Indian Journal of Paediatric Dermatology

: 2018  |  Volume : 19  |  Issue : 4  |  Page : 315--320

Does color really matter? Reliability of transcutaneous bilirubinometry in different skin-colored babies

Pearl Mary Varughese, Lalitha Krishnan, Ravichandran 
 Department of Pediatrics, Pondicherry Institute of Medical Sciences, Puducherry, India

Correspondence Address:
Dr. Pearl Mary Varughese
Department of Pediatrics, Pondicherry Institute of Medical Sciences, Kalapet, Puducherry - 605 014


Background: Transcutaneous bilirubinometry, in keeping with rapid technological advances, has come a long way as an effective tool for assessing bilirubin levels in newborns. Though the early devices showed changes due to melanin concentration, the new generation devices were based on micro-spectrometry. Color coded scales were rarely used for the comparison. Hence the primary outcome is that TcBI is more reliable in lighter skin color babies than darker skin color babies.Objective: To study the reliability of TcB in different skin color babies. Materials and Methods: The study was conducted in a tertiary newborn center from November 2014 to June 2016. The inclusion criteria included all babies above 34 weeks gestation and exclusion criteria included babies with established direct hyperbilirubinemia, neonatal septicemia, major congenital/ gastrointestinal malformations and those on phototherapy. 396 babies were recruited. At 24 hours, babies were categorized based on Fitzpatrick skin color chart. Statistical analysis was done using the ROC curves Bland Altman charts. Results: Mean TcB was found to be uniformly higher than TSB for all variables like sex, birth weight, gestational age and growth of the baby with an excellent correlation (r = 0.698-0.932). 335 babies (74.4%) were falling in the color code 3. 113 babies (25.1%) were in the color code 4 and 2 babies (0.4%) were in the color code 5. TcB correlates better in light skin tone babies (color code 3) than dark skin tone babies (color code 4) with r= 0.874 and r= 0.856 respectively. Conclusion: Though TcB overestimates; it correlates well with TSB in lighter skin tone babies than babies with darker skin tone.

How to cite this article:
Varughese PM, Krishnan L, Ravichandran. Does color really matter? Reliability of transcutaneous bilirubinometry in different skin-colored babies.Indian J Paediatr Dermatol 2018;19:315-320

How to cite this URL:
Varughese PM, Krishnan L, Ravichandran. Does color really matter? Reliability of transcutaneous bilirubinometry in different skin-colored babies. Indian J Paediatr Dermatol [serial online] 2018 [cited 2020 Nov 29 ];19:315-320
Available from:

Full Text


Neonatal hyperbilirubinemia occurs in approximately 60% of all term babies and most preterms.[1] Neonatal hyperbilirubinemia occurs when there is imbalance between the production and elimination of bilirubin, a breakdown product of hemoglobin. The production of bilirubin in neonates is 2–3 fold versus that of adults.[2] Visual assessment of jaundice by measuring the yellowness of the skin can be unreliable and can cause errors in judgment and therefore delay in treatment. Errors are more likely to occur in darkly pigmented infants.[3]

Transcutaneous bilirubinometry, in keeping with rapid technological advances, has come a long way as an effective tool for assessing bilirubin levels in newborns since its introduction about four decades ago. Complete adoption of transcutaneous bilirubin (TcB) in clinical practice is still resisted by clinicians as results vary considerably depending on the gestational age, race, and other genetic and epidemiological factors.

Although the early devices showed changes due to melanin concentration, the new generation devices were based on microspectrometry which enables determining the optical density of bilirubin, hemoglobin, and melanin in the subcutaneous layer of the newborn infant skin. Excluding the factors that interfere in the determination of bilirubin leads to measuring the optical density in the subcutaneous capillaries and tissues with greater accuracy.[4]

Despite JM 103 and BiliChek accounting for skin pigmentation in their design, the relationship between TcB and total serum bilirubin (TSB) is different in black infants. Color-coded scales were rarely used for the comparison, and more studies showed that the imprecision of the bilirubinometers mostly based on race and ethnic backgrounds. The primary outcome is that transcutaneous bilirubinometer index (TcBI) is more reliable in lighter skin color babies than darker skin color babies.


Study population

This was a prospective observational study on inborn babies more than 34 weeks gestational age from November 2014 to June 2016 in a neonatal unit of a medical college hospital in South India. The exclusion criteria included babies with established direct hyperbilirubinemia, neonatal septicemia, major congenital/gastrointestinal malformations and those started on phototherapy.

