Home Print this page Email this page Small font sizeDefault font sizeIncrease font size
Users Online: 609
About us Editorial board Search Ahead of print Current issue Archives Submit article Instructions Subscribe Contacts Advertise Login 

 Table of Contents  
Year : 2017  |  Volume : 6  |  Issue : 3  |  Page : 148-153

Effect of phototherapy on the diagnostic accuracy of transcutaneous bilirubin in preterm infants

1 Department of Pediatrics, Division of Neonatology, University of Alberta, Alberta, Canada
2 Department of Pediatrics, Division of Neonatology, University of Alberta; Staff Neonatologist, Royal Alexandra Hospital, Alberta Health Services, Edmonton, Alberta, Canada

Date of Web Publication11-Jul-2017

Correspondence Address:
Manoj Kumar
Department of Pediatrics, Edmonton Clinical Health Academy, Room 3-528, 11405. 87 Avenue NW, Edmonton, Alberta T6G 1C9
Login to access the Email id

Source of Support: None, Conflict of Interest: None

DOI: 10.4103/jcn.JCN_94_16

Rights and Permissions

Objectives: The objective of this study was to evaluate the diagnostic accuracy of JM-103 transcutaneous bilirubinmeter (TcB) in preterm infants during phototherapy (PT) and in post-PT phase. Methods: This was a prospective cohort study. Infants born between 28 and 35 weeks of gestation and at <28 days of postnatal age were eligible if they required bilirubin estimation during the NICU stay. TcB was measured within 30 minutes of the blood sampling for serum bilirubin, at forehead and sternum. Mean difference (±standard deviation [SD]) and 95% limits of agreement between TcB and total serum bilirubin (TSB) were calculated by analyzing Bland–Altman difference plots. We also calculated correlation coefficients. Results: Ninety infants were enrolled. During PT, the difference plots revealed a wide TcB-TSB disagreement; 95% agreement limits of data indicated that TcB could underestimate bilirubin levels by up to 132 μmol/L at forehead and 157 μmol/L at sternum (Mean difference ± SD of −52.4 mmol/L ± 40.7 at forehead and −69.2 mmol/L ± 42.5 at sternum) and a poor TcB-TSB correlation (r = 0.72 at forehead and 0.51 at sternum). In post-PT phase, correlation coefficients improved significantly and were equivalent to the estimates before the onset of PT (r = 0.88 at forehead and 0.87 at sternum). The analyses of difference plots revealed that TcB could underestimate bilirubin levels up to 88 μmol/L in this phase (Mean difference ± SD of −28.8 mmol/L ± 30.5 at forehead and −19.6 mmol/L ± 34.7 at sternum). Conclusions: JM-103 device is unreliable for estimating bilirubin during PT. However, TcB-TSB agreement improved substantially in the post-PT phase such that use of the TcB device could lead to a reduction in blood sampling during this phase.

Keywords: Bland–Altman difference plots, correlation coefficient, hyperbilirubinemia, phototherapy, transcutaneous bilirubin

How to cite this article:
Nagar G, Kumar M. Effect of phototherapy on the diagnostic accuracy of transcutaneous bilirubin in preterm infants. J Clin Neonatol 2017;6:148-53

How to cite this URL:
Nagar G, Kumar M. Effect of phototherapy on the diagnostic accuracy of transcutaneous bilirubin in preterm infants. J Clin Neonatol [serial online] 2017 [cited 2020 Jul 14];6:148-53. Available from: http://www.jcnonweb.com/text.asp?2017/6/3/148/210147

  Introduction Top

Jaundice is common in newborn period with a majority of neonates developing visible jaundice within the first few days of birth.[1],[2] A small proportion of these infants would develop significant bilirubin levels requiring treatment with phototherapy (PT).[3] Preterm infants are at a greater risk for developing severe jaundice requiring treatment [4] due to a slower postnatal maturation of hepatic bilirubin uptake and conjugation mechanisms in this population.[5] In addition, delay in establishing enteral feeding in preterm infants contributes further to severe jaundice in this population.[5] At present, these infants are screened for hyperbilirubinemia in the majority of the neonatal Intensive Care Units (NICUs) by measuring serum bilirubin levels which require frequent blood sampling. There are reports of potential adverse long-term consequences in neonates subjected to repeated painful stimuli.[6]

