Paediatrics and International Child Health

ISSN: 2046-9047 (Print) 2046-9055 (Online) Journal homepage: https://www.tandfonline.com/loi/ypch20

Validation of transcutaneous bilirubinometry

during phototherapy for detection and monitoring
of neonatal jaundice in a low-income setting

S. M. Johnson, V. Vasu, C. Marseille, C. Hill, L. Janvier, P. Toussaint & C.

Battersby

To cite this article: S. M. Johnson, V. Vasu, C. Marseille, C. Hill, L. Janvier, P. Toussaint & C.
Battersby (2019): Validation of transcutaneous bilirubinometry during phototherapy for detection
and monitoring of neonatal jaundice in a low-income setting, Paediatrics and International Child
Health, DOI: 10.1080/20469047.2019.1598126

To link to this article: https://doi.org/10.1080/20469047.2019.1598126

Published online: 11 Apr 2019.

Submit your article to this journal

Article views: 20

View Crossmark data

Full Terms & Conditions of access and use can be found at

https://www.tandfonline.com/action/journalInformation?journalCode=ypch20
PAEDIATRICS AND INTERNATIONAL CHILD HEALTH
https://doi.org/10.1080/20469047.2019.1598126

Validation of transcutaneous bilirubinometry during phototherapy for

detection and monitoring of neonatal jaundice in a low-income setting
a b
S. M. Johnson *, V. Vasu *, C. Marseillec, C. Hillc, L. Janvierc, P. Toussaintc and C. Battersby d

a
Department of Paediatrics, Queen Elizabeth Hospital, Woolwich, UK; bNeonatal Unit, East Kent Hospitals University NHS Foundation
Trust, Ashford, UK; cNeonatal Unit, Hospital Convention Baptiste d’Haiti, Cap Haïtien, Haiti; dSection of Neonatal Medicine, Imperial
College London, London, UK

ABSTRACT ARTICLE HISTORY

Background: Severe neonatal jaundice (SNJ) and the associated long-term health sequelae Received 18 November 2018
are a significant problem in low-income countries (LIC) where measurement of total serum Accepted 13 March 2019
bilirubin (TSB) is often unavailable. Transcutaneous bilirubinometry (TcB) provides the oppor- KEYWORDS
tunity for non-invasive, point-of-care monitoring. Few studies have evaluated its agreement Jaundice; transcutaneous
with TSB levels during phototherapy in LIC. bilirubinometry;
Aim: To determine agreement between TcB and TSB during phototherapy in a Haitian new- phototherapy; low-income
born population and to establish whether TcB can be safely used to guide treatment during countries; Haiti; neonates;
phototherapy when TSB is unavailable. newborn; infants
Methods: A single-centre prospective study (February to May 2017) in Cap Haïtien, northern
Haiti was undertaken. Newborns <7 days of age with clinically detected jaundice were eligible
for inclusion. A TcB device (JM-103) was used to screen for newborn jaundice along with
a parallel TSB. A strip of black tape was placed across the sternum during phototherapy and
uncovered for subsequent TcB measurements. Decisions about phototherapy treatment were
based upon UK National Institute of Clinical Excellence (NICE) threshold criteria. Paired TSB
and TcB measurements were compared using Bland–Altman methods.
Results: The final analysis included 70 parallel TSB/TcB measurements from 35 infants within
the first 5 days of life. Nineteen (54.3%) were male and 12 (34.3%) were <35 weeks. Thirty-two
(91.4%) were receiving phototherapy. There was good agreement between TSB and TcB.
Compared with TSB, TcB tended to over-estimate bilirubin (mean difference 11.1 µmol/L, 95%
CI −10.2–32.5 µmol/L). However, at higher bilirubin levels (>250 µmol/L), TcB tended to
under-estimate bilirubin compared with TSB and the difference increased.
Conclusion: In an LIC setting in which serum bilirubin testing is not commonly available, TcB
demonstrates good agreement with TSB and can be safely used to guide jaundice treatment
during phototherapy but can lead to over-treatment at lower bilirubin levels and are more
inaccurate at higher levels. For TcB levels >250 µmol, confirmation with serum bilirubin
should be performed, if available, to avoid under-estimation.

