Early Human Development 88 (2012) 51–55

Contents lists available at ScienceDirect

Early Human Development

j ourna l homepage: www.e lsev ie r.com/ locate /ear lhumdev

Validation of transcutaneous bilirubin nomogram in identifying neonates not at riskof hyperbilirubinaemia: A prospective, observational, multicenter study

Costantino Romagnoli a,⁎, Eloisa Tiberi a, Giovanni Barone a, Mario De Curtis b, Daniela Regoli b,Piermichele Paolillo c, Simonetta Picone c, Stefano Anania d, Maurizio Finocchi e,Valentina Cardiello a, Enrico Zecca a

a Department of Paediatrics, Division of Neonatology, Policlinico A. Gemelli, Catholic University Sacred Heart, Rome, Italyb Department of Paediatrics, Division of Neonatology, Policlinico Umberto I, “Sapienza” University, Rome, Italyc Division of Neonatology, Casilino General Hospital, Rome, Italyd Division of Neonatology, San Filippo Neri Hospital, Rome, Italye Division of Neonatology, San Pietro FBF Hospital, Rome, Italy

⁎ Corresponding author at: Department of PaediatricHeart, Largo A. Gemelli, 8, 00168 Rome, Italy. Tel.: +3055301.

E-mail address: cromagnoli@rm.unicatt.it (C. Romag

0378-3782/$ – see front matter © 2011 Elsevier Irelanddoi:10.1016/j.earlhumdev.2011.07.001

a b s t r a c t

a r t i c l e i n f o

Article history:

Received 20 April 2011Received in revised form 30 June 2011Accepted 2 July 2011

Keywords:Transcutaneous bilirubinHyperbilirubinaemiaPredictive nomogramNewborn babies

Background: Transcutaneous bilirubin (TcB) measurement is widely used as screening for neonatalhyperbilirubinaemia.Aims: To prospectively validate TcB measurement using hour-specific nomogram in identifying newborninfants not at risk for severe hyperbilirubinaemia.Study design: prospective, observational, multicenter.Subjects: 2167 term and late preterm infants born in 5 neonatal units in the Lazio region of Italy.Methods: All neonates had simultaneous TcB and total serum bilirubin (TSB) measurements, when jaundiceappeared and/or before hospital discharge. TcB and TSB values were plotted on a percentile-based hour-specific transcutaneous nomogram previously developed, to identify the safe percentile able to predict

subsequent significant hyperbilirubinaemia defined as serum bilirubin N17 mg/dL or need for phototherapy.Results: Fifty-five babies (2.5%) developed significant hyperbilirubinaemia. The 50th percentile of ournomogram was able to identify all babies who were at risk of significant hyperbilirubinaemia, but with a highfalse positive rate. Using the 75th percentile, two false negatives reduced sensitivity in the first 48 hours butwe were able to detect all babies at risk after the 48th hour of age. Conclusions: This study demonstrates thatthe 75th percentile of our TcB nomogram is able to exclude any subsequent severe hyperbilirubinaemia from48 h of life ahead.

© 2011 Elsevier Ireland Ltd. All rights reserved.

1. Introduction

Early postnatal discharge from hospital, within 48 h after birth,combinedwith exclusive breast feeding anddecreased concern about theclinical importanceof toxicpotential of bilirubin, hasbeen relatedwith anincrease in readmissions for phototherapy to treat severe hyperbilir-ubinaemia [1] as well as with a re-emergence of kernicterus [2].

In order to assess the risk of developing severe hyperbilirubinaemiathe American Academy of Pediatrics (AAP) recommends measurementof total serum bilirubin (TSB) in a predischarge newborn populationand plotting the results on an hour specific nomogram (for identifica-tion of severe hyperbilirubinaemia) [3,4]. On the other hand, strategies

s – Catholic University Sacred39 6 30154786; fax: +39 6

noli).

Ltd. All rights reserved.

based on risk factors and TSB [5,6] and on TSB determination either inumbilical cord blood [7] or within the first day of life [8] failed toidentify all newborn babies at risk for developing severe hyperbilir-ubinaemia. However, the determination of TSB levels remains aninvasive, stressful and time consuming procedure.