Transcutaneous bilirubinmeter (TcB) devices estimate serum bilirubin noninvasively from the skin. Since the publication of the first report of a TcB device used in patients by Yamanouchi et al.,[7] advanced versions of these devices have been tested in clinical studies.[8],[9],[10] American Academy of Pediatrics [11] recommends the use of TcB devices for screening of hyperbilirubinemia in term and near-term infants in immediate neonatal period. Furthermore, a recent systematic review of diagnostic accuracy studies showed the excellent reliability of some of these devices in preterm population before the onset of PT.[12]

However, the reliability of TcB devices for the assessment of neonatal jaundice during PT and in the post-PT phases is debated.[13],[14],[15],[16],[17] The evidence from the available studies in this regard is limited by several factors, i.e., the use of older versions of TcB devices,[7],[16],[18] a small number of patients enrolled,[19],[20],[21] and poor statistical methods employed.[13],[22],[23] To further reduce blood sampling in neonatal units, it would helpful if accuracy of TcB devices could be demonstrated under these circumstances.

The objective of this study was to evaluate the diagnostic accuracy of a commonly used TcB device (JM-103) in preterm infants during PT and post-PT phases. We also compared TcB-total serum bilirubin (TSB) agreement at forehead, a site typically shielded from PT light, versus at sternum.

  Methods Top

Design and setting

This prospective study enrolled participants from the two large NICUs in the western Canada, from January 2014 to November 2014. Parents or guardians of all potential study subjects received written study information brochures upon admission to the NICU, and an informed consent was obtained from those who agreed to participate. The study was approved by the Institutional Review Board at the University of Alberta, Canada. The funder had no role in the conduct of the study or in publication of results.

The newborns were eligible for inclusion if they were 28–35 weeks of gestation at birth and were <28 days of age and required TSB measurement as per the discretion of the treating physicians not involved in the study. Exclusion criteria were previous exchange transfusion; suspected or confirmed hepatic disease with cholestasis, generalized skin disease; episode of necrotizing enterocolitis (modified Bell stage >=2) or sepsis during the study period.

TcB and TSB measurements

TcB was measured on the infants in quiet state by bedside nurses using JM-103 device at two body sites, i.e., forehead and sternum. During PT, the forehead of infants was shielded from bright PT light by an opaque headband (Biliband ®) thus providing TcB readings from the “covered site” for comparison with the TSB. The TcB was simultaneously measured at the sternum, providing data for the “uncovered site.” PT source was temporarily switched off for measurements during the phototherapy phase. Bedside nurses were requested to conduct TcB measurements within 30 min of blood sample collection for TSB and were unaware of the TSB results at the time of measurement. TcB devices were regularly calibrated according to the manufacturer's recommendations.[24] The TSB was measured by direct spectrophotometry using Reichert UNISTAT Spectrophotometer.

Outcome measures

The difference between TcB and TSB values (TcB-TSB) was plotted against the average of the TcB and TSB readings for each patient to calculate mean bias and the standard deviation (SD) of the differences as originally described by Bland–Altman (BA difference plots).[25] The upper and lower limits of agreement (mean bias ± 2 SD of the differences) representing 95% of the data points were calculated. We also calculated the commonly used correlation coefficient statistic. Agreement statistics were calculated separately for each of the three phases, i.e., pre-PT, PT, and post-PT phases.

A-priori subgroup analyses were planned according to the gestational age at birth (28–31 weeks, 32–35 weeks) and as per the time period after the discontinuation of PT. Statistical analyses were conducted using Microsoft Excel 2007 spreadsheet and Stata version 13 (StataCorp LP, College Station, Texas, USA). Minitab version 16 (Minitab Inc., State College, Pennsylvania, USA) was used to generate BA difference plots.

