Original Article

Journal of Perinatology (2005) 25, 486–490. doi:10.1038/sj.jp.7211333 Published online 19 May 2005

Evaluation of a Transcutaneous Jaundice Meter Following Hospital Discharge in Term and Near-Term Neonates

Presented in part at Pediatric Academic Societies meeting, May 2, 2004, San Francisco, CA, USA.

William D Engle MD1, Gregory L Jackson MD, MBA1, Elizabeth K Stehel MD1, Dorothy M Sendelbach MD1 and M Denise Manning RN1

1Department of Pediatrics, Division of Neonatal–Perinatal Medicine, University of Texas Southwestern Medical Center at Dallas, Dallas, USA

Correspondence: William D. Engle, MD, Department of Pediatrics, University of Texas Southwestern Medical Center at Dallas, 5323 Harry Hines Blvd., Dallas, TX 75390-9063, USA





To evaluate performance of the Minolta JM-103 Jaundice Meter (JM) as a predictor of total serum bilirubin (TSB) in outpatient neonates during the first week postnatal, and to estimate the number of TSB determinations that might be avoided in clinical use.



In neonates evaluated posthospital discharge, JM and TSB results were compared using linear regression and a Bland–Altman plot, and predictive indices were calculated for various JM cutoff values. Utilizing the 2004 American Academy of Pediatrics (AAP) guidelines, the ability of JM to predict risk zone status was determined.



Overall correlation between JM and TSB was 0.77 (p<0.001; n=121). When TSB was >17 mg/dl, a cutoff value for JM of 13 mg/dl had a sensitivity of 1.0, and 50% of TSB determinations would be avoided.



JM may facilitate outpatient management of hyperbilirubinemia by reducing the number of TSB determinations required; however, it does not provide a reliable substitute for laboratory measurement of TSB.



Several groups have voiced concern regarding an apparent increase in kernicterus,1, 2, 3, 4 and the American Academy of Pediatrics (AAP) has published a Clinical Practice Guideline that addresses management of hyperbilirubinemia in infants greater than or equal to35 weeks gestation.5 Given the reported limitations of visual monitoring,6 all of these groups emphasize the importance of a more formal approach to monitoring jaundice in neonates. This process is made more difficult by the fact that, for most neonates, total serum bilirubin (TSB) is rising at the time of hospital discharge.7

The use of a transcutaneous device to estimate TSB may enhance the ability of the clinician to monitor neonates for significant hyperbilirubinemia8, 9 and some investigators10, 11 have concluded that use of the BiliCheck™ (SpectRx, Inc, Norcross, GA) is an acceptable alternative to laboratory determination of TSB. Others have questioned the accuracy of transcutaneous bilirubin (TcB) measurements, particularly in neonates with relatively high TSB.12, 13, 14, 15, 16 The majority of studies of TcB have focused on hospitalized neonates prior to peak TSB values. If reliable TcB results could be demonstrated following discharge, when higher TSB values would be anticipated, this rapid, noninvasive, point-of-care technique would be attractive to parents and useful for providers who are implementing the AAP guidelines.

The Minolta/Hill-Rom Air-Shields JM-103 Jaundice Meter™ (JM; Hatboro, PA) is a refinement of previous transcutaneous models that were considered to have limitations in dark-skinned neonates.17, 18 Jaundice Meter (JM) has been shown to be in good agreement with TSB in hospitalized term and near-term neonates.17 However, this device has not been evaluated extensively following newborn hospital discharge; in addition, prior studies have included relatively few Hispanic neonates.

The purpose of this prospective study was to compare TcB values obtained using JM with TSB values in a primarily Hispanic population evaluated following hospital discharge. In addition, we sought to estimate the number of blood tests that might be avoided with typical use of this device.



The study population consisted of term and near-term neonates who had been discharged from the Newborn Nursery of a large public hospital, Parkland Memorial Hospital in Dallas, Texas, that serves a primarily indigent Hispanic population.13 Ethnicity is determined by maternal interview, and almost all of the patients' parents are of direct Mexican heritage, based on internal demographic data. Study patients were referred for outpatient follow-up of TSB by the newborn nursery provider because of clinical jaundice prior to hospital discharge, or were jaundiced during outpatient evaluation for another reason (such as provider concern regarding the adequacy of breast feeding). All patients were evaluated during the first week postnatal in a follow-up center located within the nursery. A total of 121 neonates who had not received prior phototherapy were studied between September 2002 and August 2003. None of the parents were instructed prior to discharge to shield their newborns from sunlight. In some instances, patients were evaluated more than once; however, only the initial comparison between JM and TSB is included in this report.