The determination of transcutaneous bilirubin (TcB) is frequentlyused to reduce the measurements of TSB and to assess significanthyperbilirubinaemia in term and late-preterm infants [9–17]. Untilnow the values of TcB have been used to predict severe hyperbilir-ubinaemia using TSB nomogram [9,11], while a recent trial attemptedto convert pre-test predictive ability into post-test predictivity of TcBmeasurements [18]. In 2008 we constructed a nomogram based onskin bilirubin for the first 96 h of life in a European normal healthypopulation, obtained with multiwavelength transcutaneous bilirubi-nometry from 2198 healthy newborn babies (Table 1) [19].

The aim of this study was to verify the predictive value of ournomogram in identifying a ‘safe percentile’ below which the baby isnot at risk for severe hyperbilirubinaemia.

Table 1Values of TcB corresponding at the 50th, 75th and 90th percentile of the hour-specificnomogram elaborated in our population.

Hrs 50th 75th 90th Hrs 50th 75th 90th Hrs 50th 75th 90th

24 6.3 7.8 11.1 49 7.7 10.4 11.9 73 10.0 11.7 13.725 6.3 7.8 11.1 50 7.8 10.4 12.0 74 10.0 11.8 13.726 6.4 7.8 11.1 51 8.0 10.5 12.0 75 10.1 11.9 13.827 6.4 7.9 11.2 52 8.1 10.5 12.0 76 10.1 11.9 13.828 6.4 7.9 11.2 53 8.3 10.6 12.1 77 10.2 12.0 13.929 6.5 7.9 11.2 54 8.4 10.6 12.1 78 10.2 12.1 13.930 6.5 7.9 11.2 55 8.6 10.7 12.3 79 10.3 12.2 14.031 6.6 8.1 11.2 56 8.7 10.8 12.5 80 10.4 12.2 14.032 6.6 8.4 11.2 57 8.9 11.0 12.8 81 10.5 12.3 14.133 6.7 8.6 11.2 58 9.0 11.1 13.0 82 10.5 12.3 14.234 6.7 8.8 11.2 59 9.2 11.2 13.2 83 10.6 12.4 14.235 6.8 9.1 11.2 60 9.3 11.3 13.4 84 10.7 12.4 14.336 6.8 9.3 11.2 61 9.4 11.3 13.4 85 10.7 12.4 14.237 6.9 9.4 11.3 62 9.5 11.4 13.4 86 10.8 12.4 14.238 7.1 9.5 11.4 63 9.6 11.4 13.5 87 10.8 12.4 14.139 7.2 9.7 11.5 64 9.6 11.4 13.5 88 10.8 12.4 14.040 7.3 9.8 11.5 65 9.7 11.5 13.5 89 10.9 12.4 14.041 7.5 9.9 11.6 66 9.8 11.5 13.5 90 10.9 12.4 13.942 7.6 10.0 11.7 67 9.8 11.5 13.5 91 10.9 12.5 13.943 7.6 10.1 11.7 68 9.8 11.5 13.5 92 10.9 12.5 13.944 7.6 10.1 11.8 69 9.9 11.6 13.6 93 10.9 12.6 13.945 7.6 10.2 11.8 70 9.9 11.6 13.6 94 10.9 12.6 13.946 7.5 10.2 11.8 71 9.9 11.6 13.6 95 10.9 12.7 13.947 7.5 10.3 11.9 72 9.9 11.6 13.6 96 10.9 12.7 13.948 7.5 10.3 11.9

52 C. Romagnoli et al. / Early Human Development 88 (2012) 51–55

2. Methods

2.1. Study Group

This multicenter prospective study was conducted in the region ofLazio, in Italy, through March to December 2009 involving fiveneonatal units. The Institutional Board approved this study in eachsite and all parents signed informed consent. In the study we includednewborn babies with gestational age (GA) of ≥35 weeks, based onpostmenstrual date and early gestation prenatal sonographic findings.We excluded all sick newborn babies who were admitted to theneonatal intensive care unit and those with severe congenitalanomalies. None of the studied babies received a drug therapy exceptfor 0.5–1 mg vitamin K (Konakion, Roche Laboratories, Nutley, NJ,USA) administered intramuscularly or orally soon after their birth.Westarted feeding at 1 hour of life, followed by breast or bottle feedingevery 3 h. No prophylactic intervention for hyperbilirubinaemia wasemployed. During the study period environmental lighting wasconstant.