  Results Top

A total of ninety infants were enrolled in the study contributing data for various phases of the study. The mean (SD) birth weight (BW) was 1847 g (±478 g) and mean (SD) gestational age was 32.4 (±1.89 weeks) for the study participants. The demographic characteristics of the included subjects are presented in [Table 1]. Approximately 30% of the participants were <1500 g BW and/or <32 weeks of gestation. The main results of the study are presented in the [Table 2] and [Table 3].
Table 1: Demographic characteristics of the included infants

Click here to view
Table 2: Comparison of transcutaneous bilirubinmeter measurements at forehead with total serum bilirubin

Click here to view
Table 3: Comparison of transcutaneous bilirubinmeter measurements at sternum with total serum bilirubin

Click here to view

Prephototherapy phase

There were 81 and 80 readings of TcB from the forehead and sternum sites, respectively, for comparison with the TSB data. The correlation coefficients between TcB and TSB were 0.85 at forehead and 0.87 at sternum [Figure 1]. BA difference plots revealed minimal bias between the two methods of bilirubin measurement. The results were comparable between the two measurement sites, i.e., mean difference of 0.65 μmol/L (SD: 28 μmol/L; 95% agreement limits -54 and 55 μmol/L) at forehead and 5.8 μmol/L (SD: 27.9 μmol/L; 95% agreement limits −48.1 and 60.6 μmol/L) at sternum [Figure 2].
Figure 1: Transcutaneous bilirubin and total serum bilirubin correlation before the onset of phototherapy

Click here to view
Figure 2: Bland–Altman difference plots showing the 95% limits of agreement for transcutaneous bilirubin and total serum bilirubin measurements before the onset of phototherapy

Click here to view

During phototherapy

A total of 66 and 65 pairs of TcB and TSB measurements for forehead and for sternum, respectively, were available for comparison during the PT phase. The TcB and TSB measurements correlated poorly during the PT phase particularly at sternum (r = 0.72 at forehead and 0.51 at sternum) [Figure 3]. The analyses of the BA difference plots revealed significant bias and imprecision in the TcB measurements. TcB underestimated jaundice level with mean TcB-TSB difference of −52.4 μmol/L at forehead ([SD: 40.7 μmol/L], 95% agreement limits −132 and 26 μmol/L) and −69.2 μmol/L at sternum ([SD: 42.5 μmol/L], 95% agreement limits −157and 19 μmol/L) [Figure 4].
Figure 3: Transcutaneous bilirubin and total serum bilirubin correlation at forehead and sternum during phototherapy

Click here to view
Figure 4: Bland–Altman difference plots showing the 95% limits of agreement for transcutaneous bilirubin and total serum bilirubin measurements during phototherapy

Click here to view



A total of 61 pairs of TcB and TSB measurements were available for comparison. The correlation between TSB and TcB improved in the post-PT phase, reaching similar values as obtained in the pre-PT phase (r = 0.88 for all data; 0.84 for readings within 24 h and 0.89 for readings after 24 h) [Figure 5]. The analyses of the BA difference plots revealed improved precision but continuing bias in terms of underestimation of jaundice level with TcB (mean difference −28.8 [SD: 30.5 μmol/L]; 95% agreement limits −88 and 31 μmol/L) [Figure 6].
Figure 5: Transcutaneous bilirubin and total serum bilirubin correlation at forehead and sternum in postphototherapy phase

Click here to view
Figure 6: Bland–Altman difference plots showing the 95% limits of agreement for transcutaneous bilirubin and total serum bilirubin measurements in postphototherapy phase

Click here to view


A total of 66 pairs of TSB and TcB measurements at sternum were available for comparison. The TcB and TSB data correlated as well as in the pre-PT phase (r = 0.87 for all data; 0.80 for readings within 24 h and 0.90 for readings after 24 h) [Figure 5]. The analyses of the BA difference plots revealed continuing bias toward underestimation of jaundice level with the TcB method (mean difference: −19.6 [SD: 34.7 μmol/L]; 95% agreement limits −87 and 47 μmol/L) [Figure 6]. There was no significant difference noted in agreement statistics between the exposed and unexposed site.