Abbreviations: LIC: low income countries; LMIC: low and middle income countries; TcB:
transcutaneous bilirubinometry; TSB: transcutaneous serum biliubin

Introduction care alternative. However, in preterm infants the accu-

racy of TcB at higher levels and during phototherapy
Neonatal jaundice is common worldwide and is usually
is uncertain, and therefore, in well resourced countries
physiological and transient in nature. However, severe
which use TcB to screen for jaundice, TcB is only
neonatal jaundice (SNJ), particularly if not detected and
recommended in infants of >35 weeks gestation and
treated promptly, can result in serious long-term adverse
without phototherapy and readings above 250 µmol/
health outcomes including sensorineural hearing loss,
L should be confirmed with a TSB [3,4].
kernicterus and choreoathetoid cerebral palsy [1].
There is a paucity of studies evaluating TcB accuracy
The burden of SNJ is disproportionately high in
during phototherapy in low income countries (LIC). Only
low- and middle-income countries (LMIC) compared
one of the studies included in a systematic review evalu-
with high-income countries (HIC) (incidence per
ating the effect of phototherapy on TcB was conducted in
10,000 live births: LMIC 244.1 vs HIC 3.7) [2].
an LIC. The review found moderate correlation between
A contributory factor is the unaffordability and lack
TcB and TSB, but lower agreement during phototherapy
of laboratory facilities for timely total serum bilirubin
and better correlation at sites shielded from photother-
(TSB) testing and, in LMIC settings, transcutaneous
apy [5]. A study in Malawi was the first to assess whether
bilirubinometry (TcB) offers a non-invasive, point-of-
TcB can be used to guide phototherapy treatment in

CONTACT C. Battersby c.battersby@imperial.ac.uk

*These authors contributed equally to this work.
© 2019 Informa UK Limited, trading as Taylor & Francis Group
2 S. M. JOHNSON ET AL.

a resource-poor setting [6]. The study included 129 infants

and concluded that TcB can be used safely during photo-
therapy in the absence of TSB, but recommended valida-
tion of TcB agreement in specific settings, population and
devices [7]. Theoretically, phototherapy can affect TcB as
it reflects the yellowish discolouration of blanched skin
[8,9]. TcB manufacturers therefore recommend that an
area of skin that can be used for measurements be cov-
ered during phototherapy [10].
Haiti is an LIC with a poor and fragmented health
system. Approximately 40% of the population lacks
access to essential health services and only 45% of all
children (12–23 months) are fully vaccinated [11].
Figure 1. A strip of black tape was placed across the neo-
Kernicterus and disability related to SNJ are thought to
nate’s sternum during phototherapy and only uncovered for
be common and infants with neonatal jaundice often TcB measurements.
present late and have limited access to TSB measure-
ment. In 2014, neonatal clinicians working in Haiti imple-
Because of limited laboratory capacity to measure
mented training and a neonatal guideline for the use of
TSB levels, a maximum of two infants and two paired
a TcB (Drager JM-103) donated to the neonatal unit. The
samples were enrolled each day. A parallel TSB blood
aim of this study was to evaluate the agreement
sample obtained by venepuncture was sent to the
between TSB and TcB levels measured from covered
laboratory. Paired TSB and TcB levels were ideally
skin during phototherapy in Haitian newborn infants.
obtained in the first few days of life as this is the time
of highest risk for significant and rapidly rising bilirubin
Methods levels. This was not possible for infants admitted from
the community or other hospitals at a later age.
Setting
This prospective, observational and pragmatic study was Laboratory methods
undertaken in a small rural hospital, Hospital Convention
Baptiste d’Haiti (HCBH), Cap Haïtien, Haiti between When possible, serum samples were analysed daily.
February and May 2017. There are approximately 1500 A photometric test for direct and total bilirubin using
deliveries a year and the neonatal unit has 12 cots with the modified Jenddrassik–Grof method (Bilirubin
around 40 admissions per month with over 80% born at D + T liquicolor) was used [13].
term. The diagnostic capacity for infants with jaundice is
severely limited. Testing for blood group incompatibility
Data collection
and glucose 6 phosphate dehydrogenase (G6PD) defi-
ciency is not routinely available. Data were entered into a Microsoft Office Excel spread-
sheet which recorded gestational age (completed weeks),
birthweight, gender, hour and age at onset of clinically
Protocol and patients detected jaundice and commencement of phototherapy.
A TcB (JM-103 Drager) guideline was implemented in
2014 with staff training and an adapted version of the Statistical analysis
United Kingdom’s National Institute for Clinical
Excellence’s (NICE) guidelines [4] for jaundice man- Each newborn was included once and contributed
agement used for clinical decision-making. TcB was one paired TcB/TSB result. A Bland–Altman difference
measured according to the manufacturer’s instruc- plot was used to measure agreement between TcB
tions and after daily calibration [12]. All infants of and TSB; bias was calculated as a mean of the differ-
any gestational age and <7 days old were eligible ences between the paired TSB and TcB values [14]. All
for inclusion. TcB was measured daily as a part of analyses were conducted in STATA version 11 [15].
routine care and phototherapy commenced on the
basis of the UK NICE threshold graphs. Newborns
Ethics approval
receiving phototherapy were nursed exposed with
gauze pads for eye protection. A strip of black tape The study was approved by the Haitian Ministry of
was placed across the infant’s chest during photother- Health. As TSB is considered the gold-standard monitor-
apy and this was lifted when TcB measurements were ing method and TcB is a non-invasive test and accepted
taken from the sternum (Figure 1). Black tape was as routine clinical care in HIC, ethics approval was not
chosen as this was locally available and affordable. necessary.