2.2. Measurements of TcB and TSB

The measurement of TcB was performed in newborn babies ofthree neonatal units if clinically jaundiced and/or just before thedischarge from the hospital, while in two units TcB was measuredonly in jaundiced babies. All determinations were made usingBiliCheck™ ([BC] Respironics, Marietta, GA – USA). The BC technologyallows the optical densities attributed to bilirubin and other skinpigments to be determined avoiding any interference due to severalconfounding factors: its principles of operation have been describedelsewhere [13,15]. Experienced neonatologists performed measure-ments of TcB with 5 readings in different points of neonatal forehead.Reading's locations were distanced from the hairline and free of anybruising, nevus, haemangioma or other skin anomalies. All measure-ments were performed on ambient morning light of the nursery whilethe infant was in a quiet state. The BC devices were calibrated with adisposable tip before each measurement [16]. A single device wasused for all measurements in each neonatal unit. No device failure wasnoticed. In all newborn babies blood samples (50 μL) for the

measurements of TSB were collected by heel stick puncture. Capillarytubes were protected from light exposure and after centrifugationthey were assayed with the direct spectrophotometer (Microbili-meter Dual Beam Plus model 11144A73G, Ginevri, Rome, Italy) for30 min. All measurements of TSB were performed by trainedtechnicians, blind to the value of TcB.

A total of 3241 of paired TcB/TSB determinations was obtainedfrom 2167 newborn babies because 728 babies had more than oneTcB/TSB measurements. In detail, 506 newborn babies had 2measurements, 176 had 3 measurements, and 46 had more than 3measurements.

2.3. Follow-up of studied newborn babies

In each unit, newborn babies are never discharged before 72 h ofage independently from the mode of delivery. All newborn babieswith a predischarge TcB value N75th percentile of our nomogramwere discharged only after two consecutive decreased TSB values,12 h apart, making us able to identify the peak TSB level. The newborninfants with pre-discharge TcB level between the 50th and the 75thpercentile were discharged and controlled 48 h later for hyperbilir-ubinaemia (determination of TSB). Parents of infants with TcB b50thpercentile were counseled to return to the hospital within 5 days afterthe discharge from the hospital or earlier if they observed persistentjaundice. The decision to use phototherapy was made by theattending neonatologist according to AAP guidelines. For babiesexposed to phototherapy we considered only pre-treatment mea-surements. All perinatal data were recorded in a single database foreach site with a selected log of any event occurring during the studyperiod. Care was taken that the same clinical protocol study, methodfor sample collection and strategies for patient recruitment wereprospectively maintained, so that the data from each unit could bepooled.

2.4. Outcome

In our percentile-based hour-specific nomogram the measure-ments of TcB were plotted separately by two researchers (CR and EZ)after completion of the study. Significant hyperbilirubinaemia wasdefined as TSB value N17 mg/dL, or as need for phototherapytreatment according to AAP guidelines.

2.5. Statistical Analysis

We performed a statistical analysis using Student's t-test forcontinuous predictors and Fisher's test for categorical data. Weassessed the mean difference between paired TcB and TSB, andcalculated the correlation of TSB values tomeasurements of TcB by thelinear regression analysis. We calculated the sensitivity, specificity,positive predictive and negative values plotting TcB data in the 50th,75th and 90th percentile of our TcB nomogram. Receiver operatingcharacteristic (ROC) curve analysis was performed with SPSSsoftware, which was used to assess the predictive ability of our TcBnomogram.

3. Results

A total of 2167 newborn babies (1137 males and 1030 females) ofwhich 184 (8.5%) late preterm were enrolled in the study (Table 2).Mean gestational age (±SD) was 38.9±1.5 weeks (range: 35–42)and mean birth weight (BW) was 3237±471 g (range: 2000–5090).The majority of babies were Caucasian (90.1%) and 53.5% was bornafter spontaneous delivery. Exclusive breast feeding was prevalent,but 39.1% of babies received also bottle feeding during their hospitalstay. Delayed meconium passage (N24 h of life) was observed in 248(11.4%) babies and only 127 (5.9%) experienced a weight loss greater

Table 2Baseline characteristics of the study population. Values expressed as mean±SD ornumber (%).

