 | Volume
| Issue : 3 | Page : 1071-1077
Cord blood albumin–bilirubin as a predictor for neonatal hyperbilirubinemia
Ghada M El Mashad1, Hanan M El Sayed1, Wael A El Shafie2
1 Department of Pediatrics, Faculty of Medicine, Menoufia University, Menoufia, Egypt
2 Department of Pediatrics, Ministry of Health, Menoufia, Egypt
|Date of Submission||15-Jan-2018|
|Date of Acceptance||17-Mar-2018|
|Date of Web Publication||17-Oct-2019|
Wael A El Shafie
Department of Pediatrics, Ministry of Health, Menoufia
Source of Support: None, Conflict of Interest: None
The aim was to evaluate the predictive value of cord blood albumin and bilirubin in identifying infants greater than or equal to 35 weeks for subsequent hyperbilirubinemia.
Hyperbilirubinemia is the most common cause for readmission during the early neonatal period. Early discharge of healthy term newborns after delivery has become a common practice for socioeconomic reasons.
Patients and methods
A prospective cohort study was carried out on 75 neonates, greater than or equal to 35 weeks, with 34 female and 41 male neonates, delivered in Menoufia University hospital during the period from June 2016 to December 2017. Full history taking, clinical examination immediately after birth, and laboratory investigations were recorded.
With cord albumin less than 2.8 g/dl, 81.8% of cases developed significant hyperbilirubinemia, whereas levels greater than 3.3 mg/dl were considered safe with no incidence of hyperbilirubinemia. Moreover, in the total group, the highest sensitivity (83.3%) was for cut-off value of cord bilirubin (CB) at 1.88 mg/dl, with positive predictive value of 72.9%, whereas the highest specificity (84.8%) was for cut-off value of CB–cord albumin at −0.6, with negative predictive value of 74.1. In high-risk group, the highest sensitivity (88.9%), highest positive predictive value (94.1%), highest specificity (85.7%), and negative predictive value (75%), were for cut-off value of CB–cord albumin at − 0.82.
Cord blood albumin and bilirubin examination can be used as an early useful indicator to predict significantly hyperbilirubinemia in neonates. Researches on large scale should be done for more clarification on the predictive value of cord blood albumin and bilirubin for early prediction of significant hyperbilirubinemia.
Keywords: albumin, bilirubin, cord blood, hyperbilirubinemia, neonates
|How to cite this article:|
El Mashad GM, El Sayed HM, El Shafie WA. Cord blood albumin–bilirubin as a predictor for neonatal hyperbilirubinemia. Menoufia Med J 2019;32:1071-7
|How to cite this URL:|
El Mashad GM, El Sayed HM, El Shafie WA. Cord blood albumin–bilirubin as a predictor for neonatal hyperbilirubinemia. Menoufia Med J [serial online] 2019 [cited 2019 Nov 14];32:1071-7. Available from: http://www.mmj.eg.net/text.asp?2019/32/3/1071/268804
| Introduction|| |
Hyperbilirubinemia is also the most common cause for readmission during the early neonatal period . Approximately 50% of term and 80% of preterm newborns develop hyperbilirubinemia, which usually appears 2–4 days after birth, and resolves spontaneously after 1–2 weeks . Because of the increasing number of early discharged newborns, there is a corresponding danger of failing to diagnose severe hyperbilirubinemia in time, as reports about kernicterus in full-term healthy newborns demonstrate . In Egypt, it is common that infants are discharged at less than 24 h of age with little or no evaluation for the risk of developing jaundice or any instructions for follow-up. In fact, clinical or laboratory assessment of jaundice was almost nonexistent even among those who were discharged after the second day of life ,,. There have been a few studies that predicted postnatal hyperbilirubinemia by estimating cord blood bilirubin levels, but they vary in opinions . Early detection of risk factors is the first step toward prevention of hyperbilirubinemia and a step ahead in protecting newborns from complication at later age. It was shown that the dissociated bilirubin concentration can be increased in extracellular fluids independently of total bilirubin concentration by reducing the binding capacity of plasma proteins for bilirubin. This situation occurs clinically (a) in premature babies who have a relative hypoalbuminemia, and (b) in some newborn infants whose plasma contains substances that reduce the bilirubin-binding capacity of albumin. Such substances are hematin, sulfonamides, salicylate, caffeine sodium benzoate, and increased hydrogen ion concentration . There is paucity of reports on cord blood albumin level as predictor of hyperbilirubinemia . The aim of this study was to evaluate the cord blood albumin–bilirubin as a predictor for neonatal hyperbilirubinemia.