Results
Forty-three infants were enrolled into the study. Eight
were excluded as it was not possible to obtain TcB/TSB
paired measurements owing to equipment and labora-
tory failures. The final analysis included 70 TSB/TcB
measurements (paired measurements from 35 neo-
nates). The infants’ clinical characteristics are shown in
Table 1. Around two-thirds were ≥35 weeks. The exact
gestational week was not available. Of 35, 32 (91.4%)
were <5 days old and receiving phototherapy. Median
(interquartile range) time to TcB/TSB sampling was 72
(48–96) hours.
A Bland–Altman plot of TcB versus TSB demon-
strated good agreement between the methods with
only one TcB/TSB pair falling outside the mean 95%
confidence interval (Figure 2). Overall, TcB over-
estimated the bilirubin level in compared with TSB
[mean difference 11.1 µmol/L (95% CI −10.2–32.5)].
Compared with TSB, at higher bilirubin levels
(>250 µmol/L) TcB tended to underestimate biliru-
bin and the magnitude of the difference increased
(Figure 2). The mean difference between TcB and
Table 1. Infants’ clinical characteristics.
Total n = 35 n (%)
Gestation (completed weeks)
<35 12 (34.3)
≥35 23 (65.7)
Male 19 (54.4)
Receiving phototherapy at time of testing
Yes 32 (91.4)
No 3 (8.6)
Age phototherapy commenced, days
1 8 (22.9)
2 2 (5.7)
3 13 (37.2)
4 9 (25.7)
≥5 3 (8.6)
Median day of phototherapy 3 (2–4)
Birthweight, g, mean (SD) 2437.5 (856.8)
Admitted from community/other hospital 9 (26)
Figure 2. Bland–Altman plot of TcB versus TSB. The dotted
line is the mean difference (11.1 µmol/L); shaded grey areas
are bordered by the 95% lower and upper confidence
intervals.
PAEDIATRICS AND INTERNATIONAL CHILD HEALTH 3
TSB was increased in infants <35 weeks [16.4
(−24.8–57.5)] compared with infants ≥35 weeks 8.4
(−18.5–35.3).
Discussion
As far as we know, this is the first reported study in
Haitian newborns to compare TcB with TSB. Thirty-five
paired TcB/TSB readings were obtained during photo-
therapy in a single Haitian neonatal unit in which TSB is
not readily available. There was good agreement
between TSB and TcB measured from the sternum, cov-
ered with black tape during phototherapy. TcB generally
overestimates TSB but underestimates TSB at higher
levels (>250 µmol/L). Although this means that some
infants will receive unnecessary phototherapy, we believe
this is an acceptable compromise when the alternative is
unreliable visual assessment, delayed diagnosis and treat-
ment and irreversible kernicterus. Rather than impose
additional pressure for phototherapy devices, as sug-
gested by some authors [9], TcB enables clinicians to
prioritise the limited resources for those with dangerously
high bilirubin levels.
These findings have reassured us of the safety of daily
TcB screening in this unit and the protocol guiding the
use of TcB in the absence of TSB. In this study, the
Draeger JM-103 was validated but other TcB devices
with different accuracies are available. Validation of TcB
devices with TSB should be an integral part of implemen-
tation. Ongoing manufacturer validation of TcB devices in
different income settings and for different gestational
ages and ethnicities is recommended. Additional benefits
of TcB include reduced pain in the newborn and blood
sampling [5,16–19]. It also allows indirect savings such as
of manpower and laboratory resources, although the
upfront cost of TcB is expensive for most hospitals in
the low-resource setting and there is an urgent need for
more affordable equipment. There are no prevalence
data for G6PD deficiency in the newborn Haitian popula-
tion, but data on older age groups suggest a high pre-
valence (22.7%) [20]. In the absence of readily available
TSB testing, these findings confirm that TcB would also
be useful for the earlier detection of newborn jaundice in
the community setting both in the period after birth and
following discharge from hospital.
The study has several strengths and limitations. It was
possible to incorporate into routine practice covering the
sternum with a strip of black tape. Inconsistent covering
was acknowledged as a limitation in the study by Rylance
et al. [6]. As TcB reflects the yellowish discolouration of
blanched skin and subcutaneous tissue rather than intra-
vascular plasma or serum bilirubin concentration, photo-
therapy and skin pigmentation can theoretically affect
results [8,9]. For this study, the sternum was chosen as
previous studies have shown that TcB readings were not
significantly affected by the site covered (e.g. sternum or
head) [5,8] and these results confirm that covering the
4 S. M. JOHNSON ET AL.
sternum with black tape produces TcB results compar-
able to that of TSB with a smaller mean difference com-
pared with other studies [6]. The findings corroborate