Variables Neonates (2167)

Gestational age, weeks 38.9±1.5N37 weeks 1983 (91.5)35–36 weeks 184 (8.5)Birth weight, grams 3237±471Small for gestational age 115 (5.1)Male 1137 (52.5)Mode of deliveryVaginal 1159 (53.5)Cesarean section 953 (44)Ventouse 55 (2.5)FeedingExclusive breast feeding 1267 (58.5)Breast+bottle feeding 848 (39.1)Bottle feeding 52 (2.4)Delayed meconium passage (N24 hrs) 248 (11.4)Weight loss N10% 127 (5.9)Age at TcB/TSB hours 63±21TcB value, mg/dL 8.5±0.8TSB value, mean±SD, mg/dL 9.4±0.6Significant hyperbilirubinemia 55 (2.5)Required phototherapy 46 (2.1)TSB N17 mg/dL 9 (0.4)

53C. Romagnoli et al. / Early Human Development 88 (2012) 51–55

than 10%. Out of the total population studied, 55 newborn babies(2.5%) were diagnosed as having significant hyperbilirubinaemia(3.6% in the 2 units where measurements were performed only injaundiced babies, and 1.2% in the other 3 units). According to AAPguidelines 46 newborn babies (2.1%) received phototherapy while 9(0.4%) reached TSB values N17 mg/dL but were not treated. None ofthe babies required an exchange transfusion. No cases of significanthyperbilirubinaemia were noticed among not jaundiced babies whohad the routine pre-discharge test, or after discharge. Linearregression showed a significant correlation between TcB and TSB(r: 0.8570; pb0.0001) with slight differences among neonatal units(r value ranging from 0.800 to 0.896). The correlation between TcBand TSB values was independent by GA, gender, hour of age, TSBlevel except for values N15 mg/dL, and ethnic group except forHispanic babies. The mean difference between paired TcB and TSBmeasurements was 1.0 mg/dL with a 95% confidence interval of −2.4and+4.5 mg/dL. Table 3 shows the predictive ability of the percentilesof our TcB nomogram to identify babies with significant hyperbilir-ubinaemia (TSB N17 mg/dL or need for phototherapy). In the first 48 hof age 100% of sensitivity was reached only with the 50th percentile ofour nomogram,while after the 48th hour the 75th percentile allows usto predict all newborn babies without subsequent significant hyperbi-lirubinaemia. Two false negatives have been observed between 24 and48 h of age, but a subsequent TcBmeasurement after 24 h showed truepositive results.

Table 3Ability of TcB measurements below the 50th, 75th and 90th percentile of TcB nomogram to pfalse negative; TN true negative; FP false positive; PPV positive predictive value; NPV nega

Hours of age TP (n) FN (n) TN (n) FP (n)

24 to 48 hb50th percentile 27 0 161 460b75th percentile 25 2 429 192b90th percentile 14 13 564 5749 to 72 hb50th percentile 21 0 410 615b75th percentile 21 0 702 323b90th percentile 18 3 888 13773 to 96 hb50th percentile 7 0 246 220b75th percentile 7 0 348 118b90th percentile 5 2 409 57

Fig. 1 shows the ROC curves in which the true positive rate(sensitivity) is plotted in functionof the falsepositive rate (1-specificity)for the 50th, 75th and 90th percentiles of TcB. The area under the ROCcurve (AUC ) measures the accuracy of the percentiles in predictingsignificant hyperbilirubinaemia from 24 to 96 h of life (A), and from 49to 96 h of life (B). The AUC for the 49–96 h of age is somewhat higherthan that for the 24–96 h of age, and shows better values of sensitivity(100% vs. 96.4%) and specificity (70.4% vs. 70.0%).

4. Discussion

The identification of newborn infants at risk for developingsignificant hyperbilirubinaemia based on predischarge risk assess-ment using measurements of TSB plotted with hour-specific nomo-gram developed by Bhutani et al. has been validated and widelyaccepted [4,20]. This approach led to a significant reduction of babiesreaching TSB levels that are known to be at risk for bilirubin braindamage without any effect on kernicterus [21–24].