  Discussion Top

The results of our study demonstrate a large bias and imprecision with TcB measurements in preterm infants during PT, both at forehead and at sternum. The 95% limits of agreements indicate that the TcB device could underestimate bilirubin level by up to 132 and 157 μmol/L at forehead and at sternum, respectively, during PT. In comparison, in the pre-PT phase, the risk of underestimation by TcB was <55 μmol/L for both sites, similar to those reported in the literature.[12]

In the post-PT phase, we noted significant improvement in bias and precision in the TcB estimates as compared to estimates during the PT phase. The 95% limits of agreements indicate that the underestimation of bilirubin level with the use of TcB device during post-PT phase could be up to 88 μmol/L. These magnitude of underestimation noted here are similar those noted in a recent study involving term and preterm infants. Grabenhenrich et al.[26] in this study showed that the risk of underestimation is particularly high in the first 8 h after stopping PT. Although we were unable to analyze our data according to the time cutoffs used in that study due to very few assessments conducted within 8 h of stopping PT, we also noted a better correlation between TcB and TSB when subjects were off-PT for ≥24 h as compared to <24 h.

Based on the results of our study, JM-103 device cannot be recommended for the estimation of bilirubin during PT due to a significant risk of underestimation of serum bilirubin. However, it may be possible to employ these devices in the post-PT phase to reduce blood sampling for some infants. Based on the data presented here, and in another recent study [26] using similar TcB device, we propose that the blood sampling for bilirubin estimation in the post-PT phase could be avoided if the TcB reading is more than 85 μmol/L below the PT threshold. Similarly, a TcB reading above the PT threshold may be sufficient grounds to initiate PT without the confirmatory blood test (theoretically, some of these infants could be below the PT line as per their serum levels but are still likely to be reasonably close to the threshold for starting PT). In the borderline cases, following a trend of TcB values over a period of time could help with a decision regarding the need for blood sampling.

Our results add to the limited literature on the use of the TcB devices in infants during and after PT. Other authors have also demonstrated poor TcB-TSB correlation in preterm infants receiving PT [13],[14],[16] with some improvement during the post-PT phase.[16] However, these studies presented data in terms of correlation coefficients which are difficult to interpret in clinical practice as those data do not provide information about the expected magnitude of difference between the TcB and TSB measurements. On the other hand, the data presented here as BA difference plots allow prediction of a range of actual bilirubin levels likely for each TcB measurement.

A few studies, presenting the TcB and TSB differences in absolute terms during PT, were able to show better agreement as compared to our results when the site of measurement was shielded from the direct PT light. Nanjundaswamy et al.[14] in their study enrolling seventy preterm infants, reported only 2.8% of the TcB readings from the unexposed areas of forehead underestimated serum bilirubin levels by >25 μmol/L. Jangaard et al.,[20] in a small study enrolling 24 term and preterm infants, reported a variation (mean ± 1.96 SD) of −40.4 to +31 μmol/L between TcB and TSB. Similarly, Zecca et al.,[27] in their study enrolling 253 preterm infants showed negligible bias between TSB and TcB values from a patched area (however, as per the variance data provided by authors, TcB could have underestimated bilirubin level by over 100 μmol/L). Each of these studies used Bilicheck device which measures bilirubin by analyzing multiwavelength reflectance data, as compared to JM-103 which provides its results based on the analysis of optical density data from only two wavelengths of light. Of note, no difference in terms of the accuracy between these two devices has been reported in the pre-PT phase.[12]

The present study to the best of our knowledge is the first study of sufficient size, enrolling an exclusive preterm population, evaluating JM-103 device both during and in post-PT phases. We provided data for BA difference plots along with correlation coefficients, which are superior statistical method to evaluate statistical agreement between diagnostic tests than correlation coefficients. In addition, BA plots provide precision of agreement in numerical terms which is more useful in clinical practice.

