Cord Serum Bilirubin Level in Predicting the Development of Significant Hyperbilirubinemia in Newborns with ABO

Address for correspondence: Dr. Sunita Arora, Professor Department of Paediatrics Sri Guru Ram Das Institute of Medical Sciences and Research Sri Amritsar, Amritsar, India E-mail: dr.sunita1@yahoo.com Tel: +91-9814710274 1Dr. Sunita Arora, MBBS.MD, Professor. 2Dr. Shifali, MBBS, MD Resident. Both from the Department of Paediatrics, Sri Guru Ram Das Institute of Medical Sciences and Research, Sri Amritsar, Amritsar, India. Abstract


Introduction
N eonatal hyperbilirubinaemia is common problem which is benign in majority of neonates.In the era of early discharge from the hospital, in view of increase work load and risk of nosocomial infec on, there is increasing number of readmission of these neonates with signifi cant hyperbilirubinaemia 1 .A reasonable strategy would be required to decrease incidence of severe hyperbilirubinaemia and bilirubin encephalopathy while minimizing risk of unintended harm such as maternal anxiety, decreasing breast feeding and unnecessary cost of treatment.Rh iso immune hemoly c disease is becoming nearly nonexistent due to the use of prophylac c an D. Hence Isoimmune haemoly c disease due to ABO incompa bility assumes signifi cance as a cause of signifi cant hyperbilirubinaemia.
Approximately 15% of live births are at increased risk but jaundice develops in only 0.3 to 2.2% 2 .An bodies against A and B an gens are natural an bodies which occur without previous immuniza on.Most of these an bodies are Ig M type which do not cross placenta.However Ig G an bodies to A or B an gen may be present which can cross placenta.Thus ABO iso immune hemoly c disease can be found in fi rst born infants.Low an genicity of A and B factors and the wide distribu on in placenta and other body ssues apart from red cell accounts for rela vely low incidence and milder nature of ABO hemoly c disease 3 .
Presump ve diagnosis is based on the presence of ABO incompa bility, elevated unconjugated serum bilirubin level, weakly to moderately posi ve DCT, spherocytosis in blood smear, increased number of nucleated red blood cells and increased re culocyte count with marked polychromasia 2 .
To target health care resources towards high risk newborns cord bilirubin, 1 st day bilirubin and pre discharge bilirubin values have been used to predict signifi cant hyperbilirubinaemia in healthy term new born to keep a close follow up and plan interven on if needed 4,6,7,8,9 .These studies did not include ABO or RH incompa bility.S.Bilirubin of > 6mg/dl in fi rst 24 hrs was found to be the risk factor for signifi cant hperbilirubinemia in healthy term newborn.Posi ve direct coombs test, high maternal IgG An A and An B, high re culocyte count and a sibling with jaundice are all predictors of signifi cant jaundice in ABO incompa bility 10,11 .Six hour bilirubin levels of 4mg /dl and 6mg/dl are predictors for signifi cant hyperbilirubinaemia and severe haemoly c disease of new born respec vely, as highlighted by S Umit Sarci et al. 12 End dal carbondioxide measurement in direct an globulin test nega ve ABO new born with signifi cant jaundice points to a cause other than iso immuniza on.
This study was conducted to determine the incidence of ABO incompa bility, ABO iso immune disease in new born, to determine cri cal cord serum bilirubin level to predict subsequent signifi cant hyperbilirubinemia and to evaluate correla on of laboratory markers of haemolysis and development of signifi cant hyperbilirubinemia.

Material and Methods
This was a planned prospec ve hospital based follow up study of 15 months dura on on 100 cases conducted in department of paediatrics in collabora on with department of pathology and biochemistry at Sri Guru Ram Das Ins tute of Medical Sciences and Research, Amritsar.A total of 100 healthy newborn with birth weight ≥2500 gm and gesta on ≥37 wks with blood group A,B,AB born to mother with blood group O without simultaneous Rh incompa bility were taken.100 full term healthy newborns with B.W≥2500gm and gesta onal age ≥37 wk with blood group A, B, AB, born to mothers with O blood group without simultaneous Rh incompa bility at SGRDIMSR from 1 st Feb 2012 to June 2013 were included.Hb, re culocyte count, blood group (ABO and Rh), direct an globulin test, peripheral blood fi lm, Serum bilirubin (total and direct) were performed in all cases from cord blood.Subsequently serum bilirubin was measured approximately at 12-24 hrs, 36-48hrs,60-72hrs.The exclusion criteria were; Sick newborns with respiratory distress, asphyxia, sepsis, major congenital anomalies, Rh incompa bility, Cephalhematoma and G6PD defi ciency Guidelines for phototherapy and exchange transfusion were according to recommenda ons of American Acade my of Paediatrics 13 .Serum bilirubin measurement was done by Jendrassic and Grof method, Coombs test was performed using an human globulin reagent, Glucose6 phosphate dehydrogenase es ma on was done by dye decolorisa on method.
Sta s cal analysis was done by so ware SPSS version 16 on comple on of study.
Permission from the ins tu onal ethical commi ee was taken before conduc ong the study study.