With the advent of transcutaneous determination of bilirubin wehave a tool able to reduce the number of invasive and painful bloodsampling. Many authors tried to identify risk groups using measure-ments of TcB for predischarge bilirubin screening. Firstly, Bhutani etal. suggested that babies with predischarge BC values above the 75thpercentile of hour-specific TSB nomogram were considered at highrisk for subsequent significant hyperbilirubinaemia because none ofthe babies with significant hyperbilirubinaemia had a BC measure-ment b75th percentile on the same TSB nomogram [9]. Worst resultswere obtained by Dalal et al. [25] plotting TcB measurements onBhutani's nomogramusing the 75th percentile as the cut-off thresholdfor risk of hyperbilirubinaemia. Varvarigou et al. [18] developed apredictive nomogram, based on a large number of measurements ofTcB, in order to assess the risk for significant hyperbilirubinaemia inhealthy term and late pre-term newborn babies. With the use oflikelihood ratios, the high- and low-risk demarcators for eachdesignated timewere calculated, but unfortunately 100% of sensitivitywas not reached, calculating post-test probability of significanthyperbilirubinaemia on the same population.

More recently, Yu et al. [26] developed an hour specific TcBnomogram for Chinese newborn infants. A 100% sensitivity wasreported only using the 40th percentile calculating the predictiveability on the same population used to elaborate nomogram. Maiselsand Kring obtained measurements of TcB, at 6-hour intervals, in thefirst 96 hours after birth in a population of babies with GA ≥35 weeksand they elaborated hour-specific nomograms for babies of different GA[11]. However, the predictive ability of this nomogram was tested onlyin combination with clinical risk factors [6]. Hour-specific nomogrambased on TcB measurements have been elaborated by Sanpavat et al.[27] in a Thai population but a prospective validation is lacking.

Concerns regarding the utility of TcB or TSB values in predictingsubsequent severe hyperbilirubinaemia as a consequence of false-

redict significant hyperbilirubinaemia, for designated time periods. TP true positive; FNtive predictive value.

Sensitivity (%) Specificity (%) PPV (%) NPV (%)

100 25.9 5.5 10092.6 69.1 11.5 99.551.9 90.8 19.7 97.7

100 40.0 3.3 100100 68.5 6.1 10085.7 86.6 11.6 99.7

100 52.8 3.1 100100 74.7 5.6 10071.4 87.8 8.1 99.5

Fig. 1. ROC curve and AUC from 24 to 96 h of life (A) and from 49 to 96 h (B). Sensitivityand specificity of 75th percentile in predicting severe hyperbilirubinaemia.

54 C. Romagnoli et al. / Early Human Development 88 (2012) 51–55

negative results from pre-discharge neonatal bilirubin screening wereexpressed in two recent papers. A study from South Carolina [28]documented that of 6220 newborn babies discharged after routinepredischarge TcB testing 28 (0.43%) were readmitted because ofsevere hyperbilirubinaemia of which 13 showed pre-dischargebilirubin b75th percentile, and one b40th percentile of Bhutani'snomogram. False negative results have been reported also byRodriguez-Capote et al. [29]. These authors have questioned whetherplotting TcB value on TSB nomogram is a correct procedure.

We previously developed our hour-specific percentile-basednomogram using serial measurements of TcB in healthy term andlate pre-term babies in the first 96 h of age [19]. In the present studywe report the prospective validation of our TcB nomogram to identifynewborn babies at risk for significant hyperbilirubinaemia in amulticenter study involving a large neonatal population. Our resultssuggest that individual hour specific TcB values b75th percentilebetween 49 and 96 h of age are able to predict all babies who will not

develop subsequent significant hyperbilirubinaemia, defined as TSBvalue N17 mg/dL or as need for phototherapy treatment.

Sensitivity before 48 h of age (92.6%) was affected by two falsenegatives. Of these, the first baby was a white infant born at 38 weeksgestationwith BWof 3090 grams after spontaneous delivery, who hadhis first TcB measurement at 44 h of age whose result was falselynegative, but with a second measurement of TcB at 68 h of age whoseresult was true positive. This baby was bottle fed and reachedmaximum TSB value of 17.7 mg/dL at 68 h of age. The second was ablack baby born at 39 weeks gestation with BW of 3460 g aftercaesarean section, who had his first measurement of TcB at 46 h of agewhose result was falsely negative, but with a second measurement ofTcB at 65 h of age whose result was true positive. The baby wasexclusively breast fed and reached maximum TSB value of 18.5 mg/dLat 82 h of age. Interestingly, both the two false negative results turnedinto true positive after a second TcB determination within thefollowing 24 h. This observation suggests the possibility that morethan one measurement obtained 12–24 h apart could optimize thepredictive ability of TcBmeasurements as suggested also by Dalal et al.[25]. However, that may lead to repeated bilirubin determinationand/or a prolongation of hospital stay and a specific follow-up, withan increase in hospital costs. It is important to consider that our TcBdata obtained soon before discharge plotted on our TcB nomogram ledto identify all babies at no risk for significant hyperbilirubinaemia.