studies which demonstrated that TcB significantly over-
estimates TSB in dark-skinned African neonates com-
pared with under-estimation in white neonates.
A further strength of the study is that Bland–Altman
plots were used to provide an estimation of the impreci-
sion; only two studies in the systematic review reported
this [5]. Correlation coefficients do not provide ranges
between which TcB can safely be used and a ‘moderate
correlation’ has minimal clinical application.
The main study limitation was the small study num-
bers, constrained by unreliable laboratory equipment
and electrical supplies. Only 12 infants were <35 weeks
gestation and therefore meaningful conclusions regard-
ing this age group cannot be drawn. However, the
higher mean difference between TcB and TSB in infants
of <35 weeks gestation has been reported in other
studies and may indicate the need for caution in pre-
term infants and those weighing <2.5 kg [21]. Further
studies are warranted in preterm population.
In summary, there was good agreement between
TSB and TcB measured on skin which has been covered
in dark-skinned infants receiving phototherapy. In the
absence of TSB, it is safe to use TcB to screen for and
guide jaundice management; TcB screening as a part of
routine care in LIC neonatal units in which TSB is not
available is therefore recommended. Whilst it is more
likely to over-treat when TcB is <250 µmol/L, the oppo-
site is true at higher levels and hence a laboratory TSB is
recommended when it is >250 µmol/L or in the absence
of laboratory facilities to treat with phototherapy as the
alternative outcomes are far more dangerous. Whilst it is
easy to use and implement and was well received by
local neonatal staff, the TcB device is expensive and
there is an urgent need to make it more affordable.
In the absence of TSB in low-resource settings, TcB
measured on skin which has been covered appears
both feasible and safe to guide jaundice treatment
during phototherapy. When TcB levels are >250 µmol,
confirmation with TSB should be performed, if avail-
able, to avoid under-estimation.
Acknowledgments
We acknowledge the support of Drager for the charitable dona-
tion of a Drager JM-103 transcutaneous bilirubinometer. Drager
had no involvement in the design and funding of the study or in
the interpretation of the study results.This study is dedicated in
loving memory to Noah Amos Damien Hill, 10 January to 29
March 2019.
Author Contributions
CB and VV identified the need for and conceptualised and
designed the study. SJ and CM acquired the data. VV imple-
mented the device into routine practice and LJ, PT were local
leads. SJ wrote the first draft of the manuscript. PT and CH
provided administrative and laboratory support enabling the
study objectives to be met. CB conducted the statistical ana-
lysis and critical revision of the manuscript. All individuals will
be involved in reporting and disseminating the findings in
peer-reviewed journals and at conferences.
Consent
Parental informed verbal consent was obtained to repro-
duce the photograph which is Figure 1
Disclosure statement
No potential conflict of interest was reported by the
authors.
Funding
Hope Health Action, a UK registered non-government orga-
nisation www.hopehealthaction.org provided funding. HHA
has been providing maternal and infant health care with its
local partner hospital Hospital Baptiste d’Haiti since its
establishment in 2012.
Notes on contributor
SM Johnson Paediatrician, V Vasu Consultant Neonatologist
& Honorary Senior Lecturer, C Marseille Paediatrician, C Hill
Founder and Chief Executive Officer of Hope Health Action,
L Janvier Consultant Paediatrician and Head of Service, P
Toussaint Consultant Paediatrician and Medical Director, C
Battersby Honorary Consultant Neonatologist and Clinical
Senior Lecturer.
ORCID
S. M. Johnson http://orcid.org/0000-0001-8848-9135
V. Vasu http://orcid.org/0000-0003-2606-9797
C. Battersby http://orcid.org/0000-0002-2898-553X
References
[1] Mwaniki MK, Atieno M, Lawn JE, et al. Long-term
neurodevelopmental outcomes after intrauterine
and neonatal insults: a systematic review. Lancet.
2012;379:445–452.
[2] Slusher TM, Zamora TG, Appiah D, et al. Burden of
severe neonatal jaundice: a systematic review and
meta-analysis. BMJ Paediatr Open. 2017;1:e000105.
[3] National Institute for Health and Care Excellence.
Jaundice in newborn babies under 28 days
[Internet]. [cited 2018 Aug 2]. Available from: https://
www.nice.org.uk/guidance/Cg98
[4] Rennie J, Burman-Roy S, Murphy MS. Neonatal jaun-
dice: summary of NICE guidance. BMJ. 2010;340:
c2409.
[5] Nagar G, Vandermeer B, Campbell S, et al. Effect of
phototherapy on the reliability of transcutaneous
bilirubin devices in term and near-term infants:
a systematic review and meta-analysis. Neonatology.
2016;109:203–212.