Results
Out of total 608 term deliveries, 106 (17.4%) were cases of ABO incompa bility.Out 100 ABO incompa ble newborns 33(33%) developed ABO isoimmune disease manifes ng as signifi cant hyperbilirubinaemia with any of the four total serum bilirubin levels exceeding threshold levels defi ned for phototherapy.
Demographic characteris cs like sex, mode of delivery, Birth weight, type of blood group incompa bility did not seem relevant to development of signifi cant jaundice.O/A incompa bility was more common on the whole (50 Cases) but the diff erence in the two groups was sta s cally insignifi cant.DCT was nega ve in 31 newborn with signifi cant hyperbilirubinaemia and all newborn without signifi cant hyperbilirubinaemia.Two babies with posi ve DCT had signifi cant jaundice, Diff erence in DCT posi vity was sta s cally signifi cant in two group with a p-value 0.042.Diff erence in Hb value was not sta s cally signifi cant in two groups.Diff erence in re culocyte count was signifi cant in two groups.If 35 th percen le is taken as cutoff value, even a single case of signifi cant hyperbilirubinemia will not be missed but large number of newborns will be subjected to unnecessary inves ga on.Thus any serum biliirubin value below 35 th percen le cons tutes low risk group.Values between 35 th and 60 th percen le cons tutes low intermediate risk group.Cord serum bilirubin at or above 60 th percen le have a high probability of developing signifi cant hyperbilirubinemia.As the sensi vity is 100 % no neonate with signifi cant jaundice will be missed.At the same me specifi city is 90% and accuracy is 93%.This cons tutes high intermediate risk group.Cord serum bilirubin levels at or above 90 th percen le has specifi city and posi ve predic ve value of 100% but sensi vity of 30%.Newborns having cord serum bilirubin at or above 90 th cen le will defi nitely develop signifi cant hyperbilirubinemia as specifi city and PPV is 100% but we can miss upto 70% newborns who can develop signifi cant jaundice because sensi vity is only 30%.
Then it can be inferred that any newborn with serum bilirubin at or above 90 th percen le should not be discharged and the one between 60 th and 90 th percen le should be kept on close follow up.TSB of 2.16mg/d1 from cord blood has a sensi vity of 100% specifi city of 89.55%, NPV 100% and PPV of 82.50% to predict signifi cant hyperbilirubinaemia.Thus any newborn with cord serum bilirubin between 2.16 mg/d1 to 4.090 mg/d1 should be kept in close supervision.Newborns with cord serum bilirubin more than 4.09 mg/d1 should not be discharged and S.Bilirubin should be repeated in next 24hrs.