Description of JM

All TcB values were determined by one investigator (MDM) using a single JM device. This JM determines the yellowness of the subcutaneous tissue by measuring the difference in optical densities for light in the blue (450 nm) and green (550 nm) wavelength regions. The device uses a xenon lamp as the light source, and the measuring probe has two optical paths,18 allowing analysis of light scattered from shallow and deep areas of the subcutaneous tissue. This design feature is intended to minimize the influence of melanin pigment and skin maturity. The device is compact, and its operation is relatively simple.19

Based on our experience with measurement at different body sites and using varying numbers of determination, we concluded that optimal performance was obtained with a single reading over the sternum (data available on request); preferential use of the sternum site is also in agreement with the results of Maisels et al. (JM)17 and Poland et al. (BiliCheck™).20 TcB was measured within 30 minutes of collection of blood for TSB determination.

Blood was drawn by heelstick and TSB determinations were performed in the clinical chemistry laboratory at Parkland. This laboratory performs approximately 5000 total bilirubin assessments per month and utilizes the diazo Jendrassik-Grof with blank method (Olympus AU600, Olympus America Inc., Melville, NY). Results of the College of American Pathologists Comprehensive Chemistry Proficiency Surveys, C-A 2003 and C-B 2003, demonstrated that the standard deviation index always was within the acceptable range (bilirubin range 1.6 to 20.2 mg/dl). In a previous study,13 TSB was measured in 26 samples in our laboratory and by high performance liquid chromatography (HPLC); the largest difference was 1.9 mg/dl, the mean absolute difference was 0.7 mg/dl, and the bias (HPLC result — laboratory result) was 0.5 mg/dl.

Data were analyzed using linear regression and a Bland–Altman plot;21 bias was calculated as the mean of the differences between the paired TSB and JM values. The ability of various JM cutoff values to predict elevated TSB values was analyzed using standard 2 times 2 table analysis. Sensitivity, specificity, and positive and negative predictive values were calculated. In each instance, the percentage of TSB determinations that might be avoided was calculated [(number of false negatives+number of true negatives)/total number of comparisons], given the assumption that in clinical practice only neonates with a JM determination greater than a chosen cutoff value (either a true-positive or a false-positive result) would have a TSB measurement. The AAP Clinical Practice Guideline hour-specific nomogram was used to identify risk zones5, 22, 23 for bilirubin values during the first week postnatal, TcB and TSB values were plotted, and predictive indices were calculated.

The Institutional Review Board at the University of Texas Southwestern Medical Center at Dallas approved this study, and informed consent was obtained for all neonates.



Demographics of the study population (n=121) are shown in Table 1. Median gestational age was 40 weeks in the study neonates; 92% were Hispanic, and 78% were either exclusively or partially breast-fed. TSB values were greater than or equal to15 mg/dl in 47% of the comparisons. There was a significant correlation between JM and TSB with R=0.77 (p<0.001), R2=0.59, adjusted R2=0.59, and standard error of estimate=1.55 mg/dl. The equation for the best-fit line was TSB=4.94+0.75JM, and F was 170.9 (p<0.001). When non-Hispanic neonates were excluded, R was 0.76 (p<0.001). A weak (R=0.36, R2=0.13) but significant (p<0.001) correlation between postnatal age and the absolute difference between JM and TSB was noted.

A Bland–Altman plot of all 121 comparisons is shown in Figure 1. The average of JM and TSB values is shown on the x-axis, and the difference is displayed on the y-axis. Since the TSB value is subtracted from the JM determination, a point below the zero line indicates that the TSB for that pair of measurements was greater than JM. TSB was greater than JM in the majority of comparisons, and the bias was -1.6 mg/dl. However, the tendency for JM to underestimate TSB was uniform throughout the range of bilirubin values studied.

Figure 1.
Figure 1 - Unfortunately we are unable to provide accessible alternative text for this. If you require assistance to access this image, please contact help@nature.com or the author

Bland–Altman plot comparing TSB and JM.

Full figure and legend (61K)

Sensitivity, specificity, and positive and negative predictive values for various JM cutoff values and TSB levels ranging from >15 to >18 mg/dl are shown in Table 2. TSB was >15, >16, >17, and >18 mg/dl in 47, 29, 13, and 9%, respectively. For TSB >17 mg/dl, sensitivity was 1.0 (i.e., no false-negative values were observed) with a JM cutoff value of 13 mg/dl; at this cutoff value, 50% of the TSB determinations would be avoided. For TSB >17 mg/dl and a JM cutoff value of 14 mg/dl, sensitivity was 0.94 (a single false-negative value, with JM=13.5 mg/dl), and 70% of TSB determinations would be avoided.