This study has some limitations. Firstly, it is not a population-basedstudy but the enrolment is based on clinical practice of each neonatalunit. Secondly, we have a low incidence of significant hyperbilir-ubinaemia. Thirdly, our study includes a majority of Caucasiannewborn babies, but BC has been shown not to be affected by skinpigmentation [13,15]. The strengths of this study are its robust design,large sample size and follow-up rate for enrolled newborn babies. Thisis a prospective observational study in which physician performingmeasurement of TcB were blind to the TSB values and vice versa.Moreover, we plotted the obtained data on TcB hour-specificnomogram by two physicians after the conclusion of the study. Allbabies enrolled were studied during their hospital stay (N72 h of age)and that made the studied population unbiased. Finally, all babiesreferred to the hospital birth for clinical evaluation after discharge andno cases of readmission for hyperbilirubinaemia were noticed.However, we do not claim for universal use of our nomogram and itcould be useful for each neonatal unit to develop its own nomogram.

In summary, with a percentile-based hour-specific TcB nomogramwe showed that after 48 h of age the 75th percentile is useful to selectnewborn babies with no risk for subsequent significant hyperbilir-ubinaemia. Our data can encourage the use of TcB for a safe dischargeof late preterm and term newborn babies between 48 and 72 h of life.

Competing interests

There are no conflicts of interest to disclose, financial or otherwise.

Acknowledgements

The authors would like to thank Miss Marta Romagnoli for herhelpful reviews of the manuscript.

References

[1] Maisels MJ, Kring E. Length of stay, jaundice, and hospital readmission. Pediatrics1998;101:995–8.

[2] Maisels MJ, Newman TB. Kernicterus in otherwise healthy, breast-fed termnewborns. Pediatrics 1995;96:730–3.

[3] AAP Subcommittee on Hyperbilirubinaemia. Management of hyperbilirubinaemiain the newborn infant 35 or more weeks of gestation. Pediatrics 2004;114:297–316.

[4] Bhutani VK, Johnson L, Sivieri EM. Predictive ability of a predischarge hour specificserum bilirubin for subsequent significant hyperbilirubinaemia in healthy termand near-term newborns. Pediatrics 1999;103:6–14.

55C. Romagnoli et al. / Early Human Development 88 (2012) 51–55

[5] Keren R, Luan X, Friedman S, Saddlemire S, Cnaan A, Bhutani VK. A comparison ofalternative risk assessment strategies for predicting significant neonatal hyperbi-lirubinaemia in term and near-term infants. Pediatrics 2008;121:e170–9.

[6] Maisels MJ, DeRidder JM, Kring EA, Balasubramaniam M. Routine transcutaneousbilirubin measurements combined with clinical risk factors improve theprediction of subsequent hyperbilirubinaemia. J Perinatol 2009;29:612–7.

[7] Knupfer M, Pulzer F, Gebauer C, Robel-Tillig E, Vogtmann C. Predictive value ofumbilical cord blood bilirubin for postnatal hyperbilirubinaemia. Acta Paediatr2005;94:581–7.

[8] Alpay F, Sarici SU, Deniz Tosuncuk H, Serdar MA, Inanc N, Gokcay E. The value offirst-day bilirubin measurement in predicting the development of significanthyperbilirubinaemia in healthy term newborns. Pediatrics 2000;106(2). Availableat http://www.pediatrics.org/cgi/content/full/106/2/e16.

[9] Bhutani VK, Gourley GR, Adler S, Kreamer B, Dalman C, Johnson LH. Noninvasivemeasurement of total serum bilirubin in a multiracial predischarge newbornpopulation to assess the risk of severe hyperbilirubinaemia. Pediatrics 2000;106(2). Available at www.pediatrics.org/cgi/content/full/106/2/e17.

[10] EngleWD, Jackson GL, Stehel EK, Sendelbach DM,ManningMD, Frawle. Evaluationof a transcutaneous jaundice meter following hospital discharge in term and near-term neonates. J Perinatol 2005;25:486–90.

[11] Maisels MJ, Kring E. Transcutaneous bilirubin levels in the first 96 hours in anormal newborn population of ≥35 weeks' gestation. Pediatrics 2006;117:1169–73.