| Patients and Methods|| |
A prospective cohort study was conducted on 75 neonates, with 34 females and 41 males, delivered in Menoufia University hospital during the period from June 2016 to December 2017.
The study was approved by the Ethical Committee of Menoufia Faculty of Medicine, and an informed consent was obtained from all participants' guardian before the study commenced.
Selection criteria for the patients
The participants included in this study were selected according to inclusion and exclusion criteria.
The inclusion criteria included neonates with gestational age 35 weeks or more, any mode of delivery, both sexes, APGAR score over 7 at the first minute and 10 at fifth minute of life, absence of significant illness, or a major congenital malformation.
The exclusion criteria were significant illness (sepsis, respiratory distress syndrome, hypoxia, and infant of a diabetic mother) that could aggravate hyperbilirubinemia.
All cases were subjected to the following:
- Full history taking with emphasis on the antenatal and perinatal history (maternal illness, maternal drugs, and fever); risk factors for hyperbilirubinemia, for example, ABO or Rh incompatibility; and individual data from the mothers' medical records, for example, mode of delivery, the cause for lower segment cesarean section, recent and previous obstetric history, single or multiple pregnancies, and past medical history
- Full clinical examination immediately after birth such as general examination (vital signs, anthropometric measures, and presence of cephalohematoma), systemic examination (cardiac, abdominal, chest, and neurological examination), gestational age assessment using Ballard score, and clinical and laboratory follow-up for level of hyperbilirubinemia
- Laboratory investigations: cord blood samples (3 ml) will be collected from all newborns that complied with the protocol inclusion criteria. The samples will be sent for the assaying of total, unconjugated, conjugated bilirubin and albumin levels and baby's blood group and rhesus.
Method of umbilical cord blood collection
After delivery of the newborn, the umbilical cord was double clamped and transected within 10 s. As soon as the newborn was removed from the operative field, cord blood was collected as follows. The placenta-side part of the umbilical cord was held straight at a slight angle downward. After identification of a suitable puncture site, the umbilical cord was cleaned, the umbilical vein was then punctured with sterile syringe, ∼3 ml of blood was withdrawn from umbilical vein and then put in a plain tube to be separated by contribution and then the serum tested for albumin and bilirubin.
Follow-up of babies was done after 72 h of age for serum bilirubin (total and direct) using a peripheral venous sample.
Bilirubin was analyzed by Colorimetric Diazo method using semiautomated analysis (5010), and diazotized sulphanilic acid reaction using Diamond reagent (Diamond Company, Cairo, Egypt).
Albumin was analyzed by semiautomated bromocresol green dye-binding technique using Diamond reagent (Diamond Company).
Babies' blood groups were done manually using both anti-A and anti-B sera against the babies' serum.
We used the following definitions for our study:
- Significant hyperbilirubinemia: serum bilirubin level greater than or equal to 17 mg/dl after 72 h of life
- High-risk group: neonates who have Rh blood group incompatibility, and also newborn infants with maternal–fetal ABO incompatibility are at a greater risk for developing subsequent significant hyperbilirubinemia, and therefore, prediction of probable risk factors, such as the degree of hemolysis, gains importance
- Criteria for diagnosis of ABO and Rh incompatibility: diagnosis is usually made by investigation of a newborn baby who has developed jaundice during the first week of life. Coombs test – after birth of the baby, a direct Coombs test is run to confirm antibodies attached to the infant's red blood cells. This test is run from cord blood. In some cases, the direct coombs will be negative but severe, and even fatal hemolytic disease of newborn (HDN) can occur. Hemoglobin: the infant's hemoglobin should be tested from cord blood. Reticulocyte count: reticulocytes are elevated when the infant is producing more blood to combat anemia. A rise in the retic count can mean that an infant may not need additional transfusions
- Neutrophils: as neutropenia is one of the complications of HDN, the neutrophil count should be checked. Thrombocytes: as thrombocytopenia is one of the complications of HDN, the thrombocyte count should be checked. Bilirubin should be tested from cord blood. Ferritin: because most infants affected by HDN have iron overload, a ferritin must be run before giving the infant any additional iron.