Sample size was calculated using the formula: n = 4pq/(d)2 where P = Sensitivity (80%); q = 1 − p (20%) and d= p × absolute precision. This was done for an anticipated sensitivity of 80% with absolute precision 5% and 0.05 level of significance considering the incidence of hyperbilirubinemia to be 20% and lost to follow- up as 5%. This gave a sample size of 400. The study protocol was approved by the Institutional Review Board and Ethics Committee. Written informed consent was obtained from the mothers for using their baby's de-identified data. Confidentiality was maintained throughout the study. The clinical and dimorphic profile of the mother and the baby was collected using a pro forma.

At 24 h, the babies were assigned a color code based on Fitzpatrick skin tone scale.[5] The color chart was made on a transparent sheet with each color code separated by half an inch width, and the color codes assigned for each baby were given by the principal investigating officer to avoid the chances of investigator bias.

TcB levels were estimated with Drager Jaundice Meter (JM)-105 by placing the instrument on the baby's sternum. Sternum was taken as the principal site of measurement as several studies have shown excellent correlation with TSB compared to the other sites.[6],[7] All measurements were performed with the same device according to the standard described technique. An average of three readings was taken as the TcB value. After each baby, the probe was cleaned with a sterile gauze before using for the next baby.

Approximately 1 ml of venous blood was collected in a microtainer clot activator tube for assessing TSB level under strict aseptic precautions after the mother was explained about the procedure. Special care was taken to avoid exposure of the collected samples to light. The blood samples were taken to the hospital laboratory within an hour to prevent degradation and processed. Serum bilirubin measurements were measured using the Diazo method (modified Jendrassik-Grof method) in the automated analyzer Cobas Integra 400 plus from Roche Diagnostics. Although high-performance liquid chromatography (HPLC) estimation is considered the gold standard, “delta bilirubin” or “bilirubin-albumin,” is measured by the diazo but not by the HPLC method.[8] The maximum interval of time between the transcutaneous measurement and the collection of blood for TSB was 30 min.

All babies were visually examined every 6 h on the 1st day of life by a trained physician and twice a day thereafter. At 24 h, TSB and TcB were done on all babies and later repeated as per attending clinician's discretion.

Data were entered in Microsoft Excel and analyzed using the SPSS version 20.0 for Windows software. Pearson's correlation and Bland Altman analysis were used for studying the data.


Three hundred and ninety-six babies were finally recruited after applying the inclusion and the exclusion criteria demography of the cohort is given in [Table 1].{Table 1}

The mean serum bilirubin of the entire cohort at 24 h was 6.2 ± 1.4 mg/dl, and simultaneous mean TcB value was 7.7 ± 1.4 mg/dl. Mean TcB was found to be uniformly higher than TSB for all variables such as sex, birth weight, gestational age, and growth of the baby [Table 1]. However, among all variables, there was excellent correlation between TcBI and TSB (r = 0.698–0.932).

In our study, most of the babies belonged to the Fitzpatrick skin color codes 3 and 4 with 299 babies (75.5%) in color code 3 and 95 babies (24%) in color code 4. TcBI correlates better in light-skin tone babies (color code 3) than dark-skin tone babies (color code 4) with r = 0.874 and r = 0.856, respectively [Figure 1], [Figure 2], [Figure 3], [Figure 4], [Figure 5].{Figure 1}{Figure 2}{Figure 3}{Figure 4}{Figure 5}

On the whole, TcBI seems to have a higher value than TSB. Sex and birth weight does not seem to affect the correlation. All the babies in the study were grouped based on Fitzpatrick skin color code. A total of 335 babies (74.4%) were falling in the color code 3. A total of 113 babies (25.1%) were in the color code 4, and 2 babies (0.4%) were in the color code 5.

Color code 5 was not taken into consideration as only two babies fell in this group, so correlation was not done. The area under the curve for TcBI with color code 3 is 0.931 and with color code 4 is 0.853. This shows that TcBI for lighter skin is a reliable index for estimating serum bilirubin levels than dark skin tone babies.