The AAP risk zone distribution of TSB values was: low-risk zone=16 (13%); low-intermediate risk zone=38 (32%); high-intermediate risk zone=39 (32%); high-risk zone=28 (23%). For the 28 TSB values in the high-risk zone, the corresponding distribution of JM values was 6 (21%) in the low-intermediate risk zone, 15 (54%) in the high-intermediate risk zone, and 7 (25%) in the high-risk zone. Results of data analysis using hour-specific risk zones are shown in Table 3. A TSB value in the high-risk zone could be predicted by a JM value in or above the high-intermediate risk zone with a sensitivity of 0.79, and 69% of TSB determinations would be avoided.



Renewed concern regarding neonatal jaundice has developed secondary to an apparent rise in the incidence of kernicterus and an increased awareness of the logistical difficulties encountered in monitoring neonates following hospital discharge. Reports describing neonates enrolled in the Pilot Kernicterus Registry24 indicate that more rigorous follow-up might have prevented devastating consequences in some cases, and several groups have advocated more frequent monitoring of neonates for hyperbilirubinemia.1, 2, 3, 4 Although visual evaluation of neonates for jaundice is a time-honored method with certain advantages, the disadvantage is that agreement with TSB is poor, even when performed by experienced nursery personnel. The revised AAP guidelines for prevention and management of hyperbilirubinemia emphasize the importance of an objective assessment of bilirubin status for all neonates prior to hospital discharge, and use of a transcutaneous device is considered to be an option for this determination. In addition, the guidelines recommend that during follow-up assessment "If there is any doubt about the degree of jaundice, the TSB or TcB should be measured."5

Previous studies of TcB have produced conflicting results, with some investigators emphasizing the limitations of this technique,14 and others reporting that TcB is reliable enough to replace laboratory measurement of TSB entirely.10, 11 Advances in design technology appear to have improved the reliability of transcutaneous monitoring in the clinical management of neonatal jaundice. Nevertheless, as noted by Rubaltelli et al.,10 it is important to remember that TSB represents a measurement of circulating bilirubin, while TcB is a measurement of tissue bilirubin. Thus, bilirubin distribution may not always be even, and may depend on variables such as rising or falling serum bilirubin. In addition, we observed a weak relationship between postnatal age and the absolute difference between the two methods. Exact numeric equivalence between TSB and TcB may be unattainable by any method.

Prior to introduction of JM, the newest device in widespread use was the BiliCheck™.25 We reported that in a primarily Hispanic population with relatively high TSBs, the BiliCheck™ tended to underestimate TSB, particularly in those neonates with the highest TSB values;13 other investigators have reported similar findings in primarily non-Hispanic populations.16 Schumacher12 noted that this device may be very useful as a screening tool, for example, in evaluating all neonates at 24 hours postnatal, when the majority of neonates will have relatively low TSB. However, he concluded that it may be less useful in neonates with relatively high TSB values.

The Minolta JM-103™ is a third-generation Minolta transcutaneous bilirubinometer. In a primarily Caucasian population, Maisels et al.17 reported that JM provided good agreement with laboratory determination of TSB, with correlations greater than or equal to0.90. A wide range of TSB values was included; however, most were <13 mg/dl. In the present study, the correlation between JM and TSB (0.77) was lower than reported by Maisels et al.;17 however, our study population was primarily Hispanic, exclusively postdischarge, and included a higher percentage of neonates with relatively high TSB values (47% were greater than or equal to15 mg/dl). Because of these differences and given the small number of non-Hispanic neonates in this study, we cannot draw a definitive conclusion regarding any possible influence of ethnicity on JM values. Despite the lower correlation coefficient, we observed no false-negative results when a JM cutoff value of 13 mg/dl was used, and the outcome of interest was a TSB >17 mg/dl. Using these criteria, and assuming that TSB would be obtained only if JM were greater than the cutoff value, the projected TSB determinations could be reduced by 50%. The reduction in estimated TSB determinations increased to 70% when a JM cutoff value of 14 mg/dl was chosen; however, one false-negative value was observed.

Using the BiliCheck™, Bhutani et al.25 reported that none of the neonates with TSB >95th percentile had a TcB measurement that was less than the 75th percentile (except for one measurement with one of the three devices used). In this study, only 1.1% of the TSB values were >15 mg/dl. Considering the AAP risk zones,5 which are based on the Bhutani nomogram,22 we observed a sensitivity of 0.79 when TcB in or above the high-intermediate risk zone was used to predict TSB in the high-risk zone (Table 3).

Although this study was not a randomized, controlled trial, our results suggest that JM may be useful in the outpatient follow-up of neonates with relatively high TSB values. JM cannot be considered a substitute for laboratory measurement of TSB, but appears to be useful as a screening tool in this population. A significant number of TSB determinations could be avoided with appropriately chosen cutoff values, and this reduction in laboratory evaluation should streamline the follow-up process for families and providers.