Results were analyzed and tabulated using Microsoft Excel version 2016 (Microsoft Corporation, New York, New York, USA) and SPSS v. 16 (SPSS Inc., Chicago, Illinois, USA). Qualitative data were expressed as number and percentage and quantitative data were expressed as mean ± SD and multiple regressions, and Wallis analysis of variance for comparing categorical data. Receiver operating characteristic (ROC) curve was constructed with area under curve analysis performed to detect best cut-off value of cord blood albumin and bilirubin for detection of positive cases. A value of P was less than 0.05 was indicated statistically significant.
| Results|| |
The current included 54 (72%) full-term and 21 (28%) late preterm, with 41 (54.7%) males and 34 (45.3%) female. Overall, 62.7% of infants were delivered by cesarean section and 28 (37.3%) were delivered by normal vaginal delivery. According to maternal blood group and Rh, the most common was O, with 29 (38.7%) cases, and approximately 66 (88%) cases had Rh positive. Moreover, there were 25 neonates at more risk to develop hyperbilirubinemia: 18 cases with ABO incompatibility and seven cases with Rh incompatibility. Within the ABO group, 12 (66.6%) cases developed significant hyperbilirubinemia, whereas in Rh group, approximately five (71.4%) cases developed significant hyperbilirubinemia. Among this group, males represented approximately 48% of cases, 64% needed phototherapy, whereas 8% needed exchange transfusion [Table 1].
|Table 1: Comparison between nonrisk and high-risk groups regarding demographics, blood groups, and treatment|
Click here to view
Regarding the clinical and laboratory data, the mean weight of the cases was 2.87 ± 0.58 kg. The mean gestational age was 37.26 ± 4.28 weeks. The mean cord blood total bilirubin was 2.13 ± 0.83 mg/dl, and mean cord blood albumin was 2.76 ± 0.71 g/dl. However, the mean total serum bilirubin after 72 h of life was 14.68 ± 5.33 mg/dl. The level of serum bilirubin after 72 h of cases life was grouped into positive cases that developed significant hyperbilirubinemia (42 cases accounts for 56%) and negative cases (33 accounts for 44% of cases). Moreover, the mean cord blood total bilirubin was 2.52 ± 0.80 mg/dl, mean cord blood albumin was 2.69 ± 0.82 g/dl, and the mean total serum bilirubin after 72 h of life was 16.76 ± 4.18 mg/dl [Table 2].
Concerning the relation between level of cord blood albumin and development of significant hyperbilirubinemia, the cases were categorized into three groups: at cord albumin less than 2.8 g/dl, 81.8% (27 cases) of cases developed significant hyperbilirubinemia, with 75.8% of them (25 cases) needing phototherapy and 6.1% (two cases) needing exchange transfusion; at cord albumin 2.8–3.3 g/dl, 46.9% (15 cases) of cases developed significant hyperbilirubinemia requiring phototherapy for all them with no need for exchange transfusion. Moreover, at cord albumin greater than 3.3 g/dl, no cases developed significant hyperbilirubinemia [Table 3].