According to Bhutani's TSB nomogram, action cutoff points for 24 h was taken at 6 mg/dl and 7.9 mg/dl (considering the 75th and 95th percentile of the Bhutani's TSB nomogram). In color code 3, when the 75th centile cutoff of the Bhutani's TSB nomogram was taken, it showed that TcBI values above 6 mg/dl have 100% sensitivity, 84.2% specificity, negative predictive value (NPV) is 100%, and positive predictive value (PPV) is 43.3%. However, when TcBI above 7.9 mg/dl (95th centile of the Bhutani's TSB nomogram) is taken, the specificity is 62.5%, NPV is 100%, and PPV is 24.3%. In our study, when values above 8.7 mg/dl are considered, it shows sensitivity of 80.5%, specificity of 85.2%, NPV of 62.5%, and PPV of 24.3% [Table 2].{Table 2}

In color code 4 babies, when values above 6 mg/dl are considered, it shows 100% sensitivity, 84.1% specificity, 100% NPV, and 54.5% PPV. If TcBI values above 7.9 mg/dl are considered (95th percentile of Bhutani's TSB nomogram), it shows 83.3% sensitivity, 50.5% specificity, NPV of 94.1%, and PPV of 24.1%. In our study, when values above 8.9 mg/dl were considered, it shows the sensitivity of 77.8%, specificity of 82.1%, NPV of 95.1%, and PPV of 45.1%.


The accuracy of transcutaneous bilirubinometry has evolved over the years. The older versions were considered less reliable as they were influenced by skin color, gestational age, birth weight, and other factors.[9],[10] The major skin components which influence the spectral reflectance in neonate are melanin, dermal maturity, hemoglobin, and bilirubin.[11] The more recent devices are based on microspectrometry, and studies have shown better correlations with serum bilirubin levels.

El-Kabbany et al. showed highly statistically significant difference between white and brown skin color cases with a better accuracy of TcB (forehead) measurement in brown color; however, there was no statistically significant difference between them at the sternum (P > 0.05).[12] There are some studies that show that skin color have no effect on the accuracy of jaundice meter.[13],[11],[14],[15]

The correlation of skin color with TcBI has yet to be defined irrespective of race and place of origin. In the USA, with the increasing number of migrants, there is a whole diversity of people with different skin colors and ethnic backgrounds. However, in India, where the population is fairly distributed into major skin color groups, only a small minority having the extremes of color (too light skinned or too dark skinned). Color-coded scales were rarely used for the comparison. Kazmierczak et al. studied the accuracy of the BiliCheck with use of a skin tone chart provided by SpectRx and showed that skin color did not have an effect on the performance of the BiliCheck.[16] Samiee-Zafarghandy observed that TcB underestimated TSB in light and medium skin color and overestimated it in dark skin color with wide limits of agreement. Spearman's correlation between TcB and TSB was 0.93 in the total population and remained above 0.90 for all subgroups (0.95, 0.94, and 0.96 for light, medium, and dark skin, respectively).[10] The Fitzpatrick scales were used for this study; color code 3 has a higher correlation when compared to color code 4. For detecting hyperbilirubinemia in color codes 5 and 6, more studies are hence required with more babies falling in these two groups. A similar study was done in Kerala using the same skin color scale and out of the total 200 babies, 6 babies were in Group 2 (3%), 100 babies in Group 3 (50%), 83 babies in Group 4 (41.5%), and 11 babies in Group 5 (5.5%). The Groups 2 and 3 correlations were good with r values of 0.953 and 0.885, respectively, whereas the correlation in the dark-skinned babies (Group 4 and 5) was comparatively less, r values 0.815 and 0.798 respectively.[17] This was similar to the studies done that showed that darker skin tones have poorer correlation.[17],[18] Maisels showed that JM 103 can overestimate TSB in black infants and the imprecision of BiliChek increases as degree of skin pigment increases.[19]

This study was the first of its kind to be done in the South Indian population. There have been fewer studies done to assess the reliability of TcB using a skin color chart. However, it has some limitations too. First, it is not a population-based study and it represents the data of a single tertiary care hospital in South India. Second, most of the babies in the study belonged to Fitzpatrick skin color codes 3 and 4 with very few babies falling into the other categories. Finally, extreme high values of bilirubin were not analyzed in our study.


This study finds that the correlation of TcB and TSB is higher in lighter skin color babies than darker skin color babies. A study is needed with more babies in the other skin color tones to make a more accurate analysis. Transcutaneous bilirubinometry is most efficiently used as a screening test, in all babies irrespective of color, and values above age-specific risk cutoff points should be confirmed with serum bilirubin values.