  1. PICK website. Available from: http://www.bilirubin.net.
  2. JCAHO. Sentinel Event Alert: Kernicterus threatens healthy newborns. http://www.jcaho.org Issue 18, April 2001;.
  3. Centers for Disease Control and Prevention. Kernicterus in full-term infants, United States, 1994−1998. MMWR Morbid Mortal Wkly Rep 2001; 50: 491−494.
  4. National Quality Forum. A national framework for healthcare quality measurement and reporting: a consensus report. Available from: http://www.qualityforum.org.
  5. 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.
  6. Tayaba R, Gribetz D, Gribetz I & Holzman IR. Noninvasive estimation of serum bilirubin. Pediatrics 1998; 102 (3) Available from: http://www.pediatrics.org/cgi/content/full/102/3/e28. | ChemPort |
  7. Maisels MJ & Newman TB. Jaundice in full-term and near-term babies who leave the hospital within 36 hours. Clin Perinatol 1998; 25: 295−302. | PubMed | ChemPort |
  8. Robertson A, Kazmierczak S & Vos P. Improved transcutaneous bilirubinometry: comparison of SpectRx BiliCheck and Minolta Jaundice Meter JM-102 for estimating total serum bilirubin in a normal newborn population. J Perinatol 2002; 22: 12−14. | Article | PubMed |
  9. Schumacher RE. Noninvasive measurements of bilirubin in the newborn. Clin Perinatol 1990; 17: 417−435. | PubMed | ChemPort |
  10. Rubaltelli FF, Gourley GR & Loskamp N et al. Transcutaneous bilirubin measurement: a multicenter evaluation of a new device. Pediatrics 2001; 107: 1264−1271. | Article | PubMed | ChemPort |
  11. Kazmierczak 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: 433−435. | Article | PubMed | ChemPort |
  12. Schumacher RE. Transcutaneous bilirubinometry and diagnostic tests: "The right job for the tool". Pediatrics 2002; 110: 407−408. | Article | PubMed |
  13. Engle WD, Jackson GL, Sendelbach D, Manning D & Frawley WH. Assessment of a transcutaneous device in the evaluation of neonatal hyperbilirubinemia in a primarily Hispanic population. Pediatrics 2002; 110: 61−67. | Article | PubMed |
  14. Wong CM, van Dijk PJE & Laing IA. A comparison of transcutaneous bilirubinometers: SpectRx BiliCheck versus Minolta AirShields. Arch Dis Child Fetal Neonatal Ed 2002; 87: F137−F140. | Article | PubMed |
  15. Briscoe L, Clark S & Yoxall CW. Can transcutaneous bilirubinometery reduce the need for blood tests in jaundiced full term babies? Arch Dis Child Fetal Neonatal Ed 2002; 86: F190−F192. | Article | PubMed | ChemPort |
  16. Nanjundaswamy S, Petrova A, Mehta R, Bernstein W & Hegyi T. The accuracy of transcutaneous bilirubin measurements in neonates: a correlation study. Biol Neonate 2004; 85: 21−25. | Article | PubMed |
  17. Maisels MJ, Ostreia EM & Cepeda E et al. Evaluation of a new transcutaneous bilirubinometer. Pediatrics 2004; 113: 1628−1635. | Article | PubMed |
  18. Yasuda S, Itoh S & Isobe K et al. New transcutaneous jaundice device with two optical paths. J Perinat Med 2003; 31: 81−88. | Article | PubMed |
  19. Jaundice Meter JM-103 Instruction Manual. Minolta Co., Ltd. 2002;.
  20. Poland RL, Hartenberger C, McHenry H & Hsi A. Comparison of skin sites for estimating serum total bilirubin in in-patients and out-patients: chest is superior to brow. J Perinatol 2004; 24: 541−543. | Article | PubMed | ChemPort |
  21. Bland JM & Altman DG. Statistical methods for assessing agreement between two methods of clinical measurement. Lancet 1986; 1: 307−310. | PubMed | ISI | ChemPort |
  22. Bhutani VK, Johnson L & Sivieri EM. Predictive ability of a pre-discharge hour-specific serum bilirubin for subsequent significant hyperbilirubinemia in healthy term and near-term newborns. Pediatrics 1999; 103: 6−14. | Article | PubMed | ChemPort |
  23. Stevenson DK, Fanaroff AA & Maisels MJ et al. Prediction of hyperbilirubinemia in near-term and term infants. Pediatrics 2001; 108: 31−39. | Article | PubMed | ChemPort |
  24. Johnson LH & Bhutani VK. System-based approach to management of neonatal jaundice and prevention of kernicterus. J Pediatr 2002; 140: 396−403. | Article | PubMed |
  25. Bhutani 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: 2/e17. | Article |


The JM-103™ used in this study was provided without charge by Hill Rom Air-Shields.