A cut-off level of cord blood albumin, cord blood total bilirubin, cord blood bilirubin–albumin, and cord bilirubin (CB)/cord albumin ratio at 2.75 g/dl, 1.88 mg/dl, −0.6, and 0.78 as obtained by ROC curve was determined to have 64.3, 83.3, 73.8, and 73.8% sensitivity, respectively; 81.8, 60.6, 84.8, 86.1, and 84.4% specificity, respectively; and 81.8, 72.9, 86.1, and 0.847% positive predictive value (PPV), respectively, in the prediction of occurrence of significant hyperbilirubinemia [Table 4]. ROC curve shows that the area under the curve was 0.805, 0.756, and 0.870 of total area; P value less than 0.001 indicates the usefulness of the test in predicting hyperbilirubinemia [Figure 1].
|Table 4: The sensitivity and specificity of cord blood albumin in detecting significant hyperbilirubinemia and in group at risk for ABO and Rh incompatibility|
Click here to view
|Figure 1: Receiver operating characteristic curve showing cord blood albumin, cord total bilirubin, cord bilirubin–albumin, and cord bilirubin/cord albumin ratio sensitivity and specificity in detecting significant hyperbilirubinemia in studied groups and groups at risk for ABO and Rh incompatibility.|
Click here to view
In group at risk for ABO and Rh incompatibility, a cut-off level of cord blood albumin, cord blood bilirubin, cord blood bilirubin–albumin, and level of CB/cord albumin ratio of 3 mg/dl, 2.48 mg/dl, 0.829, −0.82, and 0.79, respectively, as obtained by ROC curve was determined to have 77.8, 72.2, 88.9, and 83.3% sensitivity, respectively; 71.4, 85.7, 85.7, and 85.7% specificity, respectively; and 87.5, 92.9, 94.1, and 93.8% PPV, respectively, in the prediction of the occurrence of significant hyperbilirubinemia [Table 4]. ROC curve shows that the area under the curve was 0.829, 0.885, and 0.849 of total area, with P values of 0.012, 0.03, and 0.008, indicating the usefulness of the test in predicting hyperbilirubinemia [Figure 1].
| Discussion|| |
In current study, the mean gestational age of our study population was 37.26 ± 4.28 weeks. Their mean birth weight was 2.87 ± 0.58 kg. We found significant hyperbilirubinemia in 42 (56%) cases; 40 (53.3%) infants underwent phototherapy, two (2.7%) infants underwent exchange transfusion, and 33 (44%) needed no intervention at all. We report quite a high incidence of significant hyperbilirubinemia in our study. Our results come in agreement with previous studies in Egypt, with occurrence from 34% up to 54.3% . On the contrary, many other studies have reported a lower incidence of 5–6, 9.5, 11.5, 19.86, 24, and 33.8% ,,. This difference may be attributed to the differences in racial and ethnic groups in different populations studied for jaundice.
It is stated clearly by several authors that male infants are more at risk of developing severe jaundice than their female counterparts ,. These findings do not coincide with our study, where we found no significant difference between the male and female distribution of our cases.
Regarding the levels of cord blood albumin, the current study showed that at cord albumin levels less than 2.8 g/dl, 81.8% (27 cases) of cases developed significant hyperbilirubinemia. At cord albumin 2.8–3.3 g/dl, 46.9% of cases developed significant hyperbilirubinemia, whereas levels greater than 3.3 g/dl were considered safe, with no incidence of hyperbilirubinemia. Our study results are similar to many other studies, where at cord serum albumin levels less than 2.8 g/dl, 58.35% , 82% , and 95%  of newborns developed significant hyperbilirubinemia in all of the aforementioned studies. At levels greater than 3.3 g/dl, no infants developed hyperbilirubinemia. However, in the study by Trivedi et al. , 12.68% developed hyperbilirubinemia at cord serum albumin greater than 3.5 g/dl.
We performed a ROC curve analysis to find a cut-off point level of cord blood albumin for development of significant hyperbilirubinemia. It was 2.75 g/dl, with a low sensitivity of 64.3%, specificity of 81.8%, PPV of 81.8%, and negative predictive value (NPV) of 64.3%. According to our study protocol, we further investigated the level of cord blood total bilirubin in our sample. The mean value was 2.13 ± 0.83 mg/dl. This is near the results found by a previous study which was 2.38 ± 0.5 mg/dl . However, other studies found a higher mean CB as 2.7 ± 0.3 and 3.7 ± 1.07 mg/dl . This higher level may be owing to the higher mean gestational age in their study group of 39.9 weeks versus 37.26 ± 4.2 weeks in our study.