Declaration of patient consent

The authors certify that they have obtained all appropriate patient consent forms. In the form, the legal guardian has given his consent for images and other clinical information to be reported in the journal. The guardian understands that names and initials will not be published and due efforts will be made to conceal patient identity, but anonymity cannot be guaranteed.

Financial support and sponsorship

This was a self- funded study.

Conflicts of interest

There are no conflicts of interest.


1Bhutani VK, Johnson L. A proposal to prevent severe neonatal hyperbilirubinemia and kernicterus. J Perinatol 2009;29 Suppl 1:S61-7.
2Gregory M, Martin C, Cloherty J. Neonatal Hyperbilirubinemia. 7th ed. London: Wolter Kluwer; 2012. p. 304-39.
3Ebbesen F, Rasmussen LM, Wimberley PD. A new transcutaneous bilirubinometer, BiliCheck, used in the neonatal Intensive Care Unit and the maternity ward. Acta Paediatr 2002;91:203-11.
4Boo NY, Ishak S. Prediction of severe hyperbilirubinaemia using the Bilicheck transcutaneous bilirubinometer. J Paediatr Child Health 2007;43:297-302.
5Fitzpatrick TB. The validity and practicality of sun-reactive skin types I through VI. Arch Dermatol 1988;124:869-71.
6Nagar G, Vandermeer B, Campbell S, Kumar M. Reliability of transcutaneous bilirubin devices in preterm infants: A systematic review. Pediatrics 2013;132:871-81.
7Kurokawa D, Nakamura H, Yokota T, Iwatani S, Morisawa T, Katayama Y, et al. Screening for hyperbilirubinemia in Japanese very low birthweight infants using transcutaneous bilirubinometry. J Pediatr 2016;168:77-810.
8Jansen PL, Peters WH, Janssens AR. Clinical value of serum bilirubin subfractionation by high-performance liquid chromatography and conventional methods in patients with primary biliary cirrhosis. J Hepatol 1986;2:485-94.
9Shah B, Gosai DD, Prajapati J. Comparison study between serum and transcutaneous bilirubin measurement with special reference to gestational age. Int J Sci Res 2013;4:2250-3.
10Samiee-Zafarghandy S, Feberova J, Williams K, Yasseen AS, Perkins SL, Lemyre B. Influence of skin colour on diagnostic accuracy of the jaundice meter JM 103 in newborns. Arch Dis Child Fetal Neonatal Ed 2014;99:F480-4.
11Bhutani VK, Gourley GR, Adler S, Kreamer B, Dalin C, Johnson LH. Noninvasive measurement of total serum bilirubin in a multiracial predischarge newborn population to assess the risk of severe hyperbilirubinemia. Pediatrics 2000;106:E17.
12El-Kabbany ZA, Toaima NN, Shedid AM. Implementation and validating transcutaneous bilirubinometry for neonates. Egyptian Pediatric Association Gazette 2017;65:38-42.
13Karen T, Bucher HU, Fauchère JC. Comparison of a new transcutaneous bilirubinometer (Bilimed) with serum bilirubin measurements in preterm and full-term infants. BMC Pediatr 2009;9:70.
14Olusanya BO, Imosemi DO, Emokpae AA. Differences between transcutaneous and serum bilirubin measurements in black African neonates. Pediatrics 2016;138. pii: e20160907.
15Slusher TM, Angyo IA, Bode-Thomas F, Akor F, Pam SD, Adetunji AA, et al. Transcutaneous bilirubin measurements and serum total bilirubin levels in indigenous African infants. Pediatrics 2004;113:1636-41.
16Kazmierczak SC, Robertson AF, Briley KP, Kreamer B, Gourley GR. Transcutaneous measurement of bilirubin in newborns: Comparison with an automated Jendrassik-Grof procedure and HPLC. Clin Chem 2004;50:433-5.
17Sujatha S, Sreenivas V, Gulvadi A, Ramara S. Accuracy of transcutaneous bilirubinometer in assessing jaundice in newborns in Indian context. Pediatr Rev Int J Pediatr Res 2017;4:270-4.
18Mahajan G, Kaushal RK, Sankhyan N, Sharma RL, Nakra M. Transcutaneous bilirubinometer in assessment of neonatal jaundice in Northern India. Indian Pediatr 2005;42:41-5.
19Maisels MJ, Ostrea EM Jr., Touch S, Clune SE, Cepeda E, Kring E, et al. Evaluation of a new transcutaneous bilirubinometer. Pediatrics 2004;113:1628-35.