Discussion
Clinical course and severity of subsequent hyperbilirubinaemia or isommune disease is diffi cult to predict in a newborn with ABO incompa bility because there is no single test that is of high predic ve value 15,16 .
Incidence of ABO incompa bility was 17.4% and signifi cant jaundice was observed in 33% of ABO incompa ble newborns in the present study.HM Rusemerg et al reported ABO incompa bility in 20-25% of deliveries and ABO haemoly c disease in less than 10% of these cases 9 .S Umit et al reported ABO incompa bility in 14.18% of deliveries out of which 21.3% had signifi cant hyperbilirubinaemia 12 .Gender and mode of delivery does not seem to have any bearing on severity of jaundice.Thus all new borns should be considered for screening irrespec ve of sex and mode of delivery.This was consistent with observa ons of S Umit et al 12,17 and Frauk Aplay et al 4 .Type of ABO incompa bility did not determine the development of signifi cant jaundice.O-A se ng was observed in 20 and O-B in 12 newborns in signifi cant hyperbilirubinaemia group.Similar observa on were made by S Umit et al who observed O-A incompa bility in 21 and O-B in 8 newborns in hyprebilirubinaemia group 12 .
Birth weight did not have any bearing on the development of signifi cant jaundice.Mean birth wt.was2870± 305 gm in signifi cant hyperbilirubinemia group Vs 2873± 254gm in non signifi cant hyperbilirubinemia group.Frauk Aplay et al reported 3310±305gm Vs 3240±34gm in hyperbilirubinaemia and non hyprebilirubinaemia group respec vely. 4Umit Sarici et al reported birth weight of 2794±418gm and 2772±157gm in two groups.17 S Umit et al observed mean birth weight of 3214±828gmgm and 3212± 196 gm in two groups 12 .
In present study only two babies had weakly posi ve DCT but both developed signifi cant hyperbilirubinaemia requiring intensive phototherapy with peak serum bilirubin of 23mg/d1 and16.5mg/d1respec vely.Thus DCT posi vity predicts development of serve hemoly c disease of newborn.S Umit et al detected posi ve DCT in 6 out of 29 newborns who developed hyperbilirubinaemia.All required intensive phototherapy and one required change trasfussion 12 .H.M. Risemberg et al al noted a strong associa on of strongly posi ve coombs test with hyperbilirubinaemia in ABO incompa bilty 9 .In Moderately aff ected group 9 (60%) were DCT posi ve out of total 15 babies.Coomb test posi vity was 15% in the group who did not develop signifi cant hyperbilirubinaemia.They concluded that DCT is not itself a method to predict hyperbilirubinaemia.Marguerite Herschel et al opined that in ABO incompa ble newborns who are DCT nega ve with signifi cant hyperbilirubinaemia, a cause other then isommunisa on should be sought like G6PD def., Elliptocytosis,Gilibert syndrome etc 14 .
Cord blood haemoglobin cannot be relied upon to predict subsequencet development of hyprebilirubinaemia.This was in concordance with observa on of S Umit al 12 .We observed sta s cally signifi cant diff erence in re culocyte count in two group (2.77±0.60% in group I versus 1.95± 0.32 in group 2).S Umit et al highlighted predic ve values of high re culocyte count for devolpement of signifi cant hyperbilirubinemia.They observed a re culocyte count of 4.39±3.446 in hyperbilirubinemia group 12 .
Mean serum bilirubin in cord blood and subsequent 3 days' serum bilirubin was signifi cantly higher in signifi cant hyperbilirubinemia group as compared to other group.This was in concordance with observa on made by S umit et al 12 .
In our study on construc ng a percen le based nomogram based on age/hour specifi c serum bilirubin levels it was observed that cord serum bilirubin of 2.16mg/dl at 60 th percen le curve has sensi vity, specifi city, NPV and PPV of 100%,91.5%,100% and 35.3% respec vely.It has a high predic ve value for subsequent hyperbilirubinemia requiring interven on.Cord serum bilirubin≥60 percen le cons tutes high intermediate risk group and value≥90 TH percen le(4.09mg/dl)cons tutes high risk group which is good predictor of developing severe hyperbilirubinaemia requiring extensive phototherapy or other appropriate interven on.Two babies in this category had posi ve DCT and developed severe haemoly c disease of new born.
Risemberg observed a strong associa on of cord serum bilirubin of ≥4mg/dl with severe hyperbilirubinemia requiring exchange transfusion necessita ng their placement in centre where frequent evalua on and appropriate therapy are available 9 .Robinsn et.al reported associa on of ABO disease with cord S.bilirubin levels above 3mg/dl 18 .
Similar observa on was made by chen JY,ling UP who suggested that ABO incompa ble babies with cord sb≥4mg/dl or posi ve DCT cons tute a high risk category 10 .S Umit et al 12 in their study of 136 healthy term newborns with ABO incompa bility observed that mean SB≥4mg/dl at 6hrs of life had sensi vity 86.2%, NPV 94.5% and PPV 39.7% and 6mg/dl had sensi vity specifi city NPV and PPV 100%,91.5%,100% and 35.3% respec vely.Using percen le curves, they observed that 35 th and 90 th percen le curves approx.3.3 and 6mg/dl at 6hrs of life can be taken as safe risk demarcators to plan a me of discharge for ABO incompa ble newborn.

Conclusion
No sta s cally signifi cant diff erence was observed in two groups regarding various demographic characteris cs like sex, birth weight, feedings and mode of delivery and type of ABO incompa bility.Cord serum bilirubin, re culocyte count and posi ve DCT could serve as good predictors for development of subsequent hyperbilirubinaemia and severe hemoly c disease of newborn in ABO incompa bility.It was also inferred that newborn with cord serum bilirubin <2.16 mg/d1 were not at risk of developing signifi cant hyperbilirubinaemia.Thus to conclude a cri cal cord S.bilirubin between 2.16 mg/d1 and 4.09mg/d1 could predict all newborns who would have signifi cant hyperbilirubinaemia and could be used as a safe demarcator to decide me of discharge.Any therapeu c interven on if necessary can be started as early as possible.

Table 1 :
Demographic characteris cs of groups with and without signifi cant hyperbilirubinaemia.

Table 2 :
Sequen al S.bilirubin values (mean±SD) in two groups

Table 3 :
Sensi vity, Specifi city, posi ve and nega ve predic ve value at diff erent percen le tracts

Table 4 :
Mean Cord and subsequent serum bilirubin values at 35 th ,60 th and 90 th percen le Risk designa on according to percen le tracks based on age specifi c serum bilirubin values HRZ -high risk zone designed above the 90 th percentile tract.HIRZ -high intermediate risk zone between 60 th and 90 th percentiles.LIRZ -low intermediate risk zone between 35 th and 60 th percentiles.LRZ -low risk zone below 35 th percentile track.