With ROC analysis, a CB level of 1.88 mg/dl was determined to have 83.3% sensitivity and 60.6% specificity. At this critical CB level, the NPV was 74.1% and the PPV was 72.9%. CB could be a useful predictor for subsequent neonatal jaundice. The study conducted by Knupfer et al.  reported a cut-off value of 1.74 mg/dl. However, higher cut-off values ranging between 2 and 2.7 mg/dl were found in other studies. This may be owing to the characteristics of the different studied groups, where the cases were all full-term infants and the cut-off values reported were 2 mg/dl and 2.54 mg/dl ,,. However, those in our study group consisted of late preterm and full-term infants, or having risk factors of ABO or Rh incompatibility as in 2.1 mg/dl with 50% sensitivity, 97% specificity, 2 and 2.6 mg/dl with 50% sensitivity and 97% specificity, PPV 41.1% and NPV 97.9% ,. In the study group of El-Gendy et al. , at cut-off value of 2.73 mg/dl, with 100% sensitivity and 90% specificity, 71.4% of males developed significant hyperbilirubinemia.
Despite most of the studies confirming the usefulness of CB in predicting the development of hyperbilirubinemia, it was concluded from a previous study by Carbonell et al.  that CB level of 2.2 mg/dl showed scant sensitivity (22.2%) but a high NPV (97.4%), which excluded CB as a useful predictor of subsequent neonatal jaundice. Moreover, another study carried out by Rostami and Mehrabi  concluded that assessment of CB level could not help identifying newborns with neonatal hyperbilirubinemia .
In this high-risk group, mean CB was 2.52 ± 0.80 mg/dl. Our results are in agreement with those of Zeitoun et al.  who recorded that the mean CB among high-risk group is significantly higher compared with the group with no risk. However, Bernaldo and Segre  had a mean CB of 1.84 ± 0.45 mg/dl, with no significant difference in the level of unconjugated bilirubin in cord blood between children with or without maternal–fetal blood incompatibility. Moreover, Ipek et al.  recorded a mean CB of 2.05 ± 0.98 mg/dl. The difference between our results, which shows a higher level, and other studies may be owing to higher percentage of ABO and Rh incompatible cases in our study (33.3%) than in others.
By using CB–cord albumin to predict significant hyperbilirubinemia, we found that a cut-off level of cord blood bilirubin–albumin of − 0.82 as obtained by ROC curve was determined to have 88.9% sensitivity, 85.7% specificity, and PPV 94.1% in the prediction of occurrence of significant hyperbilirubinemia in high-risk group, with NPV of 75%. ROC curve shows that the area under the curve was 0.885 of total area; P value of 0.003 indicated the usefulness of the test in predicting hyperbilirubinemia. A cut-off level of CB/cord albumin ratio of 0.79 in our study was determined to have 83.3% sensitivity, 85.7% specificity, and PPV 93.8% in the prediction of occurrence of significant hyperbilirubinemia in high-risk group, with NPV 66.7%. We conclude that the cut-off level of bilirubin/albumin level in both total and high-risk group is similar to each other (0.787 and 0.797, respectively). However, the cut-off value of high-risk group had a higher sensitivity (83.3%), specificity (85.7%), and PPV (93.8%) than that of total group.
We conclude from the results of our study that in the total group, the highest sensitivity (83.3%) was for cut-off value of CB (1.88 mg/dl), with PPV 72.9%, which mean that we can predict 83.3% of patients with the disease (true positives) but 16.7% of cases with the disease go undetected (false negatives), whereas the highest specificity (84.8%) was for cut-off value of CB–cord albumin (−0.6), with NPV 74.1%, which mean that we can correctly report 84.8% of patients without the disease as test negative (true negatives), but 15.2% patients without the disease are incorrectly identified as test positive (false positives). We can use the bilirubin–albumin or bilirubin/albumin ratio to get the highest combination of bilirubin and albumin.