Our study is not without limitations. First, our subjects were predominantly drawn from a white population, with majority ≥32 weeks gestation. It is, therefore, not clear how the JM-103 device will perform during post-PT phase in other populations and/or in infants of lesser gestation. Second, we were unable to collect bilirubin data during PT and post-PT phases on approximately 30% of our subjects enrolled, as there is a strong movement towards ordering less blood work in neonates in our units. A significant percentage of our neonates with DAT negative jaundice are now routinely managed with 24–48 h of PT without bilirubin estimation during PT. TcB is often estimated 24 h after stopping PT in such cases, without concurrent serum bilirubin estimation. Since the incidence of glucose-6-phosphate dehydrogenase deficiency and other conditions with DAT negative hemolysis is very low in our setting, and with the use effective LED phototherapy (neoBLUE®) in our units, we rarely encounter significant elevation of bilirubin in low-risk infants while on PT. As such, some of the attending physicians chose not to order concurrent blood sample on low-risk study infants. Third, while we noted improvement in correlation between TcB and TSB readings 24 h after discontinuation of PT, we were unable to test whether the agreement would have improved further after 48–72 h of PT discontinuation due to the paucity of data points to conduct those analyses. Finally, we did not evaluate the effect of the duration and/or multiple episodes of PT on the accuracy of TcB in the post-PT phase.

  Conclusions Top

The results from our study show that JM-103 device cannot be relied upon for the estimation of jaundice in infants receiving PT due to a significant risk of underestimation of the bilirubin level. However, it is possible to use this device in the post-PT phase to reduce blood sampling, employing the limits of agreement data presented here. Further studies are needed to test the accuracy of TcB measured by Bilicheck device during PT phase for possible use in clinical practice. Future studies should also compare Bilicheck and JM-103 devices for accuracy in the post-PT phase.


We are grateful to Barb Kamstra (RN) and Heather Rylance (RN) for helping with the data collection for this study.

Financial support and sponsorship

The grant for this study was provided by Covenant Health (Alberta, Canada).

Conflicts of interest

There are no conflicts of interest.