[6] Rylance S, Yan J, Molyneux E. Can transcutaneous
bilirubinometry safely guide phototherapy treatment
of neonatal jaundice in Malawi? Paediatr Int Child
Health. 2014;34:101–107.
[7] Nagar G, Vandermeer B, Campbell S, et al. Reliability
of transcutaneous bilirubin devices in preterm infants:
a systematic review. Pediatrics. 2013;132:871–881.
[8] Tan KL, Dong F. Transcutaneous bilirubinometry
during and after phototherapy. Acta Paediatr.
2003;92:327–331.
[9] Olusanya BO, Emokpae AA. Use of transcutaneous bilir-
ubin to determine the need for phototherapy in -
resource-limited settings. Neonatology.
2017;111:324–330.
[10] Philips BiliChek noninvasive bilirubin assessment
tool, 2017 [cited 2018 Jun 19]. Available from:
https://philipsproductcontent.blob.core.windows.
net/assets/20170616/f722df04b6ce4e7da
d79a7940147347a.pdf
[11] United States Agency for International Development
(USAIDS). Haiti health fact sheet. 2017. [cited 2018 Mar
14]. Available from: https://www.usaid.gov/sites/default/
files/documents/1862/FINAL_Health_March_2017.pdf
[12] Drager. Drager JM-103 instructions manual. 2017.
[cited 2014 Jan 2]. Available from: https://www.man
ualslib.com/manual/1302054/Dr-Ger-Jm-103.html
[13] Gambino SR, Golden WRC, Smith AW. Bilirubin (mod-
ified Jendrassik and Grof) – provisional. In: Meites S,
PAEDIATRICS AND INTERNATIONAL CHILD HEALTH 5
editor. Standard methods of clinical chemistry. Vol. 5.
New York, NY: Elsevier; 1965. p. 55–64.
[14] Bland MJ, Altman DG. Statistical methods for asses-
sing agreement between two methods of clinical
measurement. Lancet. 1986;327:307–310.
[15] StataCorp. Stata statistical software: release 11.
College Station, TX: StataCorp LP; 2009.
[16] Srinivas GL, Cuff CD, Ebeling MD, et al.
Transcutaneous bilirubinometry is a reliably conserva-
tive method of assessing neonatal jaundice. J Matern
Fetal Neonatal Med. 2016;29:2635–2639.
[17] Grabenhenrich J, Grabenhenrich L, Buhrer C, et al.
Transcutaneous bilirubin after phototherapy in term
and preterm infants. Pediatrics. 2014;134:1324–1329.
[18] Jenke A. Transcutaneous bilirubinometry for preterm
infants – are there more things to consider? J Pediatr.
2016;168:6–7.
[19] Mishra S, Chawla D, Agarwal R, et al. Transcutaneous
bilirubinometry reduces the need for blood sampling
in neonates with visible jaundice. Acta Paediatr.
2009;98:1916–1919.
[20] von Fricken ME, Weppelmann TA, Eaton WT, et al.
Prevalence of glucose-6-phosphate dehydrogenase
(G6PD) deficiency in the Ouest and Sud-Est depart-
ments of Haiti. Acta Trop. 2014;135:62–66.
[21] Maisels MJ, Ostrea EM Jr, Touch S, et al. Evaluation of
a new transcutaneous bilirubinometer. Pediatrics.
2004;113:1628–1635.