[12] Bhat YR, Rao A. Transcutaneous bilirubin in predicting hyperbilirubinaemia interm neonates. Indian J Pediatr 2008;75:119–23.

[13] Ebbesen F, Rasmussen LM, Wimberley PD. A new transcutaneous bilirubinometer,BiliCheck, used in the neonatal intensive care unit and maternity ward. ActaPaediatr 2002;91:203–11.

[14] Wong CM, van Dijk PJE, Laing IA. A comparison of transcutaneous bilirubin-ometers: SpectRx BiliCheck versus Minolta Air-Shields. Arch Dis Child FetalNeonatal Ed 2002;87:F137–40.

[15] Rubaltelli FF, Gourley GR, Loskamp N, Modi N, Roth-Kleiner M, Sender A, et al.Transcutaneous bilirubin measurement: a multicenter evaluation of a new device.Pediatrics 2001;107:1264–71.

[16] De Luca D, Zecca E, de Turris P, Barbato G, Marras M, Romagnoli C. UsingBilicheckTM for preterm neonates in a sub-intensive unit: diagnostic usefulnessand suitability. Early Hum Dev 2007;83:313–7.

[17] De Luca D, Zuppa AA, Cota F, Zecca E, Romagnoli C. Multiwavelengthtranscutaneous bilirubinometry: suitability and cost analysis. Ital J Pediatr2006;32:328–9.

[18] Varvarigou A, Fouzas S, Skylogianni E, Mantagu L, Bougioukou D, Mantagos S.Transcutaneous bilirubin nomogram for prediction of significant neonatalhyperbilirubinaemia. Pediatrics 2009;124:1052–9.

[19] De Luca D, Romagnoli C, Tiberi E, Zuppa AA, Zecca E. Skin bilirubin nomogram forthe first 96 hours of life in a European normal healthy population, obtained withmultiwavelength transcutaneous bilirubinometry. Acta Paediatr 2008;97:146–50.

[20] Johnson LH, Bhutani VK, Brown AK. System-based approach to management ofneonatal jaundice and prevention of kernicterus. J Pediatr 2002;140:396–403.

[21] Eggert LD, Wiedmeier SE, Wilson J, Christensen RD. The effect of instituting ofprehospital-discharge newborn bilirubin screening program in an 18-hospitalhealth system. Pediatrics 2006;117:e855–62.

[22] Maisels MJ, Bhutani VK, Bogen D, Newman TB, Stark AR, Watchko JF.Hyperbilirubinaemia in the newborn infants N35 weeks' gestation: an updatewith clarification. Pediatrics 2009;124:1193–8.

[23] KuzniewiczMW, Escobar GJ, Newman TB. Impact of universal bilirubin screening onsevere hyperbilirubinaemia and phototherapy use. Pediatrics 2009;124:1031–9.

[24] Newman TB. Universal bilirubin screening, guidelines and evidence. Pediatrics2009;124:1199–202.

[25] Dalal SS, Mishra S, Agarwal R, Deorari AK, Paul V. Does measuring the changes inTcB value offer better prediction of hyperbilirubinaemia in healthy neonates?Pediatrics 2009;124:e851–7.

[26] Yu Z-B, Dong X-Y, Han S-P, Chen Y-L, Qiu Y-F, Sha L, et al. Transcutaneous bilirubinnomogram for predicting neonatal hyperbilirubinemia in healthy term and late-preterm Chinese infants. Eur J Pediatr 2010, doi:10.1007/s00431-010-1281-9.

[27] Sanpavat S, Nuchprayoon I, Smathakanee C, Hansuebsai R. Nomogram forprediction of the risk of neonatal hyperbilirubinaemia, using transcutaneousbilirubin. J Med Assoc Thai 2005;88:1187–93.

[28] Slaughter J, Annibale D, Suresh G. False-negative results of pre-discharge neonatalbilirubin screening to predict severe hyperbilirubinaemia: a need for caution. Eur JPediatr 2009;168:1461–6.

[29] Rodriguez-Capote K, Kim K, Paes B, Turner D, Grey V. Clinical implication of thedifference between transcutaneous bilirubinometry and total serum bilirubin forthe classification of newborns at risk of hyperbilirubinaemia. Clin Biochem2009;42:176–9.