  References Top

Bhutani VK, Stark AR, Lazzeroni LC, Poland R, Gourley GR, Kazmierczak S, et al. Predischarge screening for severe neonatal hyperbilirubinemia identifies infants who need phototherapy. J Pediatr 2013;162:477-82.e1.  Back to cited text no. 1
Keren R, Tremont K, Luan X, Cnaan A. Visual assessment of jaundice in term and late preterm infants. Arch Dis Child Fetal Neonatal Ed 2009;94:F317-22.  Back to cited text no. 2
Sarici SU, Serdar MA, Korkmaz A, Erdem G, Oran O, Tekinalp G, et al. Incidence, course, and prediction of hyperbilirubinemia in near-term and term newborns. Pediatrics 2004;113:775-80.  Back to cited text no. 3
Watchko JF, Maisels MJ. Jaundice in low birthweight infants: Pathobiology and outcome. Arch Dis Child Fetal Neonatal Ed 2003;88:F455-8.  Back to cited text no. 4
Cashore WJ. Bilirubin and jaundice in the micropremie. Clin Perinatol 2000;27:171-9, vii.  Back to cited text no. 5
Anand KJ. Pain, plasticity, and premature birth: A prescription for permanent suffering? Nat Med 2000;6:971-3.  Back to cited text no. 6
Yamanouchi I, Yamauchi Y, Igarashi I. Transcutaneous bilirubinometry: Preliminary studies of noninvasive transcutaneous bilirubin meter in the Okayama National Hospital. Pediatrics 1980;65:195-202.  Back to cited text no. 7
Ahmed M, Mostafa S, Fisher G, Reynolds TM. Comparison between transcutaneous bilirubinometry and total serum bilirubin measurements in preterm infants <35 weeks gestation. Ann Clin Biochem 2010;47(Pt 1):72-7.  Back to cited text no. 8
Schmidt ET, Wheeler CA, Jackson GL, Engle WD. Evaluation of transcutaneous bilirubinometry in preterm neonates. J Perinatol 2009;29:564-9.  Back to cited text no. 9
Siu LY, Siu LW, Au SK, Li KW, Tsui TK, Chang YY, et al. Evaluation of a transcutaneous bilirubinometer with two optical paths in Chinese preterm infants. HK J Paediatr New Ser 2010;2:132-40.  Back to cited text no. 10
American Academy of Pediatrics Subcommittee on Hyperbilirubinemia. Management of hyperbilirubinemia in the newborn infant 35 or more weeks of gestation. Pediatrics 2004;114:297-316.  Back to cited text no. 11
Nagar G, Vandermeer B, Campbell S, Kumar M. Reliability of transcutaneous bilirubin devices in preterm infants: A systematic review. Pediatrics 2013;132:871-81.  Back to cited text no. 12
Knüpfer M, Pulzer F, Braun L, Heilmann A, Robel-Tillig E, Vogtmann C. Transcutaneous bilirubinometry in preterm infants. Acta Paediatr 2001;90:899-903.  Back to cited text no. 13
Nanjundaswamy S, Petrova A, Mehta R, Hegyi T. Transcutaneous bilirubinometry in preterm infants receiving phototherapy. Am J Perinatol 2005;22:127-31.  Back to cited text no. 14
Reyes CA, Stednitz DR, Hahn C, Mutchie KD, McCullough SR, Kronberg K. Evaluation of the BiliChek being used on hyperbilirubinemic newborns undergoing home phototherapy. Arch Pathol Lab Med 2008;132:684-9.  Back to cited text no. 15
Tan KL, Dong F. Transcutaneous bilirubinometry during and after phototherapy. Acta Paediatr 2003;92:327-31.  Back to cited text no. 16
Yamauchi Y, Yamanouchi I. Transcutaneous bilirubinometry. Effect of irradiation on the skin bilirubin index. Biol Neonate 1988;54:314-9.  Back to cited text no. 17
Hegyi T, Hiatt IM, Gertner IM, Zanni R, Tolentino T. Transcutaneous bilirubinometry II. Dermal bilirubin kinetics during phototherapy. Pediatr Res 1983;17:888-91.  Back to cited text no. 18
Palmer DC, Zenner EM, Drew JH. Transcutaneous bilirubinometry: Use in Australia. Aust Paediatr J 1982;18:273-6.  Back to cited text no. 19
Jangaard K, Curtis H, Goldbloom R. Estimation of bilirubin using BiliChektrade mark, a transcutaneous bilirubin measurement device: Effects of gestational age and use of phototherapy. Paediatr Child Health 2006;11:79-83.  Back to cited text no. 20
Stillova L, Matasova K, Mikitova T, Stilla J, Kolarovszka H, Zibolen M. Evaluation of transcutaneous bilirubinometry in preterm infants of gestational age 32-34 weeks. Biomed Pap Med Fac Univ Palacky Olomouc Czech Repub 2007;151:267-71.  Back to cited text no. 21
Karolyi L, Pohlandt F, Muche R, Franz AR, Mihatsch WA. Transcutaneous bilirubinometry in very low birthweight infants. Acta Paediatr 2004;93:941-4.  Back to cited text no. 22
Namba F, Kitajima H. Utility of a new transcutaneous jaundice device with two optical paths in premature infants. Pediatr Int 2007;49:497-501.  Back to cited text no. 23
Screening jaundice with expertise. Jaundice Meter JM-103. Draeger Medical Systems, Inc. Telford, PA 18969, USA. Available from: https://www.draeger.com/Products/Content/jm103_br_9050610_en.pdf. [Last accessed on 2017 May 25].  Back to cited text no. 24
Bland JM, Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986;1:307-10.  Back to cited text no. 25
Grabenhenrich J, Grabenhenrich L, Bührer C, Berns M. Transcutaneous bilirubin after phototherapy in term and preterm infants. Pediatrics 2014;134:e1324-9.  Back to cited text no. 26
Zecca E, Barone G, De Luca D, Marra R, Tiberi E, Romagnoli C. Skin bilirubin measurement during phototherapy in preterm and term newborn infants. Early Hum Dev 2009;85:537-40.  Back to cited text no. 27


  [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5], [Figure 6]

  [Table 1], [Table 2], [Table 3]

This article has been cited by
1 Accuracy of transcutaneous bilirubinometry in the preterm infants: a comprehensive meta-analysis
Mohammad Hassan Shabuj,Jesmin Hossain,Sanjoy Dey
The Journal of Maternal-Fetal & Neonatal Medicine. 2017; : 1
[Pubmed] | [DOI]


Similar in PUBMED
  Search Pubmed for
  Search in Google Scholar for
Related articles
Access Statistics
Email Alert *
Add to My List *
* Registration required (free)

  In this article
Article Figures
Article Tables

 Article Access Statistics
    PDF Downloaded328    
    Comments [Add]    
    Cited by others 1    

Recommend this journal