In high-risk group, the highest sensitivity (88.9%), the highest PPV (94.1%), and highest specificity (85.7%), with NPV of 75%, is for cut-off value of CB–cord albumin (−0.82). The highest PPV in both total (86.1%) and high-risk group (94.1%) is for cut-off value of CB–cord albumin, which is −0.6 and −0.82, respectively. Although the ideal (but unrealistic) situation is for a 100% accurate test, a good alternative is to subject patients who are initially positive to a test with high sensitivity/low specificity, and to a second test with low sensitivity/high specificity. In this way, nearly all the false positives may be correctly identified as disease negative.
| Conclusion|| |
Cord blood albumin and bilirubin can be used as an early useful indicator to predict significantly per bilirubinemia in neonates. Researches on large scale should be done for more clarification on the predictive value of cord blood albumin and bilirubin for early prediction of significant hyperbilirubinemia. The value of screening infants for jaundice before hospital discharge is very important and considered as risk assessment to allow safe early discharge for the baby and to prevent readmission.
Financial support and sponsorship
Conflicts of interest
There are no conflicts of interest.
| References|| |
Bernaldo A, Segre C. Bilirubin dosage in cord blood: could it predict neonatal hyperbilirubinemia? Sao Paulo Med 2004; 122
Jardine LA, Woodgate P. Neonatal jaundice. Am Fam Physician 2012; 85
Cakmak A, Calik M, Atas A, Hirfanoglu I, Erel O. Can haptoglobin be an indicator for the early diagnosis of neonatal jaundice. J Clin Lab Anal 2008; 22
Iskander I, Gamaleldin R, Kabbani M. Root causes for late presentation of severe neonatal hyperbilirubinemia in Egypt. East Mediterr Health J 2012; 18
American Academy of Pediatrics (AAP). Subcommittee on hyperbilirubinemia. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004; 114
Rostami N, Mehrabi Y. Identifying the newborn at risk for developing significant hyperbilirubinemia by measuring cord bilirubin levels. J Arab Neonat Forum 2005; 2
Simpson L, Deoarari AK, Paul VK. Cord bilirubin as a predictor of pathological jaundice – a cohort study. Indian Pediatr 2002; 39
Venkatamurthy M, Murali SM, Hemachandra K. Development of neonatal hyperbilirubinemia by increased umbilical cord bilirubin. Int J Health Inform Med Res 2014; 1
Trivedi DJ, Markande DM, Vidya BU, Bhat M, Hegde PR. Umbilical cord blood bilirubin level measurement in predicting the development of significant hyperbilirubinemia. J Int Sci Inn Tech Sec 2013; 2
Satrya R, Effendi SH, Gurnida DA. Correlation between cord blood bilirubin level and incidence of hyperbilirubinemia in term newborns. Paediatr Indones 2009; 49
Zoubir S, Mieth RA, Berrut S, Roth-Kleiner M. Incidence of severe hyperbilirubinaemia in Switzerland: a nationwide population-based prospective study. Arch Dis Child Fetal Neonatal Ed 2011; 96
Sahu S, Abraham R, John J, Mathew AA. Cord blood albumin as a predictor of neonatal jaundice. Int J Biol Med Res 2011; 2
Knupfer M, Pulzer F, Gebauer C, Robel-Tillig E, Vogtmann C. Predictive value of umbilical cord blood bilirubin for postnatal hyperbilirubinaemia. Acta Paediatr 2005; 94
Ipek IO, Bozaykut A, Cagril S, Sezer R. Does cord blood bilirubin level help the physician in the decision of early postnatal discharge? J Matern Fetal Neonatal Med 2012; 25
Carbonell X, Botet F, Figueras J, Riu-God A. Prediction of hyperbilirubinemia in the healthy term newborn. Acta Paediatr 2001; 90
El-Gendy FM, Hassane FM, Khattab AA, El-Lahony DM, Ashour NM. Predictive ability of first-day serum bilirubin and haptoglobin for subsequent significant hyperbilirubinemia in healthy-term and near-term newborn. Menoufia Med J 2013; 26
Zeitoun AA, Elhagrasy HF, Abdelsatar DM. Prediction of hyperbilirubinemia in the newborn. Egy Pediatr Associ Gaz 2013; 61
[Table 1], [Table 2], [Table 3], [Table 4]